TRENDS AND ELEVATIONAL DEPENDENCE OF THE HYDROCLIMATOLOGY OF THE CARIBOO MOUNTAINS, BRITISH COLUMBIA by Aseem Raj Sharma B.Sc., Tri-Chandra Multiple Campus, 2005 M.Sc., Tribhuvan University, 2007 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN NATURAL RESOURCES AND ENVIRONMENTAL STUDIES UNIVERSITY OF NORTHERN BRITISH COLUMBIA December 2014 © Aseem Raj Sharma, 2014 UMI Number: 1526513 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. Di!ss0?t&iori Publishing UMI 1526513 Published by ProQuest LLC 2015. Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code. ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 A bstract Pristine mountain environments are more sensitive to climate change than other land surfaces. The understanding of climatic variations in mountainous terrain is still uncertain. Previous studies reveal inconsistent findings on the elevational dependency of warming in the mountains. In this study, the trends and elevational dependence o f climatic variables in the Cariboo Mountains Region (CMR) of British Columbia are explored. A high resolution 10 km x 10 km gridded data set of climate variables over the period o f 1950-2010 is used. The Mann-Kendall test is performed for evaluation o f trends and their significance. The CMR is warming at a faster rate in recent decades than regional and global warming. The minimum air temperature trend shows significant amplified warming at higher elevations. Precipitation does not show any significant trend across the study area. The possible physical mechanisms for such wanning trends and the potential impacts of these changes on the endangered mountain caribou and water resources of the area are discussed. Table o f Contents Abstract....................................................................................................................................................i Table of Contents...................................................................................................................................ii List of Tables........................................................................................................................................vi List of Figures......................................................................................................................................vii List of Appendices.............................................................................................................................. xii Dedication...........................................................................................................................................xiii Acknowledgements............................................................................................................................ xiv 1. Introduction.................................................................................................................................... 1 1.1 Overview............................................................................................................................1 1.2 Background....................................................................................................................... 3 1.2.1 Climate change and mountains.................................................................................3 1.2.2 Physical processes for climate change in mountains.............................................. 9 1.3 Goal and research questions of the thesis...................................................................... 13 1.4 Structure of the thesis......................................................................................................13 2. Study Area: The CaribooMountains Region..............................................................................15 3. Data Collection and Methods..................................................................................................... 20 3.1 Data Collection................................................................................................................ 20 3.1.1 Data Sources............................................................................................................22 3.1.1.1 The Cariboo Alpine Mesonet Observed Data...................................................... 24 3.1.1.2 ANUSPLIN Canadian Climate Coverage gridded d ata..................................... 25 3.1.2 3.2 4. Data quality assessmentand control.......................................................................27 Methods...........................................................................................................................28 3.2.1 T ools........................................................................................................................28 3.2.2 General statistics..................................................................................................... 29 3.2.3 Validation................................................................................................................ 32 3.2.4 Lapse Rate................................................................................................................32 3.2.5 Anomaly...................................................................................................................33 3.2.6 Trend Analysis........................................................................................................ 33 3.2.6.1 Mann-Kendall Trend test...................................................................................... 34 3.2.6.2 Theil-Sen trend estimate (Median based trend estimate)................................... 36 3.2.6.3 Trend Distribution................................................................................................. 37 3.2.6.4 Hydroclimatic trends with elevation....................................................................37 3.2.6.5 LOESS.................................................................................................................... 38 Results.......................................................................................................................................... 39 4.1 Observed air temperature/precipitation variations.......................................................39 4.2 Validation of Interpolated (Gridded) Data................................................................... 46 4.3 General hydroclimatology of the region...................................................................... 52 4.3.1 General statistics..................................................................................................... 52 4.3.2 Average Hydroclimatic variation......................................................................... 54 4.4 Hydroclimatic trends in the Cariboo MountainsRegion............................................. 61 4.4.1 Temporal hydroclimatic trends in the CM R......................................................... 61 iii 4.4.1.1 Annual Trends....................................................................................................... 61 4.4.1.2 Seasonal Trends..................................................................................................... 64 4.4.2 4.5 5. Spatial hydroclimatological trends in the CMR....................................................69 4.4.2.1 Annual Trends....................................................................................................... 69 4.4.2.2 Seasonal trends...................................................................................................... 73 Elevation dependence of hydroclimatic trends............................................................77 4.5.1 Elevational dependence of annual trends.............................................................. 78 4.5.2 Elevational dependence of seasonal trends........................................................... 80 4.5.3 Elevational dependence of precipitation trends....................................................82 Discussion.....................................................................................................................................85 5.1 Hydroclimatic trends......................................................................................................85 5.1.1 Magnitude and significance of linear trends.........................................................85 5.1.2 Magnitude and distribution of spatial trends........................................................86 5.2 Physical mechanisms associated with trends............................................................... 92 5.2.1 Snow-albedo feedback (SAF)................................................................................ 94 5.2.2 Clouds......................................................................................................................95 5.2.3 Soil moisture and specific humidity......................................................................96 5.2.4 Other Factors........................................................................................................... 97 5.3 Elevational dependency of hydroclimatic trends.........................................................98 5.4 Implications of trends...................................................................................................100 5.4.1 Seasonal shifting................................................................................................... 100 iv 5.4.2 5.5 Implications to the ecohydrology..........................................................................104 5.4.2.1 Impacts on water resources.................................................................................. 104 5.4.2.2 Impacts on animal species................................................................................... 106 Limitations of this study..............................................................................................107 6. Conclusions, Recommendations and Future W ork.................................................................. 109 6.1 Conclusion......................................................................................................................109 6.1.1 General Hydroclimatology.................................................................................. 109 6.1.2 Linear and spatial trends....................................................................................... 110 6.1.3 Elevational dependency of hydroclimate.............................................................112 6.1.4 Impacts of hydroclimatic variations..................................................................... 113 6.2 Recommendations and Future Work..........................................................................114 7. References................................................................................................................................. 117 8. Appendices.................................................................................................................................. 128 v List of Tables Table 1: Primary and secondary climatic effects of the basic controls of mountain climate. The effects refer to an increase in the factor listed (Barry, 1992)............................... 10 Table 2: Climatic drivers and associated mechanisms that can produce elevation sensitive air temperature responses with seasonal variations at the land surface. Tmin and Tmax are daily minimum and daily maximum air temperatures, respectively (Rangwala & Miller, 2012)...................................................................................................................... 11 Table 3: Sources and agencies collecting a.) Observed and b.) Gridded climate data in the Cariboo Mountains........................................................................................................... 22 Table f: Details of the stations used for the preliminaiy analysis............................................. 41 Table 5: Monthly lapse rate in the Cariboo Mountains..............................................................45 Table 5: Elevation of stations used for validation with elevation difference from gridded data. ............................................................................................................................................47 Table 7: Basic statistics of hydroclimatic parameters in the CMR, 1950-2010....................... 52 Table 3: Average minimum and maximum air temperatures trends over the CMR, 19502010. In parenthesis are percentage of the significant values across the domain of the CMR...................................................................................................................................86 Table 9: Possible factors responsible for elevational warming in the CMR.............................93 List of Figures Figure 1: Global air temperature trend and decadal average as given in IPCC working group I AR5 report (IPCC, 2013). The y-axis represents the temperature anomaly (°C) relative to 1961-1990..........................................................................................................4 Figure 2: The Cariboo Mountains as seen from Browntop Mountain [elevation 2031 m] 15 Figure 3: Topographic map of the Cariboo Mountains Region (CMR) and its location within BC, Canada.......................................................................................................................17 Figure 4: Weather stations (both active and inactive) operated by different weather agencies in the CMR. The elevational distribution o f these weather stations is given in Figure 5 .........................................................................................................................................................................21 Figure 5 : Distribution of weather stations with elevation operated by different agencies in the CMR. The spatial distribution and the agencies operating these stations are shown in Figure 4.............................................................................................................................. 23 Figure 6: DEMs and corresponding histograms of elevational distribution of the CMR. a. DEM and elevational distribution of CCC data with mean elevation, b. High resolution DEM (~ 1 km x 1 km) and its elevational distribution with mean elevation. Elevations are in metres a.s.l............................................................................................26 Figure 7: Location of the weather stations used for preliminary analysis of hydroclimatology of the CMR........................................................................................................................40 Figure 8: Cross correlation of daily mean air temperature among the selected stations analyzed for the study period of September 2011 to June 2012................................... 42 Figure 9: Daily air temperature variations in the Quesnel River Basin (QRB), on the western slopes of the CMR, September 2011 to June 2012........................................................ 43 Figure 10: Elevational dependence of mean monthly air temperature on the western slopes of the CMR (monthly average of September 2011 - June 2012 observational data) 44 Figure 11: Monthly total precipitation at the Likely RS and QRRC stations in the QRB that lies on the western slopes of the CMR, September 2011 to June 2012........................46 Figure 12: Monthly minimum air temperature (CCC and CAMnet) with NSE, RMSE, r, and p values.............................................................................................................................. 48 Figure 13: Monthly maximum air temperature (CCC and CAMnet) with NSE, RMSE, r, and p values.............................................................................................................................. 49 Figure 14: Comparison of mean monthly minimum and maximum air temperature between CAMnet observed and CCC gridded data for the period of 2007 - 2010 at four locations in the CMR. The solid diagonal line is the 1:1 line........................................50 Figure 15: CCC and CAMnet monthly precipitation data with NSE, RMSE, r, and p values. Gaps are due to missing observational data as remote CAMnet stations record only rainfall................................................................................................................................ 51 Figure 16: Box plots of monthly maximum (top) and minimum (bottom) air temperatures over the CMR, 1950-2010. The small white diamond in the centre shows the mean, the black line in the centre shows the median, the notch shows the 95% confidence interval of the median, the vertical line shows range of minimum and maximum values excluding outliers, and the black dot shows the outliers defined as the values greater than 1.5 times the interquartile rangeof corresponding air temperature 55 Figure 17: Mean annual minimum and maximum air temperatures and standard deviations (SD) over the CMR, 1950-2010.......................................................................................56 Figure 18: Seasonal mean minimum and maximum air temperatures and their SD over the CMR, 1950-2010...............................................................................................................57 Figure 19: Boxplot of monthly precipitation in the CMR, 1950-2010. The small white diamond in the centre shows the mean, the black line in the centre shows the median, the notch shows 95% confidence interval of the median, the vertical line shows range of minimum and maximum values excluding outliers, and the black dot shows the outliers defined as the values greater than 1.5 times the interquartile range o f the precipitation.......................................................................................................................58 Figure 20: Mean annual precipitation and variation over the CMR, 1950-2010......................59 Figure 21: Mean seasonal precipitation with seasonal variation expressed by the CV (%), 1950-2010.......................................................................................................................... 60 Figure 22: Annual minimum air temperature anomalies and trend for the CMR, 1950-2010. 61 Figure 23: Annual maximum air temperature anomalies and trend with moving average for the CMR, 1950-2010........................................................................................................ 62 Figure 24: Annual precipitation anomalies and trend with 5-year moving average in the CMR, 1950-2010...............................................................................................................63 Figure 25: Seasonal minimum air temperature anomalies and trends in the CMR, 1950-2010. ............................................................................................................................................ 64 Figure 26: Seasonal maximum air temperature anomalies and trends in the CMR, 1950-2010. 66 ix Figure 27: Seasonal total precipitation anomalies and trends over the CMR, 1950-2010...... 68 Figure 28: Minimum air temperature trends and their significance over the CMR, 1950-2010. ............................................................................................................................................ 69 Figure 29: Maximum air temperature trends and their significance over the CMR, 1950-2010. ............................................................................................................................................ 70 Figure 30: Annual total precipitation trends and their significance over the CMR, 1950-2010. ............................................................................................................................................ 72 Figure 31: Spatial variation of seasonal minimum air temperature trends over the CMR, 1950-2010.......................................................................................................................... 73 Figure 32: Spatial variation of seasonal maximum air temperature trends over the CMR, 1950-2010.......................................................................................................................... 75 Figure 33: Spatial variation of seasonal precipitation trends over the CMR, 1950-2010........76 Figure 34: Minimum air temperature trend versus elevation, 1950-2010. Each point represents the elevation of a grid cell while the solid blue line representsthe LOESS fit.........................................................................................................................................78 Figure 35: Maximum air temperature trend versus elevation, 1950-2010. Each point represents the elevation of a grid cell while the solid blue line represents the LOESS fit.........................................................................................................................................79 Figure 36: Elevational dependence of seasonal minimum air temperature trends, 1950-2010. The solid blue line shows the LOESS fit of trends with elevation................................80 Figure 37: Elevational dependence of seasonal maximum air temperature trends, 1950-2010. The solid blue line shows the LOESS fit of trends with elevation................................81 Figure 38: Annual precipitation trend versus elevation, 1950-2010. The solid blue line shows the LOESS fit of trends with elevation............................................................................83 Figure 39: Elevational dependence of seasonal precipitation trends, 1950-2010. The solid blue line shows the LOESS fit of trends with elevation................................................ 84 Figure 40: Histogram o f the distribution of air temperature trend magnitudes with mean trend (°C decade'1) and its standard deviation (SD) (°C decade'1) over the CMR, 19502010. The solid red line shows the mean trend magnitude and the solid black curve shows the Gaussian distribution of the trend magnitudes..............................................88 Figure 41: Histogram o f the distribution of annual total precipitation trend magnitudes with mean trend (mm decade'1) and its standard deviation (SD) (mm decade'1) over the CMR, 1950-2010. The solid red line shows the mean trend magnitude and the solid black curve shows the Gaussian distribution of the trend magnitudes......................... 89 Figure 42: Seasonal trend distributions for minimum and maximum air temperatures in the CMR, 1950-2010. The solid red lines show seasonal mean trends for each season. The mean (°C decade'1) and SD (°C decade'1) of each season is given. Relative magnitudes of seasonal trends are also presented...........................................................90 Figure 43: Linear trends of start of days with greater than 0°C temperature threshold for minimum, maximum, and mean air temperature (areal average) in the CMR, 19502010 101 Figure 44: Linear trend of start of freezing days with temperature threshold o f less than 0°C for minimum, maximum, and mean air temperature (areal average) over the CMR, 1950-2010....................................................................................................................... 102 xi List of Appendices Appendix A Validation of gridded data with observed data (daily frequency)...................... 128 Appendix B Spatial plots of monthly minimum and maximum air temperatures.................. 130 Appendix C Spatial plots of monthly precipitation...................................................................132 Appendix D Minimum and maximum annual temperature and trend..................................... 133 Appendix E Mean annual air temperature anomaly and trend................................................. 134 Appendix F Seasonal minimum, maximum, and mean temperatures and trends................... 135 Appendix G Spatial trend of mean annual air temperature...................................................... 138 Appendix H Spatial trend of mean seasonal temperature in the CMR, 1950-2010.............. 139 Appendix I Annual and seasonal mean temperature trends vs elevation................................ 140 Dedication I would like to respectfully dedicate this work to my late mother Harimaya Ghimire. I miss you Aama! Acknowledgements There are a large number of people without whom this thesis might not have been written, and to whom I am greatly indebted. First and foremost, I would like to thank sincerely my supervisor, Dr. Stephen Dery, for providing me an opportunity to work with him. His guidance, encouragements, and inspiration through in-depth knowledge of research and professionalism had a great influence on this research. It has expanded my horizons beyond what I imagined when I began this degree. I also thank my supervisory committee members, Dr. Peter Jackson and Dr. Margot Parkes, for their valuable comments on the proposal and thesis. In addition, thanks go to Dr. Andrew Bush, University of Alberta, for his valuable comments on this thesis. I would like to thank Dr. Ian Picketts (UNBC) and Wyatt Klopp (UNBC) for helping me proof read the thesis and Stefanie LaZerte (UNBC) who always helped me to fix R problems during my data analysis. I would like to acknowledge my colleagues and the members of the Northern Hydrometeorology Group (NHG) of UNBC: Dr. Do Hyuk Kang, Michael Allchin, Pabitra J. Gurung, Joseph B. McGrath (Ben), and James Fraser. The data for this analysis were generously provided by Dr. Alex Cannon, Pacific Climate Impacts Consortium (PCIC). I am thankful to him. Special thanks are due to the Natural Sciences and Engineering Research Council of Canada (NSERC) for financial support. I am obliged to Bunu, my wife whose love, help and care are invaluable. I could not have completed this work without a supportive family and friends, I acknowledge them. Thank you all ©. Aseem Raj Sharma xiv 1. Introduction 1.1 O verview Mountain regions are often pristine environments that play a vital role in the sustainable functioning of the earth system. It is generally accepted that mountainous regions are more sensitive to global scale climate change than other land surfaces at the same latitude (Barry, 2008; Beniston, Diaz, & Bradley, 1997; Beniston, 2003; Rangwala & Miller, 2012). There are limited studies that examine how higher elevation regions are warming compared to lower elevations. Those studies that analyze the elevational dependence o f climate change do not necessarily show consistent findings (Rangwala & Miller, 2012). Thus, there is a need to further explore how climate variables are interrelated and change with time, space, and elevation in mountainous regions. Information on how the leeward/windward aspects of mountains influence the climate variables will help to better understand their complex environment. The mechanisms that drive enhanced warming in the mountains and change in orographic precipitation will improve knowledge of mountain climate and its role in the environment as a whole. Furthermore, development of the altitudinal profiles of climate parameters for mountainous terrain will lead to a better understanding of the mountain climate and environment in an integrated approach. Mountains cover about 25% of the earth’s land surface and have substantial environmental and social significance; they are important sources of natural resources such as: freshwater, biodiversity and an essential part of the global ecosystem (Barry, 2008, 2012; Beniston, 2003). However, mountains are experiencing the compounding impacts o f climate change more than any other land surfaces (Beniston, 2003). Paucity of observation-based data in mountainous regions, the model limitations to capture the complex mountain topography, and high uncertainties on model outputs make the understanding of climate variations and trends in these regions uncertain (Bradley, Keimig, & Diaz, 2004; Luce, Abatzoglou, & Holden, 2013; Rangwala & Miller, 2012; Q. Wang, Fan, & Wang, 2013). In this context, it is important to investigate how the principal climate variables such as air temperature and precipitation differ with elevation in mountains. To further understand this complex interaction between the mountains and climate variables, the Cariboo Mountains Region (CMR) of northern British Columbia (BC) is selected as a study site. This study incorporates historical daily climatological measurements interpolated spatially from observed data from different data generating agencies such as Environment Canada. Furthermore, hydrometeorological data collected at the Cariboo Alpine Mesonet (CAMnet) sites are used to compare with interpolated data in the CMR o f northern BC. A longitudinal and elevational profile of air temperature and precipitation of the CMR is developed. Spatial variation of the CMR is of special interest because the mountains are oriented approximately Northwest (NW)-Southeast (SE) with prevailing upper winds from the west that make spatial variations prominent in the region. Windward/leeward effects and the scales at which these processes operate are explored by contrasting the meteorological conditions on the eastern and western slopes of the mountains. Long-term time series of climate data are analyzed to observe the temporal and spatial patterns of climate variability and trends over the region. Elevational dependency on the historical climate data is analyzed. The mechanisms responsible for trends of climate variables at different elevations are explored and analyzed. The potential consequences of such changes over the local ecological and hydrological processes in the region are discussed. 1.2 Background This section describes climate change and discusses the studies that explore its relation to mountains. Furthermore, different physical processes responsible for differences in mountain climate are discussed. Many contributions to the literature show enhanced warming in the higher elevation regions. This study looks at the case of the CMR, a typical mountain range of northwestern North America with limited climate studies, and explores how different physical phenomena have acted to cause hydroclimatological variations in the region. 1.2.1 Climate change and mountains It is generally accepted that the earth is warming at a faster rate in recent decades than any time during the past thousands of years. The 5th assessment report of the Intergovernmental Panel on Climate Change (IPCC, 2013) states that “ warming o f the climate system is unequivocal and since the 1950s, many o f the observed changes are unprecedented over decades to millennia, the atmosphere and ocean have warmed". O b serv e d globally a v e ra g e d com bined land an d o c e a n su rfa c e te m p e ra tu re an o m aly 1850— 2012 o.e Annual a v e ra g e 0.4 0.2 0.0 - 0 .2 - 0.4 0.6 D ecadal a v e ra g e 0 .0 - 0.6 1 8 SO Year 2000 Figure l : Global air temperature trend and decadal average as given in IPCC working group I AR5 report (IPCC, 2013). The y-axis represents the temperature anomaly (°C) relative to 1961-1990. Figure l shows the global air temperature trend and decadal averages from 1850 to 2010 as given in the IPCC AR5 report. Globally-averaged land and ocean combined temperatures show a warming trend of 0.72 [0.49 to 0.89]°C over the period 1951-2012 (Stocker et al., 2013). Recent decades, especially after the 1990s, are consecutively warmer than previous decades globally. The compounding impacts of such warming in different components of the earth system are already observed and projected to intensify. Mountain climate is still not well understood, although there are many recent advances in several areas of research in mountain climatology (Barry, 2012). Elevation, latitude, continentality, topography, and wind are the major controlling factors o f the mountain climate. How atmospheric climate variables, namely surface air temperature and precipitation are influenced by these controlling factors is still uncertain (Barry, 2008, 2012; Beniston & Fox, 1995). The relation of climate variables among each other and how they are influenced by elevation, slope, and aspect of the mountains is still not entirely clear (Pepin & Lundquist, 2008; T. Zhang, Osterkamp, & Stamnes, 1996). In this context, the CMR is selected as a study area to better understand the influence of major controls of mountains on climate variables. In particular, there are few studies on the climatology of the Cariboo Mountains of BC. An exception includes Burford, Dery, & Holmes (2009) who analyze the long term hydroclimatology of the Quesnel River Basin (QRB). They find an increasing mean minimum air temperature trend, an earlier spring freshet, and increasing river runoff in the QRB. Most of the climate studies and information available for the Cariboo Mountains are supplementary to other studies such as those on the ecology or cryosphere of the area. These studies cover either a wide range of mountains of that region or relatively small specific areas such as particular glaciers (Beedle, 2014; Burford et al., 2009; Davis & Reed, 2013; Dery, Clifton, MacLeod, & Beedle, 2010; Eyles, 1995; Hageli & McClung, 2003; Maurer et al., 2012; Tong, Dery, & Jackson, 2009a). Mountains are an important area of climate research not only because o f their presence, but also because they are not much disturbed by human influence and are a good indicator o f the state of the planet with regards to global warming (Barry, 2012; Pepin & Lundquist, 2008; Whiteman, 2000). Environmental behaviour and rapidly changing biodiversity and hydrology along with the changing elevation in the short horizontal span of mountains can be better understood with mountain climate studies (Beniston, 2003; Whiteman, 2000). Detailed studies of the changes of climatic variables in the pristine mountain environments remain limited despite their direct and indirect importance (Bradley et al., 2004). The global surface air temperature has been rising over the past century and it is expected that there will be amplified warming in mountainous regions. The extent of the impacts of these changes is highly uncertain (Rangwala & Miller, 2012). Most of the climate downscaling approaches do not capture details of the complex mountain topography, increasing the uncertainties of their projections (Bradley et al., 2004). There are studies that show variations in the warming trend with elevation (Rangwala & Miller, 2012). Some studies show the elevation dependency o f surface warming with greater warming rates at higher altitudes (Bradley et al., 2004; Ceppi, Scherrer, Fischer, & Appenzeller, 2012; Daly et al., 2008; Dong, Huang, Qu, Tao, & Fan, 2014; Ghatak, Sinsky, & Miller, 2014; Liu, Cheng, Yan, & Yin, 2009; Pepin & Lundquist, 2008; Rangwala, Miller, & Xu, 2009; Rangwala, Sinsky, & Miller, 2013; Williams, Losleben, Caine, & Greenland, 1996); however, there are other studies where either no elevation dependent warming is observed (Ceppi et al., 2012; Pepin & Lundquist, 2008; Vuille, Bradley, Werner, & Keiming, 2003; Vuille & Bradley, 2000; You et al., 2010) or even decreasing air temperature trends with elevation are reported (Beniston & Rebetez, 1996; Ceppi et al., 2012; Lu, Kang, Li, & Theakstone, 2010; Vuille & Bradley, 2000). Some of the studies that compare mountain air temperature trends with global air temperature trends find that mountains are warming faster than low-lying areas. In addition, mountain air temperature trends are higher than the same latitude Northern Hemisphere air temperature trends (Beniston et al., 1997; Liu & Chen, 2000). The Southern Hemisphere mountain regions’ air temperature trends are also higher than the trends in low lying areas (Urrutia & Vuille, 2009; Vuille et al., 2008). However, differences in spatial scales and the paucity of climate stations in mountain regions make comparison of climate trends among different land surfaces with mountains difficult. Pepin & Lundquist (2008) used comprehensive, homogeneity-adjusted air temperature records from over 1000 high elevation stations across the globe to analyze air temperature trends at high elevations. They could not find a global relationship between elevation and warming rates. However, they find that mountains show more spatially consistent air temperature trends. Furthermore, they argue that high mountains are representative indicators of the status of global warming on the earth. An annual mean air temperature analysis shows significant altitudinal amplification of warming trends compared to low-lying areas globally (Q. Wang et al., 2013). The surface mean temperature lapse rate has decreased at a rate of about 0.2°C km'1 over the last 50 years (Q. Wang et al., 2013). This indicates that higher elevations have experienced more warming compared to lower elevations, i.e. the air temperature decreases less rapidly with height than previously. A strong positive correlation, especially in the minimum air temperature and increasing elevation, is observed in the winter months on the Tibetan Plateau using the Coupled Model Intercomparison Project Phase 5 (CMIP5) by Rangwala et al. (2013). Fyfe & Flato (1999) using the Coupled Climate Model (CCM) showed that there is an elevation dependency in surface climate change in the winter and spring seasons in the Rocky Mountains. Increased air temperatures with increasing elevation is observed along the American Cordillera (Alaska to Chile) through the analysis of seven General Circulation Models (GCMs) with simulations at 2 x CO2 levels (Bradley et al., 2004). Based on limited available observational data, Beniston et al. (1997) find higher increases in daily minimum air temperature at higher altitudes, especially in Europe and Asia. Using long term observed data along a 2077 m elevational transect in the Rocky Mountain Front Range of Colorado, USA, McGuire et al. (2012) showed that warming signals are strongest at mid-elevations over the long-term (56 years) and short-term (20 years) temporal scales. Based on observational data, Rangwala, Miller, & Xu (2009) found that the Tibetan Plateau high elevation regions are warming at a greater rate than the low lying regions. Further, they suggested that a continuous warming of the atmosphere caused by an increasing greenhouse gas forcing would further increase the atmospheric water vapour content. Therefore, for most of the 21st century, they expect relatively large rates of warming over the Tibetan Plateau. Scherrer, Ceppi, Croci-Maspoli, & Appenzeller (2012) used observational data in the Swiss Alps and found that the snow-albedo feedback (SAF) is responsible for an increasing trend o f air temperature during spring in that region. An analysis based on observational data shows a general decrease o f warming trends with elevation in the Tropical Andes. Using surface data of daily maximum and minimum air temperatures, Pepin & Losleben (2002) found the local surface air temperature trends at high elevation are opposite to global increasing warming trends. Lu, Kang, Li, & Theakstone (2010) through the analysis of 46-year January mean observed air temperature data between 1000 m and 5000 m in elevation, found significant altitudinal effects of temperature warming onset time on the Tibetan Plateau. They concluded that the warming in high elevations (below 5000 m) is less sensitive than low lying nearby areas. They suggested that could be caused by high albedo and the large thermal capacity of ice/snow cover on the higher part of the plateau and a possible heat island effect over the lower part. Furthermore, many studies focused on recent warming show earlier melting of snow and significant decline in snow cover in the mountains of North America and an increased number of snow-free days in recent years (Choi, Robinson, & Kang, 2010; Shi, Dery, Groisman, & Lettenmaier, 2013; Tong, Dery, & Jackson, 2009b). Tong, Dery, & Jackson (2009b) show linkage between snow cover fraction and the hydrology of the QRB region using MODIS snow products. However, there are no specific studies on how the snow cover days and melting and freezing days are changing in the Cariboo Mountains and the surrounding region. These studies raise a concern about whether or not there exists an elevation dependency on warming. Furthermore, it is possible that some regions may be experiencing elevational warming and other regions may not be at a global level. This study, therefore, examines important questions regarding elevational dependence on warming in the CMR, including: • Is this type of warming different with elevation and the leeward and windward sides of a given mountain region? • What are the physical processes responsible for such different outcomes in the mountain environment? 1.2.2 Physical processes for climate change in mountains There are a number o f factors associated with elevation that may lead to different responses to global warming. Indeed, different physical mechanisms and processes associated with either elevational differences or sensitivity of surface warming may be responsible for enhanced air temperature trends at high elevations (Rangwala & Miller, 2012). Some of these factors are universal for all land surfaces while others are particular to the mountains. Some are seasonal mechanisms and others operate year-round (Barry, 2008; Rangwala & Miller, 2012). Physical mechanisms along with the atmospheric circulation such as the SAF, cloud formation/concentration, soil moisture, etc. may contribute to seasonal trends (Ceppi, Scherrer, Fischer, & Appenzeller, 2012). Table 1 shows a summary o f climatic effects of the basic control of the mountain climate system. The altitude, continentality, and topography are the major physical factors that affect, alone or in combinations, the climate of mountain regions and therefore make them unique and sensitive to changes. Table 1: Primary and secondary climatic effects o f the basic controls o f mountain climate. The effects refer to an increase in the factor listed (Barry, 1992). Altitude reduced air density; increased wind velocity; vapour pressure; increase precipitation (mid­ increased solar radiation receipts (cloud latitude); dependent); reduced evaporation; lower air temperatures. physiological stress. annual/diurnal air temperature range Continentality increased; snow line altitude rises. cloud and precipitation regimes modified. snowfall proportion increases; day length and solar radiation totals vary Latitude annual air temperatures seasonally. decrease. Topography spatial contrasts in solar radiation and air diurnal wind regimes; temperature regimes, and precipitation as snow cover related to a result of slope and aspect. topography. Table 2 shows some of the possible climate drivers and mechanisms responsible for enhanced air temperature responses at high elevations. This study will further explore how these drivers are acting on the micro-scale and meso-scale in the Cariboo Mountains and how they influence air temperature variations in that region at different elevations, longitudes and exposures. Table 2: Climatic drivers and associated mechanisms that can produce elevation sensitive air temperature responses with seasonal variations at the land surface. Tm,„and Tmaxare daily minimum and daily maximum air temperatures, respectively (Rangwala & Miller, 2012). Primarily spring; but also Increases surface Decreases in Snow/Ice Albedo important in winter at lower elevations, summer absorption of Increases Tmax; suppressed effect if soil moisture also increases and at higher elevations, in causes daytime insolation association with the 0°C evaporative isotherm cooling Decreases Tmax; Increases in Cloud Decreases surface All seasons but greater strongest effect Cover (Daytime) insolation effects in summer when the cloud base is low Increases Increases in Cloud Cover (Nighttime) All seasons but greater Increases Tnun downwelling effects in winter longwave radiation Increases in Increases Primarily winter; smaller Specific downwelling effects are possible in fall Humidity (q) longwave radiation and spring | H| Increases Tmin Decreases Tmax; Small increases in Tminwhen cloud Decreases surface lifetime is Increases in insolation; Dependent on seasonal Aerosols Increases cloud emissions (non-absorbing) albedo and cloud enhanced; Effect is somewhat lifetime localized to near the emission source Increases T mjn; Decreases surface Increases T max insolation but when cloud cover increases midIncreases in tropospheric heating; Aerosols Decreases albedo of (absorbing) clouds; is reduced; Dependent on seasonal Effect is emissions and insolation somewhat localized to near Decreases albedo of the emission snow on ground; source Decreases cloud cover Decreases diurnal Increases latent heat Snowmelt effects are Increases in Soil fluxes and decreases strongest in spring and Moisture sensible heat fluxes winter; rainfall effects during the day are strongest in summer temperature range; Strong Tmax - soil moisture link in summer SAF, cloud cover, soil moisture, and aerosols all play important roles for air temperature trends in elevated areas. This study identifies the major climate drivers for the CMR and discusses their role in the region. 1.3 Goal and research questions of the thesis The goal of this study is to develop the longitudinal and elevational profiles of trends in hydroclimatic parameters across the spine of the Cariboo Mountains. To achieve this goal the following research questions are addressed: 1. What are the variability and trends (annual and seasonal) of climate variables at different elevations across the spine of the Cariboo Mountains? Do leeward/windward effects of the Cariboo Mountains influence this variability? 2. Is there an enhanced climate change signal in the higher elevations of the Cariboo Mountains compared to lower elevations? If so, what are the possible climatic drivers for this enhancement? Do these enhancements differ between the leeward and windward sites? Further, this study discusses the potential consequences of these trends to the hydrological system and local ecosystems such as on the endangered mountain caribou o f the region. 1.4 Structure of the thesis This thesis is structured as follows: the first chapter introduces mountain climate research and current knowledge gaps in understanding climate change and elevational warming related processes in mountains. The basics of climate change and hydroclimatic variations and the physical mechanisms associated with mountain climate are described in this chapter. In addition, the rationale and goal of the thesis are discussed. Chapter 2 provides an introduction to the characteristics of the study area that makes it especially relevant to learn about elevational dependence of hydroclimatology. Chapter 3 describes the nature of the data, tools, and methods used for analysis. The remainder of this thesis consists of the results of analysis and discussions based on the results. Chapter 4 details the results of local climate for short periods, comparisons of observed data with interpolated data, and temporal and spatial trends, and elevation dependency of trends. Further explanations and comparison of the results with other similar studies are presented in Chapter 5. A description of the role of different physical factors responsible for the mountain climate trends is given in this chapter. Furthermore, the impacts of these trends are discussed. The first part of Chapter 6 is the concluding summary of the results and discussions while the second part proposes some recommendations and future work. 2. Study Area: The Cariboo Mountains Region The study area is the CMR of east-central BC, Canada. Located to the east o f the Quesnel Highlands, these mountains form the northern extension of the Columbia Mountains and lie between the interior plateau and the Rocky Mountain Trench in BC. Figure 2 shows a typical view of the Cariboo Mountains from Browntop Mountain. The study domain o f the CMR spans approximately 245 km in length and 44,150 km in area. The study area extends from 51°37’00” N to 53°30’00” N latitude and 119°6’00” W to 122°33’00” W longitude with elevations ranging from 330 m to the highest peak at 3,520 m average sea level (a.s.l.) at Mt. Sir Wilfrid Laurier (Figure 3). Figure 2: The Cariboo Mountains as seen from Browntop Mountain [elevation 2031 m]. Generally, the climate of BC including the CMR is influenced by the presence of mountainous regions, the coastline of the Pacific Ocean, and its location in the Northern Hemisphere (latitude, longitude). The steep elevational gradient across the Cariboo Mountains also shows variation in soils, climate, and vegetation. The geology of these mountains is composed of sedimentary and metamorphosed sedimentary rocks, dominantly quartzite and some limestone. Valleys in this region are steep sided with shallow surficial deposits (MoF, 1979). Location o f the Cariboo Mountains Region A Alberta Brrfeah Cohim&ia U s A Legend A Prince George Elevation (m) High: 3496 Mt.Sir Wilfrid Laurier j Low: 382 # Castle Creek Glaciers A Barkerville z b b* 0 25 50 I i 100 km i i I Quesnel River Basin 123°0’0"W 122°30'0“W 122°6'0”W 121"30*0*W 121°0'0"W 120°30,0’W 12000’0"W 119"30'0*W Figure 3: Topographic map o f the Cariboo Mountains Region (CM R) and its location within BC, Canada. The CMR experiences a transitional climate and is wetter than the Rocky Mountains to the east and interior plateau to the west but drier than the Coast Mountains further to the west (Beedle, Menounos, & Wheate, 2014). The Cariboo Mountains lie in the prevailing westerlies of the Northern Hemisphere mid-latitudes and form a natural obstacle to the transport of moisture from the Pacific Ocean, forcing it upward. As a result, sharp longitudinal and elevational gradients in air temperature, precipitation, and snow accumulation are observed in the CMR. Furthermore, the Northern Hydrometeorology Group (NHG) of the University of Northern British Columbia (UNBC) operates the Cariboo Alpine Mesonet (CAMnet) weather stations in the region since 2006. The hydroclimatic data generated from these stations are useful to perform short-term hydrometeorological analyses of the region. The inland temperate rainforest is unique to this region and lies in and around the CMR providing habitat to mountain caribou and many other diverse plant and animal species (ForestEthics, 2014; Stevenson et al., 2011). Cedar (Thujaplicata) and hemlock (Tsuga heterophylla) dominated forests are found in the lower, nutrient-rich valleys of the region, logdepole pine (Pinus contorta), Engelmann spruce (Picea engelmannii), and subalpine fir (Abies lasiocarpa) are found in the mid-altitudes, and only alpine meadows and lichens are found above the tree line (1700 m a.s.l.) (Dery et al., 2010). The Cariboo Mountains form habitat of the endangered mountain caribou, a mountain ecotype of woodland caribou (Rangifer tarandus caribou) (Mountain Caribou Science Team, 2005). Tributaries of the Fraser River around the CMR such as the Quesnel River are important habitats of keystone salmon species. These are the spawning habitat of keystone salmon species, especially sockeye salmon (Oncorhynchus nerka) and Chinook salmon (O. tshawysscha) (Beacham et al., 2004; Dery, Hemandez-Henriquez, Owens, Parkes, & Petticrew, 2012; English et al., 2005). The Fraser River is an important source of freshwater for downstream regions in BC’s interior and plays a significant role in its ecohydrological system (Burford et al., 2009). Thus the study of climatic variability and trends over the Cariboo Mountains would not only make a contribution to a further understanding of mountain climate, but will also have broader implications to the ecohydrology of the surrounding areas. There are many glaciers (536 in the Cariboo Mountains; Bolch, Menounos, & Wheate, 2010) such as Castle Creek, Quanstrom, and Premier glaciers, in the CMR that form the headwaters of the Fraser, the North Thompson, and the Columbia Rivers (Beedle et al., 2014; Maurer et al., 2012). The CMR has great environmental and socio-economic importance. The QRB and historical places such as Barkerville lie within the western slopes of the Cariboo Mountains. th Barkerville was the focus o f the Cariboo Gold Rush in BC during the late 19 and early 20 centuries and these days it exists as one of the famous historical ghost towns of Canada (GeoBC, 2014; Wikipedia, 2014). Some of the major settlements in and around the CMR include Likely, Horsefly, McBride, Quesnel, Wells, Clearwater, and Williams Lake. These settlements lie within the Cariboo Regional District of BC and have a total population of about 60,000 (Statistics Canada, 2014). The CMR also provides high recreational values through scenic areas such as the Bowron Lake Provincial Park, Cariboo Mountains Provincial Park, and Wells Gray Provincial Park (Figure 3) (BC Parks, 2014; Wikipedia, 2013). I 191 tV \ 3. Data Collection and Methods This chapter describes the data and methods used in analyzing hydroclimatic variations in the CMR. The first part looks at different sources o f data considered for this study and a description of the data used for analysis. The second part is about the methods used for long-term trend analyses and their statistical significance. 3.1 D ata C ollection The climate system is the fundamental driver of life on earth and the understanding of such a complex system comes through the analysis of records of climate variables at different locations (Goosse, Barriat, Lefebvre, Loutre, & Zunz, 2010; McKenney et al., 2011). Unfortunately, the challenge for climatic studies in remote regions is the lack o f sufficient weather stations to monitor the local climate (Beniston, 2006; Bradley et al., 2004). This limitation is also evident in the chosen study area of the CMR where there are few stations, especially in the higher elevations, with both long term and/or reliable quality data (Figure Weather stations in the Cariboo Mountains Region N A Z b O o" 3 O O. oCO CO lO o O' C O Fa M » 1 £& o ©. co 6 tO CN o O' CN lO Legend o © co A A CAMnet Stations Environment C anada Stations A BCRFC's Snow Pillow Stations A wihtfire Mgmt. Branch Stations Elevation (m) H igh:3496 25 0 j i 100 km 50 i L_ j i I Low:382 -w 123°0'0"W 122°30'0"W 1 2 2 W W 12r30'0"W 1 2 1 W W 120°30’0'*W 120o0'0''W 119“30’0"W Figure 4: Weather stations (both active and inactive) operated by different weather agencies in the CMR. The elevational distribution o f these weather stations is given in Figure 5. Furthermore, historical daily climate data provide information on past daily air temperature and precipitation over a selected period. Such data are particularly valuable as input for trend analysis, model simulation processes such as plant growth, fire severity, and plant phenology (McKenney et al., 2011). In this study, both observed and gridded climate data from different sources are collected and compiled. The gridded climate data are then validated using the observation-based, independent dataset before analysis. 3.1.1 Data Sources Different types of data, both observed and modelled, from different sources are collected and examined to select long-term good quality hydroclimatic data of the CMR for detailed analysis (Table 3). Detailed inspections of the observed data from various agencies (presented in Table 3) show that there are large data gaps and lack o f continuous data beyond 30 years from the present in the CMR. Table 3: Sources and agencies collecting a.) Observed and b.) Gridded climate data in the Cariboo Mountains. The Cariboo Alpine Mesonet (CAMnet) ANUSPLIN observed gridded data (Macleod & D&y, 2007) (NRCan, 2014) Environment Canada (Environment Canada, 2014) BC River Forecast Centre’s snow pillow stations (BCRFC, 2014) BC Ministry of Forests, Lands, and Natural Resource Operations’ (MLNRO) Wildfire Management Branch (WMB) (WMB, 2014) There are few operational weather stations in the CMR covering higher elevations (Figure 5). Most stations lie below 1500 m elevation and do not have long-term continuous data. Therefore, data from those stations are not used for detailed hydroclimatic trend analysis of the CMR. Distribution of weather stations with elevation[m] in the Cariboo Region 0 5 10 15 20 25 30 35 40 45 50 55 60 65 No. of Stations Figure 5: Distribution o f weather stations with elevation operated by different agencies in the CMR. The spatial distribution and the agencies operating these stations are shown in Figure 4. The source o f gridded data for the region is the Australian National University SPLINe (ANUSPLIN) observed Canadian Coverage Climate (CCC) data developed by Natural Resources Canada (NRCan). The CCC data were available at the Pacific Climate Impacts Consortium (PCIC) web portal http://medusa.pcic.uvic.ca/dataportal/bcsd downscale canada/map. They can also be obtained through special request to Natural Resources Canada (NRC, 2014). These are the interpolated data from observed station data. The CCC data are used instead of other gridded data available for the region such as ClimateWNA (T. Wang, Hamann, Spittlehouse, & Murdock, 2012) because they are available at a higher temporal resolution (daily frequency) and based on solely observed data. 3.1.1.1 The Cariboo Alpine Mesonet Observed Data To monitor the long-term, amplified impact of global warming on the hydrometeorology of the mountainous terrain of northern BC, CAMnet, a mesoscale network of weather stations, has been developed by the NHG at UNBC (Macleod & Dery, 2007). This network of stations collects different climatic variables such as air temperature, rainfall, snow depth, wind speed and direction, and atmospheric pressure at 15 minute intervals in the CMR (Macleod & Dery, 2007). The CAMnet data, aggregated to daily totals or means, are used for preliminary climatic analysis of the CMR and validation o f interpolated data. The preliminary analysis is conducted on the data from September 2011 to June 2012 because of the availability of recent data at the time of analysis. 3.1.1.2 ANUSPLIN Canadian Climate Coverage gridded data The observation-based ANUSPLIN interpolated grids developed by NRCan span all of Canada (NRCan, 2014). The hydroclimatic data of this study domain are clipped from this nation-wide CCC gridded data set. The CCC data are generated using the ANUSPLIN climate modelling software (Hopkinson et al., 2011; McKenney et al., 2011). The ANUSPLIN climate modelling method is a multidimensional, nonparametric surface fitting method that is suitable for the interpolation of climate parameters such as minimum and maximum air temperature, and precipitation. In ANUSPLIN a trivariate thin-plate smoothing splines is applied to model the complex spatial patterns associated with daily data as spatially continuous functions of longitude, latitude, and elevation (Hutchinson et al., 2009). This method can be observed as a simplification of multivariate linear regression in which a parametric model is replaced by a smooth nonparametric function (Hutchinson et al., 2009). The ANUSPLIN method depends on every data point observed that gives the robust and stable determination of dependencies on the variable, particularly in data sparse, high elevation regions (McKenney et al., 2011). This type of interpolation can account for spatially-varying dependence on elevation that has control over different physical factors and hence overall climate (Daly, 2006). The details of the ANUSPLIN interpolation can be found in Hopkinson et al. (2011), Hutchinson et al. (2009), and McKenney et al. (2011). a. o o o CO o o tO CM £ o 2000 * t C§ .2 CM 1500 * 5 5 1000 UJg > 0) Mean * 1292 m 0 20 40 60 80 100 Frequency b. § ' O CO 3000 2500 2000 1500 1000 £o- C§ . 9 CM ro > 0 UJg' 500 m r~ 0 2500 5000 7500 10000 12500 Frequency Figure 6: DEM s and corresponding histograms o f elevational distribution o f the CMR. a. DEM and elevational distribution o f CCC data with mean elevation, b. High resolution DEM (~ 1 km x 1 km) and its elevational distribution with mean elevation. Elevations are in metres a.s.l.. The Digital Elevational Model (DEM) and elevational distribution o f the ANUSPLIN data of the study domain along with high resolution DEM and elevational distribution are shown in Figure 6. The majority o f the grid cells lie in the elevations between 800 m a.s.l. to 2000 m a.s.l. with mean grid cell elevation of 1292 m for the ANUSPLIN data. For this analysis, the ANUSPLIN interpolated data (daily models) based on station data from Environment Canada are used (Hutchinson et al., 2009; McKenney et al., 2011; NRCan, 2014). The ANUSPLIN gridded data comprise daily minimum temperature, daily maximum temperature, and precipitation. The precipitation includes all form of precipitation such as snow, rain, hail, etc. These daily frequency gridded data are at the resolution of 300 arc-seconds (0.0833° or about 10 km) for the period of 1950-2010. 3.1.2 Data quality assessment and control Quality assessment of data is an important factor for hydroclimatic analysis. For CAMnet observed station data, quality assessment is carried out for individual parameters at each station through statistical quality control procedures, analysis o f frequency intervals, values, manual observations, etc. Erroneous values and outliers are detected by defining the range of data values to be considered. Missing time gaps are identified and filled with “NA” using codes developed in R (R Core Team, 2014). Data with frequency less than daily are processed for quality control and summarized to their daily average for air temperature and daily total for precipitation. Interpolated data are also assessed for their quality through visual inspection and scatter plots. Missing gaps are filled using different techniques. Air temperature data with short gaps (<1 day), had their missing values filled in by performing linear interpolation over time. Here the differences between the end points are computed and divided by the number of times for which data are missing. A given meteorological quantity is in-filled by adding the computed increment to the point at the beginning of the gap, and is done so iteratively until the end of the gap. For longer gaps (>1 day), cross-correlation is used to fill the missing data. Air temperature is calculated by correlating daily air temperatures seen at nearby stations. Missing gaps are reconstructed based on linear regressions. Precipitation data gaps are not filled because of lack of reference precipitation data from other nearby stations. During interpretation of the results, this is clearly considered. For the ANUSPLIN data, detailed inspection is conducted for missing gaps in all parameters, namely daily minimum air temperature, daily maximum air temperature, and precipitation at each grid cell within the study domain. The inspection shows that there are no missing values in the data. Therefore, the ANUSPLIN observation based interpolated CCC data are directly used for analysis after an independent validation. 3.2 3.2.1 Methods Tools All analyses/calculations in this work are performed using R and ArcGIS. R is a freely available software language and environment for statistical computing and graphics that provides a wide variety of statistical and graphical techniques through its base and associated packages: ggplot2, raster, Kendall, etc. (Hijmans, 2014; Mcleod, 2013; R Core Team, 2014; Wickham, 2009, 2011). The Environmental Systems Research Institute (ESRI)’s ArcGIS 10.1 is a powerful tool for spatial analysis and presentation such as watershed delineation, digital elevation model (DEM) analysis, area calculation, etc. (ESRI, 3.2.2 General statistics Descriptive statistical values such as means, standard deviations (SDs), and coefficient of variations (CVs) of the hydroclimatic data are calculated for seasonal and annual time series. The four seasons mentioned in this study are: winter (DecemberFebruary), spring (March-May), summer (June -August), and fall (September-November). The quality-controlled data are transferred to the R statistical software and functions and codes are executed to calculate these statistics. Means and Standard Deviations Let Xi, i = 1 , , n be the series of the hydroclimatic data then the mean (x) and standard deviation (a) of this time series are calculated as given in equations (1) and (2) respectively (Dalgaard, 2008; Storch & Zwiers, 1999). n (i) i= 1 ( 2) The percentage of the coefficient of variation (CV) of the data series is calculated as in equation (3). Correlation The Pearson’s correlation coefficient (r) is calculated between observed and modelled data to a single-valued measure of association. The Pearson’s correlation (rxy) between two series x (, i = 1 , , n and y it i = 1 , ,n with means x and y and standard deviations ax and ay respectively is given as in equation (4) (Wilks, 2011). IT=i[(*j - *)(y« - y ) ] n - iZT-iK* (4) °x°y The r is considered statistically significant ifp< 0.05 in the present work. The p ( P-value) is the probability quantifying the strength of the evidence against the null hypothesis in favour of the alternative. Moving Average Five years moving average of the hydroclimatic time series is calculated to reduce the effects of nonsystematic variations as below. Let be the time series of the hydroclimatic data, then the five years moving average is a new series obtained by taking the arithmetic mean of five consecutive values o f x t as given in equation (5) (McCuen, 2003). 5 (5) I 30 | Root mean square error (RMSEt and Nash-Sutcliffe efficiency (NSE) coefficient Root mean square error (RMSE) and Nash-Sutcliffe model efficiency (NSE) coefficient values are calculated to assess how closely the gridded data represent independent observations collected in the CMR. The RMSE is the square root o f the average squared difference between the gridded and observed hydroclimatic data pairs. Therefore, the RMSE is calculated as given in equation (6) (Wilks, 2011). n RM SE = i Y 7\ x obs,i _ x model,i)V (6) i=l where x obs are observed values and x model are gridded values at time i. The NSE coefficient measures the accuracy of a model’s prediction (Krause, Boyle, & Base, 2005; Martinec & Rango, 1989). NSE coefficients range from —oo to 1; if the NSE is closer to 1, the more accurate is the model. The NSE is calculated to see how accurately the gridded data represent observed hydroclimatic parameters in the CMR. The NSE of observed data (x obs) and gridded data (xmodei) for data series x t, i = 1 ,..........., n is given as in equation (7) (Nash & Sutcliffe, 1970). NSE = Xmodel,i) S i= lC *o6s,i — x obs,i) (7 ) 3.2.3 Validation The interpolated CCC data are validated using independent data from CAMnet weather stations operating in the Cariboo Mountains. Four CAMnet stations covering different elevations and with maximum continuous data are selected to validate the gridded data. The CAMnet 15 minute frequency data from 2007 to 2010 are averaged/summed to daily values after quality control. The corresponding data from each grid nearest to the CAMnet station are extracted from CCC time series for the same period, i.e. 2007 to 2010 for daily and monthly comparisons. This validation period is limited by data availability from both CAMnet (>2006) and CCC (< 2010) data. 3.2.4 Lapse Rate The surface lapse rate of air temperature of the CMR is calculated to assess the monthly rate of decrease of air temperature with elevation and the stability of the atmosphere for different months. The surface lapse rate, i.e. the rate of decrease of temperature with elevation at the earth’s surface, is calculated as in equation (8) (Goosse et al., 2010): where T is the temperature and AZ is the change of elevation. For multiple measurements at different elevations, an average of values between neighbouring stations is used for surface lapse rate calculations. The surface lapse rate is calculated for the period of September 2011 to June 2012 because of the availability of recent data at the time o f analysis. 3.2.5 Anomaly The anomalies of the hydroclimatic parameters are calculated so that the data are less influenced by seasonal variations and therefore, more easily comparable. An anomaly, z, is calculated by subtracting the mean of the raw data (for a given year or season) as given in equation (9) (Wilks, 2011). Let x lt i = 1 ,..........., n be the series of the hydroclimatic data than the anomaly, z t is zi = x i - x (9) where x is the mean of the x t time series. 3.2.6 Trend Analysis The daily frequency CCC gridded data are summarized to annual and seasonal totals or means for each grid cell. Temporal and spatial analysis of the summarized gridded data are performed. As a first step for the temporal analyses, the mean of all the grid cells for each year or seasons of each year is calculated. For spatial analysis, calculations are executed on each grid cell. Trend analysis of a time series considers the statistical significance and magnitude of the trend. There are different techniques in analyzing these trends. In this study, nonparametric techniques for calculating the significance and magnitude o f the trend are used. The nonparametric tests rely on fewer assumptions compared to their parametric equivalents. They are applicable in the situations where less is known about the data in question, and are robust. Therefore, the Mann-Kendall trend test is performed to test the significance of the trend while the Theil-Sen trend estimate is executed to obtain the magnitude of the trend. 3.2.6.1 Mann-Kendall Trend test To detect the change in hydroclimate of the CMR, the Mann-Kendall trend test is performed (Kendall, 1975; H. B. Mann, 1945). The Mann-Kendall trend test is a popular nonparametric alternative for testing the presence of a trend or nonstationarity o f the central tendency o f a time series (Dery, Stieglitz, McKenna, & Wood, 2005; Wilks, 2011). The Mann-Kendall trend test is based on the relative ranking of data and not on the data themselves. This robust, nonparametric test is resistant to outlier effects, influential data, and non-normal data. This test shows lower sensitivity for non-homogeneous/inconsistent data, it does not require the data sets to follow any particular distribution. This test is found to be an excellent tool for trend detection by many researchers in similar hydroclimatic studies (Burford et al., 2009; Bum & Elnur, 2002; Dery & Brown, 2007; Dery et al., 2005; Dery & Wood, 2005; Gan & Kwong, 1992; Gocic & Trajkovic, 2013; Modarres & Sarhadi, 2009; Shi et al., 2013). Let x t, i = 1 , ,n be the monotonically increasing time series with time index i, then statistics for the Mann-Kendall trend test S is n -i S = 'Yusgn(.xi+1 - Xi ) ( 10) t=l w here s g n ( Ax) = '+ l,A x > 0 0, Ax = 0 (—1, Ax < 0 (11) The statistics in equation (10) counts the number of adjacent data pairs in which the first value is smaller than the second, and subtracts the number of data pairs in which the first is larger than the second (Wilks, 2011). The sampling distribution of the test statistic 5 is approximately Gaussian. If the null hypothesis of no trend is true, then Gaussian null distribution will have zero mean. The variance of this distribution depends on whether all the x ’s are distinct, or if some are repeated values. If there are no ties, the variance of the sampling distribution of 5 is „ s n (n - l) ( 2 n + 5) Var(S) = ------- 15--------- (1 2 ) Otherwise, „ n ( n - l ) ( 2 n + 5 ) - E j =1t y f o - l ) ( 2 t ; + 5) Var(S) = ----------------------------—---------------------------- (13) where / indicates the number of groups or repeated values, and t* is the number of repeated values in the j th group. The test p-value is then evaluated using the standard normal distribution as, ( S - 1 [Var(S)]2 0, 5+ 1 ------------T, [Tar(5)]I 5 > 0 5 = 0 5 < 0 (1 4 ) Now for this distribution in equation (14), the significance of a trend at a level p is established if \ZS\ > p ( f ) , where / is the standard normal distribution (Dery et al., 2005; Wilks, 2011). A very high positive value of S is an indicator of an increasing trend, and a large negative value indicates a decreasing trend. In this study, significance level (a) = 0.05 is used; the null hypothesis of no trend is rejected if the standard normal test statistics \ZS\ > 1.96. This trend test has two parameters of importance for trend detection: a) Significance level that indicates the strength; and b) The slope magnitude estimate that indicates the direction as well as the magnitude of the trend (Bum & Elnur, 2002). The Mann-Kendall trend test does not calculate the trend magnitude; therefore, the nonparametric median based technique, Thei 1-Sen trend estimate, is implemented for the estimation of trend magnitude. 3.2.6.2 Theil-Sen trend estimate (Median based trend estimate) Trend magnitude is calculated using the Theil-Sen trend estimate (Dery et al., 2005; Mondal, Kundu, & Mukhopadhyay, 2012; Sen, 1968). Let t be time, y is the climate variable and b a constant, then the Kendall-Theil Robust Line develops a linear equation as y -m t + b (15) where m is the slope. To calculate this slope, the slopes m k for each tied group of a time series are computed as (1 6 ) where/c = 1,2, ,n (n — l) / 2 ; i = 1,2, — I; and j = 2 ,3 ,....... ,n. The intercept b is calculated by substituting the median time and y and solving for b in equation (15). The median slope of all elements m k is then taken as the slope of equation (15) that gives the trend magnitude of the time series. 3.2.6.3 Trend Distribution The distribution of a given hydroclimatic parameter’s trend magnitude of each grid cell is plotted with its normal distribution curve. Let Xi be the series of trend magnitudes for n number of grid cells, then the Gaussian distribution curve of trend magnitude is obtained as , -oo < x < 00 (17) where x is the mean of the series and cr is the standard deviation of the series (Wilks, 2011). 3.2.6.4 Hydroclimatic trends with elevation The annual and seasonal trends are plotted with respect to elevation o f each grid cell of the study area to better understand the influence o f altitude on the hydroclimatic trends. The elevations of each grid cell are extracted from the same DEM that is used during data I371 interpolation (see Figure 6). The calculated trend magnitudes for each grid cell are plotted against the elevation. This visually shows the relationship between the trends and elevation in the CMR. 3.2.6.5 LOESS Locally weighted regression (LOESS) is a means to estimate a regression through a multivariate smoothing procedure, fitting a function of the independent variables locally and in a moving fashion analogous to how a moving average is computed for a time series (Cleveland & Devlin, 1988). Let x be a point in a data series then the quadratic fit of x using other points around it, weighted by their distance from the x gives LOESS. In this study, LOESS is used to evaluate the elevational dependence of hydroclimatic parameters in the 4. Results The results of the analysis of hydroclimatic data o f the CMR are presented in this chapter. The patterns of air temperature and precipitation in the region based on observed data from different agencies are shown first. The results of the validation of gridded data with independent CAMnet data are followed by basic statistics of the hydroclimatology of the region. The basic hydroclimatic statistics are given in tabular format to provide general information on the climate of the CMR. Furthermore, annual, seasonal, and monthly air temperature and precipitation data are analyzed and presented. The results o f the validation of interpolated data are assessed with Pearson’s correlation coefficient (r) and corresponding /7-values, RMSEs, and NSE coefficients to ensure the reliability of the data. The temporal and spatial trends and trend plots for annual and seasonal long-term interpolated air temperature and precipitation data for the region are then presented and described. The last section of this chapter shows the results of elevational dependency on hydroclimatic variations and trends in the CMR. 4.1 Observed air temperature/precipitation variations The CAMnet data are used along with data from Environment Canada and the BC River Forecast Centre (BCRFC) to enhance the understanding of short-term climatic variations in the CMR. Location o f the stations used for preliminary analysis. Z b z b co m o bo ‘ <0 in o o. CO o 04 in o oo 04 Legend in Stations 1 1 Study area b S 0 i T" in Q uesnel River Basin Elevation (m) High : 3496 0 25 L_j i 50 i I i 100 km i i I Low: 382 124°0'0’W 123°0'0"W 122W W 12 1 W W 120°0'0"W 119°0'0"W Figure 7: Location o f the weather stations used for preliminary analysis o f hydroclimatology o f the CMR. Analysis of data collected by CAMnet stations as well as from different agencies such as the BCRFC’s Yanks Peak snow pillow site and Ministry of Forests, Lands and Natural Resources Operations’ (MLNRO) Likely Airport is performed. This is conducted to obtain the preliminary information on how annual climate varies near Likely, BC, located on the western slopes of the CMR (Figure 7). Table 4 shows the details of the location and elevation of the stations used for this analysis. These stations cover different elevations ranging from 744 m (Quesnel River Research Centre (QRRC)) to 2105 m (Upper Castle Creek (UCC)). Table 4: Details o f the stations used for the preliminary analysis. Latitude Longitude QRRC 52°37’06.4” N 121°35’23.5” W 744 m CAM net 15 minutes Browntop Mt. 52°42’28.0” N 121°20’02.4” W 2031 m CAM net 15 minutes Spanish Mt. 52°33’48.4” N 121°24’35.4” W 1511 m CAMnet 15 minutes 52°49’00.0” N 1 2 1 °2 r 0 0 .0 ” W 1683 m BCRFC 1 hour 5 2 °3 6 ’54.0” N 121°30’48.0” W 1046 m MFLNRO 1 hour 53°03’45 .0 ” N 120°26’04.0” W 1803 m CAMnet 15 minutes 53°02’36.0” N 120°26’ 18.0” W 2105 m CAM net 15 minutes Elevation Yanks Peak Snow Pillow Likely Airport Lower Castle Creek (LCC) Upper Castle Creek (UCC) C ro ss correlation of daily m ean air tem perature am ong the statio n s L ikely A irp o rt indicate statistically significant values Y a n k 's P e a k S p a n i s h Mt B ro w n to p Mt Figure 8: Cross correlation o f daily mean air temperature among the selected stations analyzed for the study period o f September 2011 to June 2012. Cross correlation of daily mean air temperature among the selected stations analyzed for the study period of September 2011 to June 2012 is shown in Figure 8. There is a significant (p<0.05) positive correlation among all stations within the region, illustrating that the closer the stations are to each other, the higher is the correlation in their records. The correlation is highest between stations located at higher elevations (e.g., Spanish and Browntop Mountains) or stations at lower elevations (e.g., QRRC and Likely). As the distance and elevation difference between the stations increases, the relation between climatic variables decreases. Daily air temperatures (°C) at QRB Sept2011 - June2012 Likely — Yanks Peak — 1 1 1 1 QRRC — Spanish — Browntop 1 1 1 1 1 t -------------------- -------------------- -------------------- -------------------- -------------------- -------------------- -------------------- -------------------- ------------------- .............. r Tlme[Day] Figure 9: Daily air temperature variations in the Quesnel River Basin (QRB), on the western slopes o f the CMR, September 2011 to June 2012. Time series plots are constructed for the observational data o f air temperature and precipitation in the region. The combined time series of air temperatures for the five stations show that, in general, the lower elevation stations (QRRC and Likely) record higher air temperatures than the high elevation stations (Figure 9). However, in the case of extreme low air temperatures, such as in the middle o f January and February 2012, the high elevation stations are warmer than lower elevation records (Figure 10). This may be due to air temperature inversions occurring during extreme winter days. 11-Oct 11-Sap L • ■ ■ 2000" • 1600- il-Oec ............ ^ _______ ... * 1..-) yi fy] k) ■~!... "I. 1200- A A 800- -8 -6 i 12-Fab ■ ■ ■ 1 ----- r 2 4 0 A -4 -7 -2 12-M»r I • • - 6 - 5 - 4 .......................... .......1 fx 1 i i -f -4 - i -f i Elevation (m) ■ 1— Tt----- r ..... ...... "T 10 11 -2 8 9 7 12-Jan _.............: L_ . • 2000“ • A A 800-9 ■ -7 __ ^6-7 -8 ^ ■ -4 _ -3 -5 ■ - 6 - 4 - 2 ■ 0-2.5 0.0 2.5 12-Jun 12-May 2000" • • S tation M i- V 1600" , ♦ Browntop Mt. -j—Spanish Mt. ^ A LikelyRS _T_ ■ ! '■ Yanks Peak East QRRC 1200* A 800- ■ ■ ......l— 2.5 A 5.0 7.5 4 6 8 ■ 10 12 Monthly Mean Temp(°C) Figure 10: Elevational dependence o f mean monthly air temperature on the western slopes o f the CMR (monthly average o f September 2011 - June 2012 observational data). The average monthly surface lapse rate on the western slopes of the CMR is calculated to see monthly temperature variations. The monthly average surface temperature lapse rate of the region is 4.7°C km'1(Min = 2.4°C km"1 in February, Max = 6.6°C km '1 in June)1 (Table 5). The atmosphere in the region is generally stable during winter months and 1 July and August excluded near neutral to unstable during the summer. The unstable atmosphere during the summer, especially during daytime, causes convective activity such as showers and thunderstorms. Table 5: Monthly lapse rate in the Cariboo Mountains. Average monthly lapse rate (°C km ’) 4.1 5.1 5.7 3.2 2.5 2.4 5.9 5,5 6.2 6.6 The monthly total precipitation is calculated for two nearby stations: the QRRC and Likely Airport stations. These two stations, the Likely RS and QRRC have a 302 m elevation difference (Table 4). There is a strong correlation in the measurement of monthly precipitation between the two stations (r = 0. 89,/? < 0.05) (Figure 11). Total monthly precipitation @ Likely and QRRC 120- Station Likely RS | io o 'E' o | 80- Q. ■ QRRC O « 40£ c O 2 20 - 0 - Tim e[M onths] Figure 11: Monthly total precipitation at the Likely RS and QRRC stations in the QRB that lies on the western slopes o f the CMR, September 2011 to June 2012. 4.2 Validation of Interpolated (Gridded) Data The observed data from four CAMnet stations (namely QRRC, Spanish Mountain, UCC, and Browntop Mountain) are used for the comparison and validation o f CCC interpolated data. These stations are located within the QRB on the western slopes of the CMR; the UCC station is on the spine o f the Cariboo Mountains chain. These stations cover different elevations ranging from 744 m (QRRC) to 2015 m (UCC). The elevations of the stations and corresponding gridded data and their elevation difference are given in Table 6. QRRC and UCC stations, which are at the lowest and highest elevations, respectively, show negative elevational difference with the gridded data. The Browntop Mountain and Spanish Mountain stations show positive elevational difference with gridded data. Table 6: Elevation o f stations used for validation with elevation difference from gridded data. QRRC 744 921 -177 Browntop Mountain 2031 1706 325 Spanish Mountain 1511 1084 427 UCC 2105 2220 -115 Figure 12 shows the comparison of gridded and observed monthly average daily minimum air temperature data for the four stations. The CCC minimum air temperature data are validated with independent observed data at all stations. The NSE coefficient is higher than 90% for all the stations except at Spanish Mountain (85%), and all correlations (r) are also higher than 0.9. The daily minimum air temperature comparison shows that NSE coefficient is higher than 70% for all the stations except Spanish Mountain (67%) and rvalues are higher than 0.8 (p<0.05) for all stations (Appendix A). The NSE and r-values are high at higher elevation stations such as Browntop Mountain and UCC for minimum air temperature. Therefore, the minimum air temperature at higher elevation regions o f the CMR is well captured by CCC data. CCC & CAMnet monthly Tmin. in th e CMR Canadian Climate Coverage — Cariboo Alpine Mesonet 10 - NSE= 0.9 , RMSE= 2.19 °C , r= 0.98 (p= 0 ) - 10 - S—20-I 10 - £ .8 °C , r= 0.99 (p= 0 ) NSE= 0.85 , R M S & S 5 J7 °C , r= 0.97 (p= 0 ) t-1 0 f-20- 10 - NSE= 0.95 , RMSE= 1.54 "C , r= 0.99 (p= 0 ) £ - 10 - -20-I Time[Months] Figure 12: Monthly minimum air temperature (CCC and CAMnet) with NSE, RMSE, r, and p values. Figure 13 shows the monthly average daily maximum gridded and observed air temperatures for four stations with their NSE, RMSE, and r- values. The CCC maximum air temperature data correlate well with independent observed data at all stations. The monthly maximum air temperature NSE is only 0.29 and 0.16 for Browntop and UCC stations respectively, but they are 0.97 and 0.79 for QRRC and Spanish Mountain, r-values are higher than 0.9 for all stations. The RMSEs for minimum temperature are low; however, there are up to about 6°C differences for some stations. This is likely due to the elevation difference between the station and the elevation of the grid cell nearest to the station whose data are used for validation. CCC & CAMnet monthly Tmax. in th e CMR Canadian Climate Coverage — Cariboo Alpine Mesonet 20 - 10 NSE= 0.29 , RMSE* 6.48 °C , (p = 0 3 20- 5® 10 Q. - 10-1 NSE= 0.97 , RMSE= 1.68 "C , r= 0.99 (p= 0 ) g 20S 10£ - 1 0 -|NSE= 0.79 , RMSE= 3.92 »C ,_r= 0.99 (p= 0 ) 5 20ioo- IU /V y sX NSE= 0.16 , RMSE= 6.28 "C , I '" I I.... . ..'T..... Time[Months] Figure 13: Monthly maximum air temperature (CCC and CAMnet) with NSE, RMSE, r, and p values. Figure 14 shows the correlation between CAMnet observed and CCC gridded air temperature (monthly average minimum and monthly average maximum). For minimum air temperatures, CCC and CAMnet data show a strong relationship, but interpolated values are generally underestimated from their observed values. For maximum air temperature, there is a strong correlation, but CCC data overestimate the daily maximum air temperature than observed. The gaps between the interpolated and observed data are not only due to the interpolation, but also because of elevational difference. The observed stations’ elevations are different than the elevation of nearest grid cells whose data are extracted for validation (Table 6). Correlation: Tmax. Correlation: Tmin • Browntop ♦ Browntop A qrrc * QRRC • Spanish • Spanish •4” UCC I UCC ■10 CAMnet Tmin.(Observed)(°C) -5 0 5 10 15 20 25 30 CAMnet Tmax.(Obserwed)(°C) Figure 14: Comparison o f mean monthly minimum and maximum air temperature between CAMnet observed and CCC gridded data for the period o f 2007 - 2010 at four locations in the CMR. The solid diagonal line is the 1:1 line. CCC data do not capture precipitation data as well as air temperature data. The monthly precipitation between observed and gridded data at the QRRC station exhibit a significant correlation of 0.64 (p <0.05) (Figure 15). There are many gaps in observed precipitation data for other stations, especially during winter as the remote stations do not have heated precipitation gauges. Even for QRRC, which has a heated precipitation gauge and thus more reliable observational data, the gridded data are underestimating the observed precipitation. Local convective events may be responsible for these differences in the observed and interpolated precipitation data. For example, for the summer months, the NSE, RMSE and r-values are -0.48,28.15 mm, and 0.2 (p>0.05) respectively, for QRRC while they are 0.4, 16.83 mm, and 0.87 (p<0.05) for the winter months. The comparison of daily minimum and maximum air temperatures and precipitation data is given in Appendix A. At daily scale both minimum and maximum air temperatures show high and significant correlations (r>0.88,/?<0.05) for all stations. Precipitation data do not exhibit good correlation except for QRRC. This is because of missing data for several days in almost all of these stations. CCC & CAMnet monthly Precip. in the CMR ■ Canadian Climate CoverageBCariboo Alpine Mesonet NSE= -1.44 , RMSE= 44.79 m m , r= 0.2 (p= 0.18 ) CO 3 100 50- 111111lllll 11Li.hkllhhLl.LLAhkilLll i u lI kill.LI m l E E *o NSE= 0.35 , RMSE= 18.53 m m , r= 0.64 (p= 0 ) 'S'lOOl 0 D 1 50 Q. 'O (I) nu . O Til NSE= -0.08 , RMSE= 30.96 m m , r= 0.4 (p= 0.01 ) g | .2100 O |P 50 IfaJiillkLlI i jiLllhLili 11illijlIli +* c a OH NSE= -1.52 , RMSE= 50.98 m m , r= 0.23 (p= ( 100 I I I I . | C * hlilliLILIIIIi.lllllull.llillillll,^JlIiIIil. 8 1--------- 1--------- 1--------- 1--------- 1--------- 1--------- 1--------- 1--------- 1---------r.......... 1---------- 1--------- 1--------- 1--------- 1--------- !— ^ ^ v9 ^ o ^ ^ N.^ X^ ^ ^ o f Vs ^ 0& ' f ^ ^ $ o Time[Months] Figure 15: CCC and CAM net monthly precipitation data with NSE, RMSE, r, and p values. Gaps are due to m issing observational data as remote CAM net stations record only rainfall. 4.3 4.3.1 General hydroclimatology of the region General statistics Table 7 shows the basic hydroclimatological statistics of the CMR, summarized seasonally and annually for the period of 1950-2010 using CCC data. The minimum, maximum, and mean values of each hydroclimatic parameters (namely minimum air temperature, maximum air temperature, and precipitation) averaged over the region is calculated and shown in the table. The annual and seasonal values are given with their standard deviations (SD) and coefficient of variations (CV) for precipitation and only SD for air temperature. Mean annual minimum air temperature over the region is below freezing while maximum air temperature is above freezing. During winter, both maximum and minimum air temperatures are below freezing while during summer both are above freezing. The total mean annual precipitation over the area is about 738 mm, with wet years receiving more than 1000 mm precipitation and dry years receiving less than 500 mm on average. Table 7: Basic statistics of hydroclimatic parameters in the CMR, 1950-2010. T m in. Annual -7.4°C -0.9°C -3.7°C 1.3°C NA Winter -16.3°C -9.8°C -12.8°C 1.2°C NA Spring -8.8°C -1.2°C -4.4°C 1.5°C NA Summer 2.0°C 8.0°C 5.2°C 1.2°C NA Fall -6.6°C -0.1 °C -2.8°C 1.3°C NA 521 Tmax. Tmean Prep. Annual 1.1°C 11.1°C 7.3°C 1.9°C NA Winter -7.5°C -1.5°C -3.8°C 1.3°C NA Spring -0.2°C 12.3°C 7.4°C 2.4°C NA Summer 10.8°C 23.4°C 18.4°C 2.3°C NA Fall 1.0°C 10.7°C 6.9°C 1.9°C NA Annual -3.2°C 5.8°C 1.8°C 1.6°C NA Winter -1 1.9°C -5.5°C -8.3°C 1.2°C NA Spring -4.6°C 6.4°C 1.6°C 1.9°C NA Summer 6.4°C 16.7°C 11.8°C 1.8°C NA Fall -2.9°C 5.8°C 2.1 °C 1.6°C NA Annual 425 mm 1008 mm 739 mm 140 mm 19% Winter 98 mm 294 mm 191 mm 46 mm 25% Spring 70 mm 213 mm 140 mm 32 mm 23% Summer 132 mm 277 mm 209 mm 31 mm 15% Fall 99 mm 283 mm 196 mm 44 mm 23% The minimum, maximum, mean values, and standard deviations of average hydroclimatic data are calculated for the CMR to capture the average annual and seasonal variability. The SD of mean annual, summer, fall, and spring air temperature is 2°C but for winter, it is 1°C. The SD of minimum air temperature is less than the maximum air temperature, indicating that minimum temperatures are less dispersed than the maximum temperatures in the region. The annual precipitation CV is 19%, during winter it is above 25%, while it is only 15% during summer. The precipitation in the region is less dispersed from its mean value during the summer months while it is highly dispersed during winter months. 4.3.2 Average Hydroclimatic variation Annual, seasonal, and mean monthly hydroclimatic variations of the CMR are analyzed. Average minimum air temperature in the region is below -15°C during winter months while the highest minimum air temperature is >5°C in July (Figure 16). The maximum average monthly air temperature in the region is below -5°C during winter and is nearly 20°C during summer months. Generally, for winter and spring the air temperature range is almost equal, but for fall months it is narrower. On average, there is almost a 40°C difference in air temperature between the warmest and coldest time of a year. This indicates that the region experiences extremes in air temperatures, reflecting its continentality. M onthly T e m p e ra tu re o v e r th e C M R O- O- O- Ja n F eb Mar Apr May Ju n Jul T im e [M onth] Figure 16: Box plots o f monthly maximum (top) and minimum (bottom) air temperatures over the CMR, 1950-2010. The small white diamond in the centre shows the mean, the black line in the centre shows the median, the notch shows the 95% confidence interval o f the median, the vertical line shows range o f minimum and maximum values excluding outliers, and the black dot show s the outliers defined as the values greater than 1.5 times the interquartile range o f corresponding air temperature. There are lower minimum air temperature variations during the summer, but higher variations during winter, whereas maximum air temperatures do not show much variation. Average annual spatial variation of minimum and maximum air temperatures show that higher elevation areas are below freezing or near to 0°C year-round with up to 2°C of standard deviation (Figure 17). The SD of average annual minimum air temperature (minimum = 0.94°C, maximum = 1.81°C, and mean = 1.20°C) is greater than that o f average annual maximum air temperature (minimum = 0.86°C, maximum - 1.04°C, and mean = 0.93°C). Therefore, average annual minimum air temperature variation in the region is higher than average annual maximum air temperature variation. T m in : M e a n T m ax: M ean X ' »' •121 -1 2 2 •1 2 0 •122 -1 2 1 L o n g i t u d e ( ”W ) L o n g itu d e (* W ) T m in : S D T m ax: SD -1 2 1 -1 2 0 L o n g itu d e (°W ) -122 -1 2 1 -1 2 0 -118 L o n g itu d e (° W ) Figure 17: Mean annual minimum and maximum air temperatures and standard deviations (SD) over the CMR, 1950-2010. Mean Tmin: Seasonal ( M ean Tmax: S e a s o n a l ! = : • S*** r \ - - / ............ / [ "1 Winter Summer o e I \II Jl i **9 I / r''1 L—y % Sum m er ;[ ’ \ 9wmm Fat ■as to «S K p -119 SD Tmax: S e a s o n a l SD Tmin: S easo n al | -120 Temp.[°CJ Temp. [‘‘Cl -v w n- r -119 -122 -121 L o n g itu d e (°W ) | / Fat P \ ) \ J -122 -121 -120 -119 -122 -121 L o n g itu d e (°W ) -120 -122 -119 -121 -120 -119 -122 -121 L on g itu d e ( aW ) -120 -119 Temp.[e,C] Tem p.f’C] Figure 18: Seasonal mean minimum and maximum air temperatures and their SD over the CMR, 1950-2010. 571 M onthly P re c ip itation o v e r the C M R o -1 Jan F eb Mar Apr May Ju n Jul Tim e[M onth] Figure 19: Boxplot o f monthly precipitation in the CMR, 1950-2010. The small white diamond in the centre show s the mean, the black line in the centre shows the median, the notch show s 95% confidence interval o f the median, the vertical line shows range o f minimum and maximum values excluding outliers, and the black dot shows the outliers defined as the values greater than 1.5 times the interquartile range o f the precipitation. The seasonal variation of mean seasonal minimum and maximum air temperatures show that winters are below freezing and summers are above freezing over the entire region. The air temperature variation in the region is highest during winter and lowest during summer (Figure 18). Figure 19 shows the mean monthly total precipitation over the CMR. Generally, winter and summer months receive higher precipitation amounts than spring and fall. For example, monthly mean precipitation is about 72 mm in January, while that in April is only 40 mm. Mean annual precipitation is low at lower elevations compared to higher elevations of the Cariboo Mountains; the southeastern part of the region receives more precipitation. Mean seasonal precipitation variations show that the spring and fall months receive less precipitation than winter and summer, with higher elevation regions receiving more precipitation (Figure 21). Seasonal total precipitation is highly varying during winter and summer followed by fall and spring, respectively (Figure 21). The wettest three months are June, January, and December. Most of the precipitation should be in the form of snow because winter months show the highest amounts of precipitation. Precipitation: Mean Precipitation: CV CV[%] -121 -120 Longitude (°W) -121 -120 Longitude (°W) Figure 20: Mean annual precipitation and variation over the CMR, 1950-2010. CV Prop : Seasonal Mean Prep.: Seasonal viUMar r i \ X Summar |1 Fal % 2 f x \ . i i . Utammr aiFwaaaaaawa | Frt aj 3 0 -122 -121 -120 -119 100 -122 -121 -120 -119 -122 -121 -120 -119 -122 -121 -120 -119 Longitude (°W) Longitude (°W) Prcp.|mm] cvr%i 150 200 250 25 30 35 40 Figure 21: Mean seasonal precipitation with seasonal variation expressed by the CV (%), 1950-2010. 4.4 Hydroclimatic trends in the Cariboo Mountains Region 4.4.1 Temporal hydroclimatic trends in the CMR 4.4.1.1 Annual Trends Average annual air temperature anomalies and trends are calculated and plotted for the CMR based on the CCC dataset. The annual minimum air temperature anomaly with the trend is shown in Figure 22 (the trends are given in Appendix D). Annual minimum temp, anomaly in the CMR 2 .0 - Trend= 0.32 °C/decade p = 8. le -0 6 1.0 - s- 0.5- 1 o.oO £-0.5- 3-1.0 -1.5- -2.5- - 5 yrs moving average — Median based Trend -3.0- ^Negative A nom aly! Positive Anomaly -3.5 Time [Year] Figure 22: Annual minimum air temperature anomalies and trend for the CMR, 1950-2010. A strong positive trend of 0.32°C decade'1(p<0.05) is observed in annual minimum air temperature over the 61-year period in the CMR. Although there is high fluctuation in these temperatures, the 5-year moving average shows a continuously increasing minimum air temperature in recent decades. The annual minimum air temperature anomaly shows positive anomalies for many recent years (24 out of 30 recent years) (Figure 22). The 1950s and 1960s show negative anomalies with low values (anomalies <3.0°C). High values (anomalies >1.5°C) occur in recent decades such as the 1990s and 2000s. After 2000 there are continuous positive anomalies except in 2008 and 2009 (cf. discussion in section 5.1). Annual maximum temp, anomaly in the CMR Trends 0.19 °C/decade 2.0 p O = 0.0073 1 .0 - >> ™ 0.5- jo -0.5 - < u o. E-1.0-1.5- 5 yrs moving average ” Median based Trend - 2 .0 - | Negative A nom alyH Positive Anomaly Time [Year] Figure 23: Annual maximum air temperature anomalies and trend with moving average for the CMR, 1950-2010. A significant positive trend of 0.19°C decade'1(p<0.05) is observed in annual maximum air temperature anomalies over the 61-year period in the CMR, a trend 0.13°C decade'1 less than for the minimum air temperature trend. Year to year maximum air temperature variations are high. The annual maximum air temperature anomalies show positive anomalies for many recent years (22 out of 30 recent years) (Figure 23). After 1986, most of the years show positive anomalies (as high as 2.0°C). The minimum and maximum air temperature anomalies coincided for 25 years (positive anomalies coincided 21 years and negative anomalies coincided 4 years) during the most recent 30 years. The mean air temperature trend and anomalies are given in Appendix E (cf. discussion in section 5.1). There is no specific trend in precipitation during the past six decades over the CMR. This is interesting given the significant changes in temperature in the CMR. Maximum total annual precipitation is >900 mm per year, while the minimum value is <500 mm. Annual total precipitation anomaly in the CMR 200 - Trend= -0.92 mm/decade p = 0.93 100 - _ E E CO o- E o cCO c o 5 - 100Q . o0) - 200- 5 yrs moving average— Median based Trend | -300- Vb Negative Anomaly | Positive Anomaly 1------- 1— \N Time [Year] Figure 24: Annual precipitation anomalies and trend with 5-year m oving average in the CMR, 19502010. 63 | 4.4.1.2 Seasonal Trends S easo n al Minimum Temp, anom alies and trend in th e CMR Trend= 0.62 °C /d ecad e p=0 Trend= 0.35 °C /d ecad e P“ 0 Trend= 0.1 ’C /d e ca d e p= 0.44 i------------ 1------------ 1------------ 1------------ 1------------ 1------------ 1------------ 1------------ 1------------ 1------------ 1------------ 1------------ r Tim e [Year] — 5 yrs moving average — Median b ased Trend | Negative Anomaly f l Positive Anomaly Figure 25: Seasonal minimum air temperature anomalies and trends in the CMR, 1950-2010. Figure 25 shows seasonal average daily minimum air temperature anomalies and trends. The minimum air temperature trends in winter (0.62°C decade'1,/><0.05) and spring (0.35°C decade'1, p<0.05) are significantly higher compared to summer (0.12°C decade'1, p<0.05) and fall (0.10°C decade'1, p>0.05). The seasonal temporal trends are statistically significant for all seasons except for fall. Furthermore, seasonal anomalies of minimum air temperatures show that, in recent years, seasons are much warmer (especially after the 1990s) compared to previous years. For example, out of the most recent 30 years, 21 years show positive winter anomalies and 26 years show positive spring anomalies. Winter has the highest positive anomaly of up to about 5°C in some years while summer has the lowest. S easo n al Maximum Temp, anom alies and trend in the CMR 2.5Winter o.o-2.5| I I | -5.0- 1 Trend= 0.39 °C /decade | p=0.01 2.5«*» o.ow 0 0 “ i 1* ^ - 2 .5 (0 ■ T re n d = 0 .1 5 °C /d e c a d e E g - 5 .0 - p= 0.15 (0 1 §5 2.5Ol | 0 .0-2.5T rend= 0.11 °C /decade p= 0.28 -5.0- j 2.5- o.o~ -2.5-5.0« # d ^ dS^ 1 Trend= -0.04 °C /d ecad e 1 p - 0.74 c # d?>S # c£ S Tim e [Year] — 5 y rs moving average — Median b a sed Trend H Negative Anomaly J | Positive Anomaly Figure 26: Seasonal maximum air temperature anomalies and trends in the CMR, 1950-2010. Figure 26 shows seasonal daily average maximum air temperature anomalies and their trends. Maximum seasonal air temperature trends are lower than minimum seasonal air temperature trends. Winter shows the highest significant maximum air temperature trends (0.39°C decade'1, /?<0.05) among all the seasons followed by spring (0.15°C decade'1, jC>>0.05). Unlike minimum air temperature trends (where only the fall temperature trend is not statistically-significant), maximum air temperature trends are only significant in winter (p<0.05). Furthermore, seasonal anomalies of maximum air temperatures show that winters in recent decades are warmer than before (20 years out of recent 30 years show positive anomalies). The mean seasonal air temperature anomalies and trends also show anomalies and trends similar to minimum air temperatures. Each year’s seasonal total precipitation shows varying trends among the different seasons (Figure 27). Winter precipitation decreases by about 6.9 mm decade'1 (p>0.05) even though temperature is increasing while spring (1.4 mm decade'1, p>0.05) and fall (5.1 mm decade'1, /?>0.05) show increasing trends, but none of these trends are statistically significant. Seasonal anomalies of total precipitation show that winters in recent years experience less abundant precipitation. Seasonal Precipitation anomalies and trend in the CMR Trend= -6.88 mtn/decade p = 0 .0 8 Trend= 1.38 mm/decade Trend=-1.31 m m/decade p= 0 . 7 | Trend=5.11 mm/decade p= 0.18 ■ ■ t------------ 1------------ 1------------1------------ 1------------1------------ 1------------ 1------------1------------ 1------------ 1------------1------------r Time [Year] — 5 yrs moving average — Median based Trend ^Negative Anomaly ( | Positive Anomaly Figure 27: Seasonal total precipitation anomalies and trends over the CMR, 1950-2010. 4.4.2 Spatial hydroclimatological trends in the CMR Annual and seasonal trend analyses for minimum and maximum air temperatures and precipitation are performed for each grid point (~ 10 km x 10 km) of the CMR. 4.4.2.1 Annual Trends Annual Tmin. trend in the CMR ■ m'* Trend (°Cdecade ) Significance S ig . ++++♦♦ co m • N o t sig. + ♦♦ + ♦* + ♦♦ +V 4- * ♦ + + + + ♦ + ♦ + ♦ + *♦♦ +B* + +♦♦ + ♦ ♦ ♦ + ♦ ♦ • ♦ + + ♦ + ♦++♦+♦♦♦♦♦+*♦♦♦+++ ♦*++♦♦+♦♦++++♦♦+♦++♦ 0 "O D 1 C mM -122 -121 L ongitude (°W ) -120 -119 Figure 28: Minimum air temperature trends and their significance over the CMR, 1950-2010. Annual Tmax. trend in the CMR ■'f 10 Trend (°Cdecade 1) 0.3 Significance *■ Sig. • Not sig. co m a) TJ 3 ro CM in -122 -121 -120 -119 L ongitude (°W ) Figure 29: Maximum air temperature trends and their significance over the CMR, 1950-2010. Figure 28 shows the annual spatial trend of minimum air temperatures while Figure 29 shows the spatial trend of maximum air temperature. For minimum air temperature, the minimum trend over the region is 0.08°C decade'1and the maximum trend is 0.69°C decade1. All grid cells show a significant increase of minimum air temperature except those grid cells represented by dots in the southwest, lower elevations of the CMR (Figure 28). Maximum air temperature trends range from 0.06°C decade'1to 0.30°C decade'1(cf. discussion in section 5.1.2). All grid cells show a significant increase of maximum air temperature except points represented by dots in the southeastern part of the domain (Figure 29). Spatial minimum air temperature trend magnitudes are higher than the maximum air temperature trend magnitudes. Air temperature is rising throughout the CMR with higher spatial variability in minimum air temperature trend than in the maximum air temperature trend. Furthermore, spatial trends of minimum and maximum air temperature show an almost opposite pattern of increase with elevation. The minimum air temperature trends are stronger at higher elevations and maximum air temperature trends are stronger in lower elevations. Further details of elevational dependency of air temperature trends are discussed in Results section 4.5 and Discussion section 5.3. Figure 30 shows the varying pattern of annual precipitation trends: it is increasing in some parts and decreasing in other parts of the CMR. The extreme values of annual precipitation trends range from -3 mm year'1 to +3 mm year'1but most of the grid cells are not significant ip >0.05) as represented by dots in Figure 30 (cf. discussion in section 5.1.2). Annual total precipitation trend in the CMR 10 Trend (mmdecade 1) ♦ +♦ + * ♦ + m. Significance + Sig. • Not sig. CO W a; ■o 3 4-» CD CN m -122 -12l' L ongitude (°W) -120 -1 1 9 Figure 30: Annual total precipitation trends and their significance over the CMR, 1950-2010. 4.4.2.2 Seasonal trends Spatial trends of seasonal minimum temperature Winter Spring co ^XXXXXXKXXX**ftXXMKft**(* pxXXKX»ttXXKX*KXXKtt 10 KXXXXKXKXXXMMXXX j N I I K N N X IX IIt ia ilt ll l1 IX IK II «M X K K K I« #t »t \*im** ******* VKRIM« CM lO D Summer CD s ' C O 10 CM If) -122 -121 -120 -119 -122 -121 -120 -119 Longitude (° W ) Trend (°C d ecad e1) 0.0 0.2 0.4 Significance * Sig. • Not sig. 0.6 Figure 32: Spatial variation o f seasonal maximum air temperature trends over the CMR, 1950-2010. Spatial variation of seasonal precipitation trends Winter Spring KVt t K « V H P KK»*XSXKXR*K*f t KXX co if) x>«x« xxi rx ite xi 2500 m). Higher and significant minimum air temperature trends arise clearly at higher elevations, with grid cells >2000 m a.s.l. exhibiting trends >0.5°C decade'1 (cf. discussion in section 5.3). Annual maximum Temp, trend vs elevation in the C M R ♦ ♦ | 1 0.05 , 0.10 , 0.15 , 1 0.20 0.25 Trend fC d e c a d e -1) , 0.30 r 0.35 Figure 35: Maximum air temperature trend versus elevation, 1950-2010. Each point represents the elevation o f a grid cell while the solid blue line represents the LOESS fit. Maximum air temperature trends range from about 0.05 °C decade'1to 0.30°C decade'1at different elevations. Higher and significant maximum air temperature trends are found at lower elevations. Those grid cells that are above 2000 m a.s.l. show mostly non-significant maximum air temperature trends with a magnitude of about 0.1 °C decade*1(Figure 35 and discussion in section 5.3). Generally, the areas below 1500 m a.s.l. show significant maximum air temperature trends >0.15°C decade'1. 4.5.2 Elevational dependence of seasonal trends To obtain further insights on the air temperature trends’ elevational dependency in the CMR, analyses are now conducted on a seasonal basis. S ea so n a l Tmin. trend vs elevation in the CMR R -0.61 n= 844 ♦ S ig .tre n d ♦ N o t s i g . tr e n d Figure 36: Elevational dependence o f seasonal minimum air temperature trends, 1950-2010. The solid blue line shows the LOESS fit o f trends with elevation. Seasonal Tmax. trend vs elevation in the CMR ♦ Sig.trend • Not sig. trend Trend [“ Cdecade ] Figure 37: Elevational dependence o f seasonal maximum air temperature trends, 1950-2010. The solid blue line shows the LOESS fit o f trends with elevation. Figures 36 and 37 show the elevational dependency of seasonal minimum and maximum air temperature trends in the CMR over 1950-2010. The blue solid lines are LOESS fits that show the general relation between elevation and air temperature trend magnitude. For minimum air temperature, winter, spring, and summer trends are greater at higher elevations. Minimum air temperature trend magnitude above 2500 m a.s.l. reaches up to 1.3°C decade'1during winter, while spring exhibits trends above 0.7°C decade'1and summer near 0.5°C decade1. For most of the grid cells, the minimum air temperature trend is not significant for fall; nevertheless three of the grid cells that lie above 2000 m a.s.l. show a significant air temperature trend of about 0.3-0.4°C decade'1. For maximum air temperature, winter trends are mostly significant but do not show a relationship with elevation. The spring, summer, and fall trends decrease with elevation, but only part of spring and summer trends are statistically significant. The mean seasonal air temperature versus elevation plots are given in Appendix I (cf. discussion in section 5.3). 4.5.3 Elevational dependence of precipitation trends Total annual precipitation trends do not show a relationship with elevation (Figure 38). The significant precipitation trends are either low or high; the higher elevation (>2000 m) grid cells do not show significant trends. Seasonal precipitation trends also do not show specific patterns with elevation (Figure 39). The significant precipitation trends grid cells are mostly in the lower elevations of the region. Total annual precipitation trend vs elevation in the CM R o o lO C M w * Sig.trend # ♦ Not sig. trend O O O C M c o 0>0 .0) Ul ♦ \ R2 = 0 02 # * * *•/*'.*. . / . / * * n = 844 • . . •4 ♦ ♦ ~f • » . » : ' V * • ♦. **v* •. O O in o o o 4 Trend (mmyear ) Figure 38: Annual precipitation trend versus elevation, 1950-2010. The solid blue line shows the LOESS fit of trends with elevation. Seasonal precipitation trend vs elevation in the CMR * Sig.trend • Not sig. trend Trend [mmdecade ] Figure 39: Elevational dependence o f seasonal precipitation trends, 1950-2010. The solid blue line shows the LOESS fit o f trends with elevation. 5. Discussion The linear and temporal hydroclimatic trends in the CMR, elevational dependence of these trends, and impacts of such trends in the region are discussed in this chapter. The spatial trend magnitude distribution is also described. Furthermore, findings o f this study are used to discuss some of the physical mechanisms in the CMR that are possibly associated with the elevational dependency of the air temperature trends. Finally, possible impacts of climate change to the region’s hydrology and ecosystems are explored. 5.1 5.1.1 Hydroclimatic trends Magnitude and significance of linear trends Annual and seasonal minimum and maximum air temperatures show significant increasing trends over the period o f 1950 to 2010 in the CMR (Figures 22 and 25) providing clear evidence that the region is warming in recent decades. The minimum and maximum air temperatures o f the region rose by 1.93°C and 1.16°C respectively in the last six decades. These warming rates are consistent with reported increased air temperatures for BC (Dawson, Werner, & Murdock, 2008; MoE, 2014; PCIC, 2014; Picketts et al., 2012). A report on BC’s climate states that its annual air temperature has warmed by 0.5-1.7°C during the 20th century (MoE, 2014). The 1PCC AR5 (2013) report suggests that warming trends over the past six decades are very likely due to positive radiative forcing. Increases in greenhouse gases (GHGs) emissions, particularly from anthropogenic sources, have been causing the positive radiative forcing and therefore, surface warming globally (IPCC, 2013). Therefore, the trends observed in the CMR during the last six decades are consistent with warming temperatures globally, and are associated with anthropogenic influences on climate. The annual total precipitation does not show a specific pattern of increasing or decreasing trend. Total annual precipitation anomalies show multiple continuous years of positive anomalies during the 1990s and recent negative anomalies (Figure 24). Strong negative anomalies in 2009 (about 150 mm) and 2010 (about 300 mm) may be associated with the moderate El Nino events of these years. S. 1.2 Magnitude and distribution of spatial trends Trend Magnitude The average annual spatial warming rates over the CMR are 0.33°C decade'1 and 0.18°C decade'1 for minimum and maximum air temperatures, respectively (Figures 28 and 29). The warming rates and spatial patterns are different across the seasons in the CMR. For example, Table 8 shows the average air temperature increases over the CMR. The minimum air temperature trends are higher than the maximum air temperature trends, but with large spatial variability. Table 8: Average minimum and maximum air temperatures trends over the CMR, 1950-2010. In parenthesis are percentage of the significant values across the domain of the CMR. Trnin. Tmax. W inter 0.63 (89 %) 0.40 (93%) Spring 0.39 (94%) 0.20 (33%) Summ er 0.15(78%) 0.10(5%) Fall 0.13 (1%) -0.01 (<1%) There is a clear difference of air temperature trends with elevation over the CMR but the trends are not distinctly different along the western and eastern aspects of the study area. Both the western and eastern slopes of the Cariboo Mountains show similar trends with elevation. The northwestern and the southeastern regions of the CMR show significant decreasing and increasing precipitation trends, respectively, but the central part o f the CMR does not show significant trends (Figure 30). Trend Distribution The distribution of hydroclimatic trends over the CMR is analyzed to understand which trend magnitudes are dominant, how they are dispersed from the mean trend, and how they are related to elevation. Distribution of minimum air temperature trend magnitude Mean=0.33 SD=0,12 OO 04 02 03 04 05 06 07 O? Trend m agnitude [ ° C d e c a d e -1 ] Distribution of maximum air tem perature trend magnitude Mean=0.18 SD=0.06 004 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 Trend m agnitude [ ° C d e c a d e -1 ] Figure 40: Histogram o f the distribution o f air temperature trend magnitudes with mean trend (°C d e c a d e 1) and its standard deviation (SD ) (°C decade'1) over the CMR, 1950-2010. The solid red line show s the mean trend magnitude and the solid black curve shows the Gaussian distribution o f the trend magnitudes. Figure 40 shows the distribution of annual minimum and maximum air temperature trends over the CMR. The mean and standard deviation of trends are fitted with the Gaussian distribution. The mean trend over the area is 0.33°C decade1 for minimum air temperature with a standard deviation of 0.12°C decade'1. Similarly, the mean trend is 0.18°C decade"1 for maximum air temperature with a standard deviation of 0.06°C decade'1. All trend magnitudes for both minimum and maximum air temperatures in the CMR are positive. The highest frequency of the trend magnitude for minimum air temperature trends is 0.35-0.40°C decade'1, which is greater than the mean minimum air temperature trend (0.33°C decade'1). For maximum air temperature, the highest frequency of trends is 0.14-0.16°C decade'1, which is less than the mean maximum air temperature trend (0.18°C decade'1). Distribution of precipitation trend magnitude j i-------- ,----------------------- ---------------------- 1 -40 -30 -20 ~ |-----------------------1-----------------------1---------------------- \ ------------------ -10 0 10 20 I-----------------------1--------- 30 40 Trend magnitude [mm decade ' 1] Figure 41: Histogram o f the distribution o f annual total precipitation trend magnitudes with mean trend (mm decade'1) and its standard deviation (SD ) (mm decade'1) over the CMR, 1950-2010. The solid red line show s the mean trend magnitude and the solid black curve show s the Gaussian distribution o f the trend magnitudes. The precipitation trend magnitude also follows the Gaussian distribution with mean trend of nearly zero (-0.49 mm year"1). The distribution of precipitation trends is highly dispersed (SD = 13.79 mm year'1) ( Figure 41). Seasonal temperatijre trend distribution Ttnm Tmax Mean= 0.64 , SD =0.02 , SD= 0 ^ ^ ^ ^ ^ ^ Mean= 0 .3 9 , SD= 0.01 ^ ^ ^ ^ ^ l e a n = 0.1 B , SD= 0.01 S' <0 S' TJ A •e 0 0 o o z CM 8" ) * I A A Mean= 0.14 , SD= 0.01 S' Mean= 0.07 , SD= 0.01 1 Mean= -0.03 , SD= 0.01 i Trend magnitude [ ° C decade ] Figure 42: Seasonal trend distributions for minimum and maximum air temperatures in the CMR, 1950-2010. The solid red lines show seasonal mean trends for each season. The mean (°C decade"1) and SD (°C decade'1) o f each season is given. Relative magnitudes o f seasonal trends are also presented. The seasonal trend distributions show large variation in terms o f trend magnitude and distribution among the seasons as well as between minimum and maximum air temperatures. Most notably, increasing winter and spring air temperature trends show nearly Gaussian distributions. The summer and fall air temperature trends are skewed; the minimum air temperature trends are positively skewed and maximum air temperature trends are negatively skewed (Figure 42). Comparison with similar studies The warming trends found in this study, especially for minimum air temperature over the CMR are not only consistent with, but also stronger than most of the regional, Northern Hemisphere, and global warming trends. The findings on hydroclimatic trends, especially air temperatures, are consistent with the findings of Bonsai & Prowse (2003), Ceppi et al. (2012), Dery & Wood (2005), Easterling (2000), Mann, Bradley, & Hughes (1999), Picketts et al. (2012), Pike et al. (2008), MWLA (2002), Rangwala et al. (2013), Trenberth (1990), Stocker et al. (2013), and X. Zhang, Vincent, Hogg, & Niitsoo (2000). O f note, the results for minimum air temperature trends in the Cariboo Mountains follow those of Liu et al. (2009), Diaz & Bradley (1997), and Fyfe & Flato (1999). In a similar study for the Tibetan Plateau, Liu et al. (2009) examined the elevational dependency of minimum surface air temperatures using station data. They found more prominent warming at higher elevations than at lower elevations, especially during winter and spring. Diaz & Bradley (1997) revealed higher increases in daily minimum air temperature than maximum air temperature with elevation in the regional high elevation stations analysis of the American and Canadian Rockies. Fyfe & Flato (1999) found enhanced warming signals in the higher elevation region over the Rocky Mountains. They used the Canadian Centre for Climate Modelling and Analysis (CCCma) Coupled Climate Model, and found marked elevational dependency of increased surface air temperature at higher elevation in winter and spring. These findings are not consistent with McGuire et al. (2012) who reported strong warming signals at mid-elevations o f the Rocky Mountain Front Range for both 56 and 20 year periods. Further, the results of this study contradict those of Vuille & Bradley (2000) who found reduced air temperature trends with increasing elevation in the tropical Andes. These contradictions are more likely due to differences in study areas or in the approaches of analysis. There are many different factors and processes responsible for complex mountain region warming. There are certain mechanisms associated with elevation that provide specific responses to daily minimum and maximum air temperatures. Processes and mechanisms possibly responsible for different air temperature responses with elevation and their role in context of the CMR elevational dependency of warming are now discussed. 5.2 Physical mechanisms associated with trends Seasonal breakdowns of air temperature trends over the CMR show spatial and seasonal variations. Therefore, there should be some physical mechanisms related to elevation of the area as well as the seasons that influence these patterns of air temperature trends. Some of the physical mechanisms at local scales driven by climatic control of mountains are presented in Tables 1 and 2. The role of SAF, clouds, soil moisture, specific humidity, and other factors are discussed as possible factors responsible for such hydroclimatic trends over the Cariboo Mountains. However, a detailed analysis and role of impacts of each of these components is beyond the scope of this study. The summary o f the possible factors responsible for elevational warming in the CMR and their seasonal relevance are presented in Table 9. Table 9: Possible factors responsible for elevational warming in the CMR. Primarily spring; SAF (Decreases in Snow/Ice Important in winter (Lower Increases Tmm Albedo) elevation) and summer (Higher Increases Tmax elevation) Increases in Cloud Cover All seasons Decreases Tmax (Daytime) Increases in Cloud All seasons (greater Cover (Nighttime) effects in winter) Increases Tmm Decreases diurnal Snowmelt effects are strongest Increases in Soil in spring and winter; rainfall Moisture temperature range (DTR); Decreases effects are strongest in summer Tmax Increases in Specific Primarily winter Increases Tmjn All seasons Increases surface Humidity Local factors (Forest die back because o f mountain pine temperature (Tmm beetle attack and impacts on and Tmax) local energy balance) 5.2.1 Snow-albedo feedback (SAF) Sharp increases in the minimum air temperature during spring in the CMR may be due to the SAF. The SAF mechanism is one of the strongest climate drivers in the natural climate system. It causes elevational gradients of warming in snow-covered regions, particularly near the 0°C isotherm in relation to the surface energy budget (Flanner, Shell, Barlage, Perovich, & Tschudi, 2011). Indeed, many studies have reported amplified warming during spring days due to the SAF. Dery and Brown (2007) find amplified trends o f negative snow cover extent (SCE) in the Northern Hemisphere, North America, and Eurasia during spring and argue that it is a possible driver contributing to recent changes in observed SCE. Analysis of more than 1000 high elevation homogenized surface air temperature records across the globe show strongest warming rates near the annual 0°C isotherm due to the SAF (Pepin & Lundquist, 2008). Scherrer et al. (2012) quantify the effect o f the SAF on Swiss spring air temperature trends using daily air temperature and snow depth measurements. They found that the daily mean 2-m temperature of a spring day without snow cover is on average 0.4°C warmer than the one with snow cover at the same location. In the context of these findings, the SAF may have played a significant role for the steep minimum spring air temperature increase with elevation in the CMR. Furthermore, Fyfe & Flato (1999) have shown that surface albedo changes are responsible for net solar radiation change, that in turn has dominated the surface energy balance change. The effects of such changes are responsible for surface warming during winters in the Rockies. The location of the Rockies is near to this study area, lying in western North America, with similar global circulation influences. Therefore, winter air temperature increases in the CMR may be related to the surface albedo changes. 5.2.2 Clouds The amplified warming in the CMR may relate to cloud-radiation feedback mechanisms. Clouds have impacts both on short and long wave radiation, and thus affect surface air temperatures (Warren, Eastman, & Hahn, 2007). Most notably they affect the DTR and precipitation directly (Free & Sun, 2014; Ramanathan et al., 1989). Milewska (2004) finds positive trends of cloud cover especially in night time in BC. Dai, Karl, Sun, & Trenberth (2006) and Free and Sun (2014) show increasing trends in cloud cover over the northwest United States and the Northern Hemisphere in recent decades. Although their findings are not specific to the CMR region, their analyses cover northwestern North America where the CMR lies. The increase of cloud concentration increases the downwelling longwave radiation that results in increased minimum air temperature over the CMR. Similarly, increased cloud concentration decreases surface insolation during daytime, especially during summer, that decreases daily maximum temperature of the CMR. Therefore, the increased cloud over the CMR in recent decades could be related with the air temperature trend patterns observed over the CMR. Furthermore, Beniston & Rebetez (1996) find that there is an increase of night time air temperatures in winters in the lower valleys of the Alps than the regions exposed to stratus clouds at higher elevations. The clouds trap outgoing infrared radiation and warm the region. Most notably in winter, it is observed that the atmosphere near the CMR is calm and there is often a temperature inversion layer that may be associated with trapped clouds (section 4.1). Liu et al. (2009) also find increasing monthly minimum air temperatures with elevation and suggest that this dependency may be caused partially by cloud/radiation feedback effects. Similar conditions may have occurred in the CMR exhibiting higher temperature trends in higher elevations. 5.2.3 Soil moisture and specific humidity Soil moisture is considered as another important factor for air temperature increases (Seneviratne, Liithi, Litschi, & Schar, 2006). Soil moisture conditions affect the energy balance resulting in a change in surface air temperatures (Meng & Shen, 2014). Soil moisture influences all regions except water bodies. Daily maximum air temperature (Tmax) increases with decreasing soil moisture whereas an increase in soil moisture decreases DTR (Durre, Wallace, & Lettenmaier, 2000). The soil moisture content of the CMR may be decreasing with decreasing precipitation during summer. This may have contributed to increased maximum air temperature during summer in the CMR. There are studies that show an increase in minimum air temperature with an increase in specific humidity, primarily in winter with smaller effects during spring and fall (Rangwala et al., 2009; Rangwala & Miller, 2012). With increasing cloud concentration and precipitation in recent decades, there should be an increase in specific humidity over the CMR. This may have contributed to increasing air temperatures, especially minimum air temperature. The role of soil moisture and specific humidity for the air temperature trends over the CMR should not be as significant as SAF and clouds because of precipitation as snow for winter months and insignificant precipitation trends over the period of study. 5.2.4 Other Factors Local and regional disturbances and changes may have also influenced the air temperature of the Cariboo Region. Maness, Kushner, & Fung (2012) find that about a 1°C increase in summertime surface air temperature is associated with shifting evapotranspiration and albedo due to forest dieback following the mountain pine beetle infestation in BC. The wildfires and modifications in climatic circulation due to teleconnections such as the Pacific Decadal Oscillation (PDO) and El Nino-Southern Oscillation (ENSO) may have also influenced the air temperature pattern in the CMR (Ropelewski & Halpert, 1986; Stewart, Cayan, & Dettinger, 2005). Furthermore, other factors such as atmospheric brown clouds, industrial pollutants, and aerosol concentrations may be considered as additional factors responsible for elevational warming because of their role in absorbing the longwave radiation and radiating back to the earth’s surface, making it warmer. For the CMR, the effect of pollutants, aerosols, and black carbon should be minimal because of the location of the study area. It is relatively less influenced by anthropogenic factors owing to the lack of large industries nearby. Precipitation variation is associated with the mountainous topography of the CMR that directly influences the precipitation pattern between the leeward and windward sides. Local convective activity and wind activity are also responsible for precipitation variation in the region. | 97 | 5.3 Elevational dependency of hydroclimatic trends General Most of the grid cells of the study domain lie at mid-elevations (900-1500 m a.s.l.), with the range being 631 m to 2365 m (Figure 6). There are fewer points with climatic records at the elevation extremes, i.e. below 800 m and above 2000 m (Figure 6). Thus, findings of this study generally report the hydroclimatic trends that occurred in the mean elevations of the CMR. If it would be only global warming and not the influence of elevation, then either uniform warming or higher warming in lower elevation throughout the region is expected. However, the results indicate that there is a spatial difference across elevations in the seasonal trend magnitudes over the study area. Therefore, there is an influence of altitude to the variation of long-term hydroclimatic trends. There may be different physical phenomena associated with this variation either combined or individually. The quantification of the contribution of each of the physical processes to test the elevational dependency of air temperature trends is important but beyond the scope o f this work. Elevational Dependency Elevational dependency o f warming in the mountains is important because it has potential to provide early and detectable climate change signals and assess environmental impacts of global warming in these regions (Fyfe & Flato, 1999; Liu et al., 2009). There is clear evidence of elevational dependence of air temperature with elevation in the CMR. For the minimum air temperature, generally, the grid point’s trends, i.e. LOESS fits, form almost a J-shaped curve showing almost flat line until a 0.2°C decade'1trend at about 1200 m a.s.l. and moving straight up (Figure 34). Therefore, the higher the elevations of grid cells, the stronger are the trends. However, the reliability o f the fits is low above 2500 m a.s.l. because of the fewer grid cells. For the maximum air temperature, generally, the annual trends are decreasing with elevation except during winter. Winter maximum air temperature trends do not show any specific pattern with elevation. Not only the annual but also seasonal minimum and maximum air temperature trends are elevation dependent. For every season, the minimum air temperature trend increases with elevation, following a J-shaped curve except for fall (Figure 36). The maximum air temperature trend decreases with elevation almost linearly for all seasons except winter. Winter maximum air temperature does not indicate a specific pattern with elevation (Figure 37). Likewise, the precipitation trend does not follow a specific pattern with elevation (Figure 39). Positive minimum air temperature trends with greater magnitude in winter and spring may be linked to a specific phenomenon occurring during these periods in the region. There may be temperature inversions and fog formation in the lower valleys but not in high elevations especially during winters. This causes a higher minimum temperature response in the higher elevations. For spring, one important phenomenon that occurs in the CMR is the SAF that may have contributed to sharp increases in minimum air temperature trends with elevation. Furthermore, generally higher elevations experience higher cloud concentration/formation, recent decades show that the rate of cloudiness is increasing in the United States and Canada; specifically during night time in BC (Dai et al., 2006; Free & Sun, 2014; Milewska, 2004). This may be happening over the CMR and have contributed to increased minimum and decreased maximum air temperature trend magnitudes with elevation. The results do not show a clear relationship between the precipitation trends and elevation in the CMR. 5.4 5.4.1 Implications of trends Seasonal shifting With increasing warming trends in cold regions such as the CMR, one of the most noted impacts would be on snowmelt initiation in spring and start of the freezing period in fall. For many practical purposes, freezing and melting temperatures are o f special interest in cold regions. For example, winters are dominated by the cryospheric components, bringing changes to ecosystems, water resources as well as society and economy (Bonsai & Prowse, 2003). This study shows a substantial increase in air temperatures similar to global warming. It is expected that there is a substantial ch an ge o v er the m eltin g and freezin g tim e in the CMR. I ioo| Trend of Spring d a y s [Julian] with Tem p, threshold of 0°C Figure 43: Linear trends o f start o f days with greater than 0°C temperature threshold for minimum, maximum, and mean air temperature (areal average) in the CMR, 1950-2010. Trend of fall d a ys [Julian] with T em p , threshold of 0°C 330 T 320 310 300 290- y =261 +0.021 x, /* = 0.00987 300- 29 0 - 280- y S174+U.055-X, r =0.0075 270 - y =*163+0.045/y, ^ = 0.013 260 - 250- Time [Year] Figure 44: Linear trend o f start o f freezing days with temperature threshold o f less than 0°C for minimum, maximum, and mean air temperature (areal average) over the CMR, 1950-2010. A trend analysis is conducted to observe changes in the timing of the transition between subfreezing and above freezing conditions and vice versa. The daily minimum, maximum, and mean air temperatures are filtered using a 31-day running mean for spring and fall. A 0°C air temperature is considered as the threshold value for melting (freezing) | 102 | start days of spring (fall). The changes in the number of threshold days with year is calculated to measure the seasonal shifting. Figures 43 and 44 show the trend of days with a 0°C air temperature threshold for spring and fall respectively. For spring, the number of days is decreasing in the CMR (-4.1 days decade'1) for minimum air temperature while the number of days for fall 0°C is increasing (0.45 days decade'1) for minimum air temperature. Generally, minimum air temperature trends are considered to refer to changes of melting (freezing) periods. The numbers of days with 0°C air temperature threshold are decreasing for spring while increasing for fall. The trends towards earlier spring and later fall days over the CMR are consistent with the hypothesis that with increasing warming there may be shorter snow seasons (Choi et al. 2010). Timing of snowmelt is affected by both the air temperature and precipitation. Early snowmelt leads to the early runoff, in turn influencing the overall hydrology of the area (Lundquist, Dettinger, Stewart, & Cayan, 2009; Pederson et al., 2011). There are other potential associated impacts such as decreases in reservoir efficiencies (Stewart, Cayan, & Dettinger, 2004), increases in the number of forest fire events (Westerling, Hidalgo, Cayan, & Swetnam, 2006), increases in river flooding, and disturbances in the aquatic habitat, etc. (Tohver, Hamlet, & Lee, 2014; Zwiers, Schnorbus, & Maruszeczka, 2011). It will also influence water rights and management issues among provinces and states in the future (Kenney, Klein, Goemans, Alvord, & Shapiro, 2008). 5.4.2 Implications to the ecohydrology This study reports almost a 2°C of minimum air temperature increase in the CMR during the last six decades with the rate of change increasing. Therefore, the region is warming and is likely to warm in many years to come. Such warming has and will have broader implications towards the local ecohydrology of the CMR and overall northern BC. There is a strong relation between climate and the overall ecosystem of a particular region (Mooney et al., 2009). There are many examples showing ecological impacts of climate change at local, regional, and global level ecosystems (Grimm et al., 2013; Mooney et al., 2009; Walther et al., 2002; Weed, Matthew, & Hicke, 2013). Increased frequency and intensity of weather events are the significant driving factors for population dynamics of different species of any region. In the case of the CMR, the mountain caribou, salmon, and other species are directly impacted by climate changes over the region (Dery et al., 2012; Vors & Boyce, 2009). Nonetheless, the consequences of short-term and long-term climatic variations on complex ecological processes are not clear. In the following sections, some of the impacts of such changes on selected animal species of the study area and water resources of the region are discussed. 5.4.2.1 Impacts on water resources Increased air temperatures and changes in the precipitation regime over western North America will affect the hydrology o f the region. This will have impacts on hydropower generation, potentially cause disasters such as floods and droughts, affect water demand for irrigation and household consumption, and disturb aquatic habitats (Zwiers et al., 2011). I 104| Glaciers cover about 3% o f the BC landmass and play a significant role in the river runoff as well as hydropower generation in BC. There are recent studies that show the accelerated retreat of glaciers in the region (Beedle, 2014; Bolch et al., 2010; Moore & Demuth, 2001). Although there are very few findings to relate these recessions with increasing air temperature and changes in precipitation patterns, it may be accelerating due to the warming in the region. The faster retreat of glaciers in the CMR will also affect the flow in the Fraser River and its tributaries that drain the region. Most of the Cariboo Mountains drain into the Fraser River, a major habitat for keystone salmon species. Many studies and model simulations show significant changes over the hydrological regime of the Fraser River Basin with increasing temperature (Dery et al., 2012; Kang, Shi, Gao, & Dery, 2014; Shrestha, Schnorbus, Werner, & Berland, 2012). Rising air temperature is one of the causes of a greater range o f annual runoff fluctuations in the Fraser River (Dery et al., 2012). Such fluctuations have detrimental effects on migrating and spawning salmon. In addition to the study area’s specific impacts, there will be implications of hydroclimatic changes over the CMR on many sectors of the CMR as well as northern and central BC as a whole. Some of the implications of increasing air temperatures over the region include the mountain pine beetle epidemic, increase in frequency and occurrence of forest fire events, extreme weather events, etc. (Kurz et al., 2008). These may cause much environmental and economic losses. Therefore, significantly increasing air temperatures, as shown in the CMR, have broader implications towards the hydrology, ecosystems and ecosystem services, and overall environmental health of the CMR and northern BC. I 105| 5.4.2.2 Impacts on animal species The endangered mountain caribou (mountain ecotype of woodland caribou) population that is found in relatively wet and mountainous terrain of BC is declining. One o f the causes for population decline may be the milder winters due to climate change along with human influences (Opdam & Wascher, 2004; Parks Canada, 2011; Vors & Boyce, 2009). Mountain caribou population distribution has declined over the past 50 to 100 years and now they are limited to only 12 recognized sub-populations (Wittmer et al., 2005). Broadly, the mountain caribou population has been stratified into four different geographic regions in BC with one of them being the Cariboo Mountains where there are now only 850 caribous or so (Mountain Caribou Science Team, 2005). In the Cariboo Mountains, seasonal migration of the caribou to lower elevations is further limited because of shallower snowpacks. This migration will be impacted even further should winter air temperatures continue to warm in the region. The primary winter food for caribou is arboreal lichen that is abundantly found in the coniferous forests o f the CMR. They walk on top of the deep snowpack (generally >2 m) in the mountains (Apps, McLellan, Kinley, & Flaa, 2001; Wittmer et al., 2005). With a late start of freezing conditions during fall and milder winters, there will be reduction in the snowpack thickness making it difficult for caribou to access the arboreal lichen. In addition, the warmer and drier winter conditions are favorable for deer, elk, and moose to move to the Cariboo Mountains. This will increase the competition for food amongst animals (Mountain Caribou Science Team, 2005). Furthermore, there are studies that show the change in nutrient cycling with warming that would affect caribou populations adversely by creating competition with browsing ungulates and eliminating food sources, especially in winters | 106| (Lenart, Bowyer, Hoef, & Ruess, 2002). Complex interactions among seasonal hydroclimatic variations, snowfall patterns, wildfire events, forest insect disease outbreaks, etc., have the potential to increase and compound the impacts on the mountain caribou. This is expected to increase with rising air temperature in the future. 5.5 Limitations of this study There are limitations to this study that require some attention. First are the data gaps in the CAMnet data used for the general climatology of the region and validation of interpolated data. Further, the accuracy of the interpolated data, ANUSPLIN, is another limitation of this study. Unfortunately, there are limited weather stations generating long­ term data to conduct a detailed and systematic analysis over the region. The nature of data sets such as their quality, time of availability, etc. influence the results based on these data. Observation-based interpolated data are used in this study, but inaccuracies arise from the fact that there are fewer stations in the higher elevations for interpolation. The data sets are crosschecked to corroborate the quality o f data in representing the study area, yet the interpolated data may not represent the real scenario exactly. This is also because the interpolated data used are not specific to the study area, but are extracted from the entire Canadian domain. Data resolution remains coarse to represent the details of complex mountainous topography of the region. In addition, the gridding process may have smoothed the data spatially. These gridded data may have some errors too. The precipitation analysis is based on interpolated total data only. It is not clear how much of the precipitation is contributed by snow and how much is from rainfall because of unavailability of snowfall data. | 107 | There is limited literature that examines the effects of glacier retreat in the CMR in relation to global warming. Further, there are limited publications emphasizing impacts of climate change on the water resources of the CMR. Therefore, the discussions in this study are limited to available studies. | 108 | 6. Conclusions, Recommendations and Future W ork The conclusions, recommendations, and future work based on the analyses are presented in this chapter. The first part summarizes the main findings of this study. The hydroclimatic state in the CMR is summarized based on linear and spatial trends, with their magnitude and significance at annual and seasonal scales. Furthermore, the elevational dependency of the hydroclimatic trends in the CMR and impacts due to these changes are presented. The second part of this chapter consists of recommendations for further research and proposes some future work based on the scope and limitations of this study. 6.1 Conclusion 6.1.1 General Hydroclimatoiogy Average annual minimum daily air temperature in the CMR is -3.7°C (minimum -16.3°C in winter and maximum 8.0°C in summer) during 1950-2010. Similarly, the average annual maximum daily air temperature is 7.3°C (minimum -7.4°C in winter and maximum 23.4°C in summer) for the same period. The total mean annual precipitation over the region is 738 mm. An analysis of observed climate data along the western slopes of the Cariboo Mountains shows that higher elevation regions have lower air temperatures than lower elevations. The monthly average surface temperature lapse rate of the region is 4.7°C km'1 during the period of September 2011 June 2012. Air temperature is found to be inversely related to elevation in all months, except during extreme weather events in January and February. For these days, it may be due to air temperature inversions occurring in the region. 6.1.2 Linear and spatial trends A hydroclimatic trend analysis of the region was performed using the daily frequency Canadian Climate Coverage (CCC)’s interpolated data from 1950 to 2010. The minimum and maximum air temperatures and precipitation were summarized for monthly, seasonal, and annual frequency. The nonparametric Mann-Kendall trend test was performed to detect the trends and their significance. The Theil-Sen estimate was used to evaluate the magnitude of the trends over the region at different temporal and spatial resolutions. The minimum and maximum air temperatures of the region have risen by 1.9°C and 1.2°C, respectively, in the last six decades. Area averaged annual minimum and maximum air temperatures show significant positive trends over the period of study. The minimum air temperature trend over the region is 0.32°C decade'1 while maximum air temperature is 0.20°C decade"1. Although the total annual precipitation does not show any significant trend, there is year-to-year variation of total precipitation by ±30% from its long-term mean. Anomalies in air temperatures showed that recent years, especially after 1996, are warmer across the CMR. There is a greater seasonal variation in the hydroclimatic trends in the CMR. The winter and spring minimum air temperature trends are 0.62°C decade'1 and 0.35°C decade'1respectively followed by the summer trend of 0.12°C decade'1. All of these trends are highly significant ip <0.05) except for the fall minimum air temperature trend. For maximum air temperature, only winter trends are significant. The seasonal precipitation does not show significant trends during last six decades. The spatial trends of the region show mostly significant positive annual trends for air temperatures, but both positive and negative trends for precipitation. The average minimum air temperature warming rate is 0.33°C decade'1. The spatial trend shows greater variability I 1101 with magnitudes ranging from 0.07-0.67°C decade'1. The mean maximum air temperature trend over the region is 0.18°C decade'1(range of 0.03-0.34°C decade'1) with less spatial variation compared to the minimum air temperature. Contrary to yearly trends, there is a clear and strong spatial variation of seasonal trends. On the seasonal scale, the analysis reveals that there are all positive trends of minimum air temperature in the region except some grid cells showing insignificant negative trends during the fall. Winter and spring minimum air temperature trends are strong and positive. They show greater spatial variability. For example, winter minimum air temperature trends range from 0.24°C decade'1to 1.18°C decade'1(mean = 0.63°C decade'1) and spring minimum air temperature trends range from 0.09°C decade'1to 0.72°C decade'1(mean = 0. 39°C decade'1). The summer and fall show little spatial variability. Maximum air temperature trends for winter and spring are all positive and mostly significant. The summer and fall trends range from negative to positive values and generally are not significant, showing little spatial variability. For example, winter trends range from 0.21°C decade'1to 0.59°C decade'1(mean = 0.39°C decade’1) while fall trends range from -0.12°C decade'1to 0.13°C decade'1(mean = -0.03°C decade'1). The precipitation trends show greater spatial variability over the study area. Most o f the grid cells do not show significant trends; in addition, there are positive trends in some areas of the CMR and negative ones in other areas. Winter precipitation trends range from -18.5 mm decade'1to 6.2 mm decade'1(mean = -6.8 mm decade'1), spring -2.2 mm decade'1to 9.3 mm decade'1(mean =1.9 mm decade'1), summer -4.2 mm decade'1to 8.5 mm decade'1(mean = -0.2 mm decade'1), and fall -6.6 mm decade'1 to 15.4 mm decade'1(mean = 5.3 mm decade'1). I 111 i 6.1.3 Elevational dependency of hydroclimate There is a clear difference in the trends with respect to elevation in the CMR. Both the western and eastern slopes of the Cariboo Mountains show similar patterns in the air temperature trend with respect to elevation. The annual minimum air temperature trends are stronger and significant at higher elevations with magnitude of more than 0.5°C decade'1above 2000 m a.s.l. The annual maximum air temperature trends show the opposite pattern of increase with elevation, i.e. stronger and significant maximum air temperature trends are observed at lower elevations, mostly below 1500 m a.s.l. There is no specific annual precipitation trend with elevation. The patterns of air temperature trends variability are related with the elevation of the region. The seasonal breakdown of minimum air temperature trends with elevation show stronger, significant trends with increasing elevations. Winter and spring trends are >0.50°C decade'1above 2000 m a.s.l. (extreme values of winter trends above 2000 m a.s.l. are 0.80°C decade'1to 1.30°C decade'1and for spring 0.60°C decade'1to about 0.80°C decade'1). For maximum air temperature, most significant trends are observed below 1500 m a.s.l. for all seasons showing decreasing trends with increasing elevation except for winter. During winter, there is no specific pattern of trend with elevation. The seasonal precipitation trends also do not show any pattern with elevation. The SAF and clouds along with other factors in the region may be responsible for sharp gradients in air temperature trends with elevation in the CMR. The role of the SAF is significant during spring; summer elevational dependency of the air temperature may be influenced by soil moisture, aerosols, and change in local surface energy balance because of mountain pine beetle disturbances. SAF and increasing cloud concentration may be responsible for winter time air temperature trend variations in the CMR. 6.1.4 Impacts of hydroclimatic variations There have been direct and indirect impacts o f hydroclimatic variations over the local and regional ecohydrology and economy o f BC. Such impacts are expected to continue into the 21st century. Spring melt is starting earlier (0.41 day decade'1) and fall freezing is delayed (0.45 day decade'1). Such timing of snowmelt change will affect runoff generation and the overall hydrology/water resources of the region. Milder winters will affect the endangered mountain caribou populations in the CMR through increases in competition with other species such as elk, moose, and deer. Further, there will be shallow snow depths that make the caribou’s major winter food, arboreal lichen, inaccessible. With a warming environment, there is increased runoff variability in the Fraser and North Thompson Rivers, major habitats for salmon. Increased mountain pine beetle epidemic frequency, forest fires, sedimentation in reservoirs, effects on hydroelectricity generation, and glacier retreat are other impacts expected to occur due to hydroclimatic changes in the CMR. The CMR is warming; the minimum air temperature is increasing at a faster rate than the lowland counter parts. There are stronger warming signals at higher elevations; minimum air temperature is increasing at a faster rate, especially in winter and spring. Such changes have impacts over the ecohydrology and overall environmental health of northern BC and region. 1113| 6.2 R ecom m endations and Future W ork To reduce the uncertainties associated with climatic variations, it is recommended to have a dense network of continuously operating weather stations in the region, especially at high elevations. It is recommended to establish and operate new weather stations at higher elevations, especially above 2000 m for the long term monitoring of mountain climate. The interpolated data were extracted from the large Canadian domain for the trend analyses. The use of a large domain may have smoothed the data, especially for the complex mountain environment of the CMR. Therefore, it is recommended and proposed as a future effort to develop specific gridded hydroclimatic datasets for the CMR. The specific gridded data should incorporate any observed data available in the region as well as model output. Results from such data would be more specific, provide more precise estimates, and reduce uncertainties. The hydroclimatology of the CMR is influenced by large scale atmospheric circulation patterns and mountain specific phenomena. Therefore, there is a need to develop multiple regression models using atmospheric circulation patterns such as 500 hPa geopotential height fields, radiation/energy distribution data of the region, etc. Further, this study showed opposite warming patterns of minimum and maximum air temperatures, with the minimum air temperature trend increasing with elevation and maximum air temperature trend decreasing with elevation. Therefore, there are still uncertainties on how mountain regions will respond to global warming. This study is a special case of a relatively small mountainous region. It is difficult to estimate the contribution of specific physical mechanisms such as the cloud concentration or SAF on air temperature changes. Therefore, further investigation of relevant physical I H4 | mechanisms and accurate quantification of processes and their role for elevational dependency o f warming in large mountain ranges is suggested. This could be achieved by developing proper regression models and performing Principal Component Analysis (PCA) analysis of 500 hPa geopotential height fields. Furthermore, development and use of high resolution climate models is recommended. There are many complexities associated with such analysis in mountain regions. The proper methods of investigation to identify the effect of each parameter to the hydroclimatology of the CMR are important. An extended analysis of daily maximum and minimum air temperature throughout many mountain chains of the globe would improve the understanding of elevational sensitivity of the mountains in the context of global warming. In addition, large-scale hydroclimatic variation analysis is recommended for a better understanding of the impacts of climate change. Similarly, case studies from large mountains or mountain ranges are recommended for improved understanding of elevational dependency of hydroclimatic parameters. For example, an analysis of elevational dependency o f trends in the Canadian Rockies or the Columbia Mountains would enhance the understanding of recent changes observed in East-Central BC. This study has explored the dependence of hydroclimatic variables in the CMR and contributed to enhance our understanding of mountain climate and their responses to the climate change. Furthermore, the roles of different physical mechanisms responsible for elevational dependency of hydroclimate in the CMR are explained. The implications of hydroclimatic changes in the ecohydrology of the CMR are discussed. This work is expected to fill the gaps of our understanding about the pristine mountain environment and provide [115| scientific information to policy makers and communities to formulate better policies and make informed decisions. S H 6 j 7. References Apps, C. D ., McLellan, B. N ., Kinley, T. A ., & Flaa, J. P. (2001). Scale-dependent habitat selection by Mountain Caribou, Columbia Mountains, British Columbia. Journal o f Wildlife M anagement, 65(1), 6 5 -7 7 . Retrieved from http://cat.inist.fr/?aM odele=afficheN&cpsidt=938318 Barry, R. G. (1992). Mountain clim atology and past and potential future changes in Mountain Regions: A Review. M ountain Research an d Developm ent, 72(1), 7 1 -8 6 . Barry, R. G. (2008). M ountain W eather an d Clim ate (3rd ed.). Cambridge, United Kingdom and N ew York, N Y , USA: Cambridge University Press. Barry, R. G. (2012). Recent advances in mountain climate research. Theoretical and A p p lied Clim atology, 7 /0 (4 ), 5 4 9 -5 5 3 . doi:10.1007/s00704-012-0695-x BC Parks. (2014). Cariboo Mountains Provincial Park. C ariboo M ountains Provincial Park. Retrieved March 29, 2014, from http://www.env.gov.bc.ca/bcparks/explore/parkpgs/cariboo_mt/ BCRFC. (2014). Automated Snow Pillow Data Archive. Autom ated Snow P illow D ata Archive. Retrieved January 10, 2014, from http://bcrfc.env.gov.bc.ca/data/asp/archive.htm Beacham, T. D., Lapointe, M ., Candy, J. R., McIntosh, B., MacConnachie, C., Tabata, A ., ... Withler, R. E. (2004). Stock identification o f Fraser River Sockeye Salmon using M icrosatellites and Major Histocompatibility com plex variation. Transactions o f the Am erican F isheries Society, 133(5), 1117-1137. doi: 10.1577/T 04-001.1 Beacham, T. D ., & Murray, C. B. (1989). Variation in developmental biology o f sockeye salmon (Oncorhynchus nerka) and chinook salmon (O. tshawytscha) in British Columbia. Canadian Journal o f Zoology, 67, 2081—2089. Beedle, M. J. (2014). G lacier change in the C ariboo M ountains o f British C olum bia,C anada (19462011). University o f Northern British Columbia. Beedle, M. J., Menounos, B., & Wheate, R. (2014). Glacier change in the Cariboo Mountains, British Columbia, Canada (1 9 5 2 -2 0 0 5 ). The Cryosphere D iscussions, 8(3), 3367-3411. doi: 10.5194/tcd-8-3 367-2014 Beniston, M. (2003). Climatic change in Mountain Regions: A Review o f possible impacts. Clim atic Change, 59, 5 -3 1 . Beniston, M. (2006). Mountain Weather and Climate: A general overview and a focus on Climatic Change in the Alps. H ydrobiologia, 562(1), 3 -1 6 . doi: 10.1007/s 10750-005-1802-0 Beniston, M., Diaz, H. F., & Bradley, R. S. (1997). Climatic change at high elevation sites: an overview. Clim atic Change, 36, 233-251. I U7 | Beniston, M ., & Fox, D. G. (1995). Impacts o f Climate Change on Mountain Regions. In R. T. Watson, M. C. Zinyowera, & R. H. M oss (Eds.), Working G roup IIR ep o rt "Impacts, A daptations an d M itigation o f Clim ate Change: Scientific-Technical Analyses ” Contribution o f Working G roup II to the Second Assessm ent Report o f the Intergovernm ental P anel on Clim ate Change. IPCC. Beniston, M., & Rebetez, M. (1996). Regional Behaviour o f Minimum Temperatures in Switzerland for the Period 1979-1993. Theoretical and A p plied Clim atology, 243, 2 31 -2 4 3 . Bolch, T., M enounos, B., & Wheate, R. (2010). Landsat-based inventory o f glaciers in western Canada, 1985-2005. Rem ote Sensing o f Environment, 114( 1), 127-137. doi: 10.1016/j.rse.2009.08.015 Bonsai, B. R., & Prowse, T. D. (2003). Trends and variability in Spring and Autumn 0°C-isotherm dates over Canada. Clim atic Change, 57, 341 -3 5 8 . Bradley, R. S. S., Keim ig, F. T., & Diaz, H. F. (2004). Projected temperature changes along the American Cordillera and the planned GCOS network. G eophysical Research L etters, 37(16), L 16210. d oi:10.1029/2004G L 020229 Burford, J. E., Dery, S. J., & Holm es, R. D. (2009). Some aspects o f the hydroclimatology o f the Quesnel River Basin, British Columbia, Canada. H ydrological P rocesses, 23, 1529-1536. doi: 10.1002/hyp Bum , D. H., & Elnur, M. A. H. (2002). Detection o f hydrologic trends and variability. Journal o f H ydrology, 255(1-4), 107-122. doi: 10.1016/S0022-1694(01)00514-5 Ceppi, P., Scherrer, S. C., Fischer, A. M., & Appenzeller, C. (2012). Revisiting Sw iss temperature trends 1959-2008. International Journal o f C lim atology, 32(2), 2 0 3 -2 1 3 . doi:10.1002/joc.2260 Choi, G., Robinson, D. a., & Kang, S. (2010). Changing Northern Hemisphere snow seasons. Journal o f Clim ate, 23(19), 5 3 0 5 -5 3 1 0 . doi:10.1175/2010JCLI3644.1 Cleveland, W. S., & Devlin, S. J. (1988). Locally weighted regression: An approach to regression analysis by local fitting. Journal o f Am erican Statistical Association, 53(403), 596 -6 1 0 . Dai, A., Karl, T. R., Sun, B., & Trenberth, K. E. (2006). Recent trends in cloudiness over the United States: A tale o f monitoring inadequacies. Bulletin o f the Am erican M eteorological Society, 57(5), 5 9 7 -6 0 6 . doi: 10.1175/BA M S-87-5-597 Dalgaard, P. (2008). Introductory Statistics with R (2nd ed.). Springer. Daly, C. (2006). Guidelines for assessing the suitability o f spatial climate data sets. International Journal o f C lim atology, 26(6), 707-721. doi: 10.1002/joc. 1322 Daly, C., Halbleib, M., Smith, J. I., Gibson, W. P., Doggett, M. K., Taylor, G. H., & Pasteris, P. P. (2008). Physiographically sensitive mapping o f clim atological temperature and precipitation across the conterminous United States. International Journal o f Clim atology, 28, 20 3 1 -2 0 6 4 . doi: 10.1002/joc i U S I Davis, E. J., & Reed, M. G. (2013). M ulti-level governance o f British Columbia’s mountain pine beetle crisis: The roles o f memory and identity. Geoforum, 47, 32 -4 1 . d oi: 10.1016/j .geoforum.2013.02.005 Dawson, R. J., Werner, A. T., & Murdock, T. Q. (2008). Prelim inary analysis o f Clim ate Change in the Cariboo-Chilcotin area o f British Columbia (p. 49). Pacific Climate Impacts Consortium, University o f Victoria, Victoria BC: Pacific Climate Impacts Consortium, University o f Victoria, Victoria BC. Dery, S. J., & Brown, R. D. (2007). Recent Northern Hemisphere snow cover extent trends and implications for the snow-albedo feedback. G eophysical Research Letters, 34(22), L22504. doi :10.1029/2007G L 031474 Dery, S. J., Clifton, A ., MacLeod, S., & Beedle, M. J. (2010). B low ing Snow Fluxes in the Cariboo Mountains o f British Columbia, Canada. Arctic, Antarctic, an d Alpine Research, 42(2), 188— 197. doi: 10.1657/1938-4246-42.2.188 Dery, S. J., Hemandez-Henriquez, M. A ., Owens, P. N., Parkes, M. W., & Petticrew, E. L. (2012). A century o f hydrological variability and trends in the Fraser River Basin. Environm ental R esearch Letters, 7(2), 024019. doi: 10.1088/1748-9326/7/2/024019 Dery, S. J., Stieglitz, M., McKenna, E. C., & Wood, E. F. (2005). Characteristics and Trends o f River Discharge into Hudson, James, and Ungava Bays, 1964 - 2000. Journal o f Clim ate, 18, 2 5 4 0 2557. D eiy, S. J., & W ood, E. F. (2005). Observed twentieth century land surface air temperature and precipitation covariability. G eophysical Research Letters, 32(21), L21414. d o i: 10.1029/2005G L 024234 Diaz, H. F., & Bradley, R. S. (1997). Temperature variations during the last century at high elevation sites. Clim atic Change, 36, 2 5 3 -279. D ong, D ., Huang, G., Qu, X., Tao, W., & Fan, G. (2014). Temperature trend-altitude relationship in China during 1963-2012. Theoretical and A p plied Clim atology, doi: 10.1007 /s0 0 7 0 4 -0 14-12869 Durre, I., W allace, J. M., & Lettenmaier, D. P. (2000). Dependence o f extreme daily Maximum temperatures on antecedent soil moisture in the contiguous United States during summer. Journal o f Clim ate, 13, 26 4 1 -2 651. Easterling, D. R. (2000). Climate extremes: Observations, modeling, and impacts. Science, 259(5487), 2 0 6 8 -2 0 7 4 . doi: 10.1126/science.289.5487.2068 English, K. K., Koski, W. R., Sliwinski, C., Blakley, A., Cass, A., & W oodey, J. C. (2005). Migration timing and river survival o f late-run Fraser river Sockeye Salmon estimated using Radiotelemetry techniques. Transactions o f the Am erican Fisheries Society, 134(5), 1342-1365. doi: 10.1577/T 04-119.1 I 119 I Environment Canada. (2014). Historical Climate Data. H istorical Clim ate D ata. Retrieved April 04, 2014, from http://climate.weather.gc.ca/data_index_e.html ESRI. (2011). ArcGIS Desktop: Release 10. Redlands, CA: Environmental System s Research Institute. Eyles, N. (1995). Characteristics and origin o f coarse gold in Late Pleistocene sediments o f the Cariboo placer mining district, British Columbia, Canada. Sedim entary G eology, 95(1-2), 6 9 95. d oi: 10.1016/0037-0738(94)00101 -Y Flanner, M. G., Shell, K. M., Barlage, M., Perovich, D. K., & Tschudi, M. A. (2011). Radiative forcing and albedo feedback from the Northern Hemisphere cryosphere between 1979 and 2008. Nature G eoscience, 4(3), 151-155. doi:10.1038/ngeo 1062 ForestEthics. (2014). Inland Temperate Rainforest: Saving the endangered Mountain Caribou. Inland Temperate Rainforest: Saving the endangered mountain caribou. Retrieved September 17, 2014, from http://www.forestethics.org/inland-temperate-rainforest Free, M., & Sun, B. (2014). Trends in U.S. total cloud cover from a Homogeneity-Adjusted Dataset. Journal o f Clim ate, 27(13), 49 5 9-4969. doi: 10.1175/JCLI-D-l 3-00722.1 Fyfe, J. C., & Flato, G. M. (1999). Enhanced Climate Change and its detection over the Rocky Mountains. Journal o f Climate, 12, 230-243. Gan, T. Y ., & Kwong, Y. T. J. (1992). Identification o f warming trends in Northern Alberta and Southern Northwest territories by the non-parametric Kendall Test U sing hydrometric data to detect and moitor climate channge. In G. E. Kite & K. D. Harvey (Eds.), P roceedings o f NHRI Symposium 8 (pp. 4 3 -5 6 ). Saskatoon. GeoBC. (2014). Barkerville. The P rovince o f British Colum bia - GeoBC. Retrieved September 17, 2014, from http://apps.gov.bc.ca/pub/bcgnws/names/l 1124.html Ghatak, D., Sinsky, E., & Miller, J. (2014). Role o f snow-albedo feedback in higher elevation warming over the Himalayas, Tibetan Plateau and Central Asia. Environm ental Research Letters, 9(11), 114008. doi: 10.1088/1748-9326/9/11/114008 G ocic, M., & Trajkovic, S. (2013). A nalysis o f changes in meteorological variables using MannKendall and Sen’s slope estimator statistical tests in Serbia. G lobal and Planetary Change, 100, 172-182. doi:10.1016/j.gloplacha.2012.10.014 G oosse, H., Barriat, P. Y., Lefebvre, W., Loutre, M. F., & Zunz, V. (2010). Introduction to clim ate dynam ics an d clim ate m odelling. Online textbook available at http://www.climate.be/textbook. Retrieved from http://www.climate.be/textbook Grimm, N. B., Chapin, F. S., Bierwagen, B., Gonzalez, P., Groffman, P. M., Luo, Y., ... W illiamson, C. E. (2013). The impacts o f climate change on ecosystem structure and function. F rontiers in E cology an d the Environment, 11(9), 4 7 4 -4 8 2 . doi: 10.1890/120282 |1 2 0 | Hageli, P., & M cClung, D. M. (2003). Avalanche characteristics o f a transitional snow climate— Columbia Mountains, British Columbia, Canada. C o ld Regions Science and Technology, 57(3), 2 5 5 -2 7 6 . doi: 10.1016/SO 165-232X (03)00069-7 Hijmans, R. J. (2014). Raster: Geographic data analysis and modeling. R package version 2.2-12. http://CRAN.R-project.org/package=raster. Hopkinson, R. F., M cKenney, D. W., M ilewska, E. J., Hutchinson, M. F., Papadopol, P., & Vincent, L. A. (2011). Impact o f A ligning Climatological Day on Gridding Daily M aximum-M inimum Temperature and Precipitation over Canada. Journal o f A p plied M eteorology and C lim atology, 50(8), 1654-1665. doi: 10.1175/2011JA M C 2684.1 Hutchinson, M. F., M cKenney, D. W., Lawrence, K., Pedlar, J. H., Hopkinson, R. F., M ilewska, E., & Papadopol, P. (2009). Development and Testing o f Canada-Wide Interpolated Spatial M odels o f Daily M inim um -M axim um Temperature and Precipitation for 1961-2003. Journal o f A p p lied M eteorology a n d C lim atology, 48(4), 725 -7 4 1 . doi: 10.1175/2008JA M C 1979.1 IPCC. (2013). Clim ate Change 2013: The Physical Science Basics. Contibution o f Working Group I to the Fifth Assessm ent R eport o f the Intergovernm ental Panel on Cliam te Change. (T. F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. B o sch u n g ,... P. M. M idgley, Eds.) (p. 1535). Cambridge, United Kingdom and N ew York, N Y , USA: Cambridge University Press. Kang, D. H., Shi, X., Gao, H., & Dery, S. J. (2014). On the changing contribution o f snow to the hydrology o f the Fraser River Basin, Canada. Journal o f H ydrom eteorology, 15, 1344—1365. doi: 1 0 .1 175/JHM -D-13-0120.1 [In Press] Kendall, M. G. (1975). Rank Correlation Methods. (4th ed.). London: Charles Griffin. Kenney, D., Klein, R., Goemans, C., Alvord, C., & Shapiro, J. (2008). The Im pact o f E arlier Spring Snowm elt on Water Rights an d A dm in istration : A Prelim inary O verview o f Issues and Circum stances in the Western States (Final P roject Report). Krause, P., B oyle, D. P., & Base, F. (2005). Comparison o f different efficiency criteria for hydrological model assessment. A dvances in G eosciences, 5, 89 -9 7 . Kurz, W. A ., Dymond, C. C., Stinson, G., Rampley, G. J., N eilson, E. T., Carroll, A. L., ... Safranyik, L. (2008). Mountain pine beetle and forest carbon feedback to climate change. Nature, 452(7190), 9 8 7 -9 0 . doi:10.1038/nature06777 Lenart, E. A ., Bowyer, R. T., Hoef, J. Ver, & Ruess, R. W. (2002). Climate change and Caribou: Effects o f summer weather on forage. Canadian Journal o f Zoology, 80, 664 -6 7 8 . doi:10.1139/Z02-034 Liu, X., & Chen, B. (2000). Climatic warming in the Tibetan Plateau during recent decades. International Journal o f Biom eteorology, 1742, 1729-1742. Liu, X ., Cheng, Z., Yan, L., & Yin, Z.-Y. (2009). Elevation dependency o f recent and future minimum surface air temperature trends in the Tibetan Plateau and its surroundings. G lobal and P lanetary Change, 68(3), 164-174. doi:10.1016/j.gloplacha.2009.03.017 Lu, A ., Kang, S., Li, Z., & Theakstone, W. H. (2010). Altitude effects o f climatic variation on Tibetan Plateau and its vicinities. Journal o f Earth Science, 21(2), 189-198. d oi:10.1007/sl2583-0100017-0 Luce, C. H., Abatzoglou, J. T., & Holden, Z. a. (2013). The m issing mountain water: slower westerlies decrease orographic enhancement in the Pacific Northwest USA. Science (New York, N.Y.), 342(1360), 1360-1363. doi: 10.1126/science. 1242335 Lundquist, J. D., Dettinger, M. D., Stewart, I. T., & Cayan, D. R. (2009). Variability and Trends in Spring R unoff in the Western United States. In F. Wagner (Ed.), Clim ate w arm ing in w estern North A m erica-E vidence and environm ental effects (pp. 6 3 -7 6 ). University o f Utah Press. M acleod, S., & Dery, S. J. (2007). The Cariboo Alpine M esonet. CM OS Bulletin SCM O, 35(2), 4 5 51. Maness, H., Kushner, P. J., & Fung, I. (2012). Summertime climate response to mountain pine beetle disturbance in British Columbia. Nature G eoscience, 5(12), 1-6. doi:10.1038/ngeol642 Mann, H. B. (1945). Nonparametric tests against trend. Econom etrica, 13(3), 24 5 -259. Mann, M. E., Bradley, R. S., & Hughes, M. K. (1999). Northern Hemisphere Temperatures during the Past Millennium: Inferences, uncertainties, and limitations. G eophysical Research Letters, 26(6), 75 9 -7 6 2 . Martinec, J., & Rango, A. (1989). Merits o f Statistical criteria for the performance o f Hydrological M odels. Journal o f the Am erican Water Resources A ssociation, 25(2), 421 —432. d oi: 10.1111 /j. 1752-1688.1989.tb03079.x Maurer, M. K., Menounos, B., Luckman, B. H., Osborn, G., Clague, J. J., Beedle, M. J., ... Atkinson, N. (2012). Late H olocene glacier expansion in the Cariboo and northern Rocky Mountains, British Columbia, Canada. Q uaternary Science Reviews, 51, 7 1 -8 0 . d oi: 10.1016/j .quascirev.2012.07.023 McCuen, R. H. (2003). M odeling H ydrologic Change: Statistical M ethods. Lewis Publishers. McGuire, C. R., N ufio, C. R., Bowers, M. D., & Guralnick, R. P. (2012). Elevation-dependent temperature trends in the Rocky Mountain Front Range: changes over a 56- and 20-year record. P lo S O n e, 7(9), e44370. doi:10.1371/journal.pone.0044370 M cKenney, D. W., Hutchinson, M. F., Papadopol, P., Lawrence, K., Pedlar, J., Campbell, K., ... Owen, T. (2011). Customized Spatial Climate M odels for North America. Bulletin o f the Am erican M eteorological Society, 92(12), 1611-1622. doi:10.1175/2011BAM S3132.1 M cleod, A. A. I. (2013). Kendall: Kendall rank correlation and Mann-Kendall trend test. Meng, L., & Shen, Y. (2014). On the Relationship o f Soil Moisture and Extreme Temperatures in East China. Earth Interactions, 18( 1), 1-20. doi: 10.1175/2013E I000551.1 M ilewska, E. J. (2004). Baseline cloudiness trends in Canada 1953-2002. Atm osphere-O cean, 42(4), 2 6 7 -2 8 0 . doi: 10.3137/ao.420404 Modarres, R., & Sarhadi, A . (2009). Rainfall trends analysis o f Iran in the last h alf o f the twentieth century. Journal o f G eophysical Research, 114, D 03101. doi:10.1029/2008JD 010707 MoE. (2014). Effects o f Climate Change. Effects o f Clim ate Change in British Colum bia. Retrieved from Ministry o f Environmet, Government o f British Columbia. http://www.livesmartbc.ca/learn/effects.html MoF. (1979). B iogeoclim atic Zones and Subzones o f the C ariboo F orest Region. (M. Annas, R & R. Coup6, Eds.). M inistiy o f Forest,Provience o f British Columbia. Mondal, A ., Kundu, S., & Mukhopadhyay, A. (2012). Rainfall trend analysis by Mann-Kendall Test: A case study o f North-Eastern part o f Cuttack District, Orissa. International Journal o f G eology, 2(1), 7 0 -7 8 . M ooney, H., Larigauderie, A ., Cesario, M., Elmquist, T., Hoegh-Guldberg, O., Lavorel, S., ... Yahara, T. (2009). Biodiversity, climate change, and ecosystem services. Current Opinion in Environm ental Sustainability, 1(1), 46 -5 4 . doi:10.1016/j.cosust.2009.07.006 Moore, R. D., & Demuth, M. N. (2001). Mass balance and streamflow variability at Place Glacier, Canada, in relation to recent climate fluctuations. H ydrological P rocesses, 75(18), 3473-3486. d oi: 10.1002/hyp. 1030 Mountain Caribou Science Team. (2005). Mountain C aribou in British C o lu m bia: A Situation A nalysis (pp. 1-9). MW LA. (2002). Indicators o f Clim ate Change fo r British Colum bia 2002. Ministry o f Water,Land and Air Protection,British Columbia. Nash, J. E., & Sutcliffe, J. V. (1970). River flow forecasting through conceptual m odels part I - A discussion o f principles. Journal o f H ydrology, 10, 2 8 2 -290. NRCan. (2014). Regional, national and international climate modeling. Regional, national an d international clim ate m odeling. Retrieved April 07, 2014, from http://cfs.nrcan.gc.ca/projects/3?lang=en_CA Opdam, P., & Wascher, D. (2004). Climate change meets habitat fragmentation: linking landscape and biogeographical scale levels in research and conservation. B iological Conservation, 777(3), 2 8 5 -2 9 7 . doi: 10.1016/j.biocon.2003.12.008 Parks Canada. (2011). Conservation Strategy f o r Southern M ountain C aribou in C a n a d a ’s N ational Parks. Parks Canada. PCIC. (2014). Clim ate Summ ary f o r C ariboo Region. Retrieved from http://www.pacificclimate.org/sites/default/files/publications/Climate_Summary-Cariboo.pdf | 123 | Pederson, G. T., Gray, S. T., Ault, T., Marsh, W., Fagre, D. B., Bunn, A . G., ... Graumlich, L. J. (2011). Climatic controls on the Snowmelt Hydrology o f the Northern Rocky Mountains. Journal o f Clim ate, 24(6), 1666-1687. doi:10.1175/2010JCLI3729.1 Pepin, N., & Losleben, M. (2002). Climate change in the Colorado Rocky Mountains: Free air versus surface Temperature Trends. International Journal o f C lim atology, 22, 311 -3 2 9 . doi: 10.1002/joc.740 Pepin, N ., & Lundquist, J. D. (2008). Temperature trends at high elevations : Patterns across the globe. G eophysical Research Letters, 35, 1-6. doi:10.1029/2008G L034026 Picketts, I. M., Werner, A . T., Murdock, T. Q., Curry, J., Dery, S. J., & Dyer, D. (2012). Planning for climate change adaptation: lessons learned from a community-based workshop. Environmental Science & Policy, 17, 8 2 -9 3 . doi:10.1016/j.envsci.2011.12.011 Pike, R. G., Bennett, K. E., Redding, T. E., Werner, A. T., Spittlehouse, D. L., Moore, R. D. D., ... Tschaplinski, P. J. (2008). Clim ate Change Effects on W atershed P rocesses in British Colum bia (V ol. 3). Ministry o f Forests, Lands and Natural Resources Operations. R Core Team. (2014). R: A language and environment for statistical computing. R Foundation fo r S tatistical Computing, Vienna, Austria. Vienna, Austria.: R Foundation for Statistical Computing. Retrieved from http://www.r-project.org/. Ramanathan, V ., Cess, R. D., Harrison, E. F., Minnis, P., Barkstrom, B. R., Ahmad, E., & Hartmann, D. (1989). Cloud-radiative forcing and climate: results from the Earth radiation budget experiment. Science (New York, N.Y.), 243(4887), 57-63. d oi: 10.1126/science.243.4887.57 Rangwala, I., & Miller, J. R. (2012). Climate change in mountains: a review o f elevation-dependent warming and its possible causes. Clim atic Change, 114(2-4), 5 2 7 -5 4 7 . doi: 10.1007/s 10584012-0419-3 Rangwala, I., Miller, J. R., & Xu, M. (2009). Warming in the Tibetan Plateau: Possible influences o f the changes in surface water vapor. G eophysical Research Letters, 36(6), L06703. d o i:10.1029/2009G L 037245 Rangwala, I., Sinsky, E,, & Miller, J. R. (2013). Amplified warming projections for high altitude regions o f the northern hemisphere mid-latitudes from CMIP5 m odels. Environm ental Research Letters, 8(2), 024040. d oi: 10.1088/1748-9326/8/2/024040 Ropelewski, C. F., & Halpert, M. S. (1986). North American Preciptation and Temperature patterns associated with the El Nino/Southern Oscillation (ENSO). M onthly Weather Review, 114, 2 3 5 2 2362. Scherrer, S. C., Ceppi, P., Croci-M aspoli, M., & Appenzeller, C. (2012). Snow-albedo feedback and Sw iss spring temperature trends. Theoretical an d A p plied Clim atology, 110(4), 509 -5 1 6 . d o i: 10.1007 /s0 0 7 0 4 -0 12-0712-0 I 124| Sen, P. K. (1968). Estimates o f the Regression Coefficient based on Kendall’s Tau. Journal o f the Am erican S tatistical A ssociation, 63(324), 1379-1389. Retrieved from http://www.jstor.org/stable/2285891 Seneviratne, S. I., Liithi, D., Litschi, M., & Schar, C. (2006). Land-atmosphere coupling and climate change in Europe. N ature, 443(7108), 2 0 5 -2 0 9 . doi:10.1038/nature05095 Shi, X., Dery, S. J., Groisman, P. Y., & Lettenmaier, D. P. (2013). Relationships between recent PanArctic snow cover and Hydroclimate Trends. Journal o f Clim ate, 26(6), 2 0 4 8-2064. doi:10.1175/JC LI-D -12-00044.1 Shrestha, R. R., Schnorbus, M. A ., Werner, A. T., & Berland, A. J. (2012). M odelling spatial and temporal variability o f hydrologic impacts o f climate change in the Fraser River basin, British Columbia, Canada. H ydrological Processes, 26(12), 1840-1860. doi:10.1002/hyp.9283 Statistics Canada. (2014). Cariboo Regional District 2011 population census. Retrieved from http://w w w l2.statcan.gc.ca/census-recensem ent/2011/dp-pd/map-carte/index-eng.cfm Stevenson, S. K., Armleder, H. M., Arsenault, A ., Coxson, D., DeLong, S. C., & Jull, M. (2011). British C olu m b ia ’s Inland Rainforest. UBC Press. Vancouver.Toronto. Stewart, I. T., Cayan, D. R., & Dettinger, M. D. (2004). Changes in snowmelt runoff tim ing in western North America under a “Business as usual” Climate change scenario. C lim atic Change, 62, 21 7 -2 3 2 . Stewart, I. T., Cayan, D. R., & Dettinger, M. D. (2005). Changes toward earlier Streamflow Timing across Western North America. Journal o f Clim ate, 18, 1136-1155. Stocker, T. F., D.Qin, Plattner, G.-K., Alexander, L. V., Allen, S. K., Bindoff, N. L., ... Gregory, J. M. (2013). Technical summary. In T. F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, ... P. M. M idgley (Eds.), Clim ate Change 2013: The Physical Science Basis. Contribution o f Working Group I to the Fifth Assessm ent R eport o f the Intergovernm ental P anel on Clim ate Change. Cambridge, United Kingdom and N ew York, N Y , USA: Cambridge University Press. Retrieved from http://www.clim atechange2013.org/im ages/report/W GlAR5_TS_FlNAL.pdf Storch, H. Von, & Zwiers, F. W. (1999). Statistical Analysis in Clim ate Research. Cambridge University Press. Tohver, I. M., Hamlet, A. F., & Lee, S.-Y. (2014). Impacts o f 21 st-Century Climate Change on Hydrologic extreme in the Pacific Northwest region o f North America. Journal o f the Am erican W ater Resources Association(JAW RA), 1-16. doi:10.111 l/jaw r.12199 Tong, J., Dery, S. J., & Jackson, P. L. (2009a). Interrelationships between MODIS / Terra remotely sensed snow cover and the hydrometeorology o f the Quesnel River Basin, British Columbia, Canada. H ydrology a n d Earth System Sciences, 13, 1439-1452. | 125 | Tong, J., Dery, S. J., & Jackson, P. L. (2009b). Topographic control o f snow distribution in an alpine watershed o f western Canada inferred from spatially-filtered M ODIS snow products. H ydrology and Earth System Sciences, 13(3), 3 1 9 -3 2 6 . doi: 10.5194/hess-13-319-2009 Trenberth, K. E. (1990). Recent Observed Interdecadal Climate Changes in the Northern Hemisphere. Bulletin o f the Am erican M eteorological Society, 71(7), 988-993. Urrutia, R., & Vuille, M. (2009). Climate change projections for the tropical Andes using a regional climate model: Temperature and precipitation simulations for the end o f the 21st century. Journal o f G eophysical Research, 114( D 2), D 02108. doi: 10.1029/2008JD 011021 Vors, L. S., & Boyce, M. S. (2009). Global declines o f caribou and reindeer. G lobal Change B iology, 75(1 1), 262 6 -2 6 3 3 . doi: 10.1111/j. 1365-2486.2009.01974.x Vuille, M., & Bradley, R. S. (2000). Mean annual temperature trends and their vertical structure in the tropical Andes. G eophysical Research Letters, 27(23), 3885-3888. Vuille, M., Bradley, R. S., Werner, M., & Keiming, F. (2003). 20th Century Climate change in the Tropical Andes : Observations and Model Results. Clim atic Change, 59, 7 5 -9 9 . Vuille, M., Francou, B., Wagnon, P., Juen, I., Kaser, G., Mark, B. G., & Bradley, R. S. (2008). Climate change and tropical Andean glaciers: Past, present and future. Earth-Science R eviews, 59(3-4), 7 9 -9 6 . doi:10.1016/j.earscirev.2008.04.002 Walther, G., Post, E., Convey, P., M enzel, A ., Parmesan, C., Beebee, T. J. C., ... Bairlein, F. (2002). Ecological responses to recent climate change. Nature, 416, 389-395. Wang, Q., Fan, X., & Wang, M. (2013). Recent warming amplification over high elevation regions across the globe. C lim ate D ynamics, 43, 87—101. doi:10.1007/s00382-013-1889-3 Wang, T., Hamann, A ., Spittlehouse, D. L., & Murdock, T. Q. (2012). ClimateW NA— HighResolution Spatial Climate Data for Western North America. Journal o f A p plied M eteorology an d Clim atology, 57(1), 16-29. doi:10.1175/JA M C -D -l 1-043.1 Warren, S. G., Eastman, R. M., & Hahn, C. J. (2007). A Survey o f changes in cloud cover and cloud types over land from surface observations, 1971—96. Journal o f Clim ate, 20(4), 717—738. d oi: 10.1175/JCL14031.1 Weed, A. S., Matthew, P. A., & Hicke, J. A. (2013). Consequences o f climate change for biotic disturbances in North American forests. E cological M onographs, 55(4), 441—470. W esterling, A. L., Hidalgo, H. G., Cayan, D. R., & Swetnam, T. W. (2006). Warming and earlier spring increase western U.S. forest wildfire activity. Science (New York, N.Y.), 575(5789), 9 4 0 3. doi: 10.1126/science. 1128834 Whiteman, C. D. (2000). M ountain M eteorology: Fundamentals an d Applications. Oxford University Press. Wickham, H. (2009). ggplot2 E legant G raphics f o r D ata Analysis Use R!. Springer. I 126| Wickham, H. (2011). The Split-Apply-Combine strategy for data analysis. Journal o f Statistical Software, 40( 1), 1-29. Retrieved from http://www.jstatsoft.org/v40/i01/ Wikipedia. (2013). Cariboo Mountains Sub-ranges. Wikipedia,the fre e encyclopedia. Retrieved Decem ber 11, 2013, from http://en.wikipedia.org/wiki/Cariboo_Mountains Wikipedia. (2014). Barkerville,BC. Barkerville, B C. Retrieved September 17, 2014, from http://www.barkerville.ca/pages/HistoryO 1.html W ilks, D. S. (2011). S tatistical M ethods in the Atm ospheric Sciences (3rd ed.). Elsevier Inc. W illiams, M. W., Losleben, M., Caine, N ., & Greenland, D. (1996). Changes in climate and hydrochemical responses in a high-elevation catchment in the Rocky Mountains, USA. Lim nology an d O ceanography, 41(5), 939-946. doi: 10.4319/lo. 1996.41.5.0939 Wittmer, H. U., M clellan, B. N ., Seip, D. R., Young, J. A., Kinley, T. A ., Watts, G. S., & Hamilton, D. (2005). Population dynamics o f the endangered mountain ecotype o f woodland caribou ( Rangifer tarandus caribou) in British C olum bia, Canada. Canadian Journal o f Zoology, 83, 4 0 7 -4 1 8 . doi:10.1139/Z05-034 WMB. (2014). Weather Stations. M inistry o f Forests , Lands an d N atural Resource O perations Wildfire M anagem ent Branch Weather Stations. Retrieved January 10, 2014, from http://bcwildfire.ca/Weather/stations.htm You, Q., Kang, S., Pepin, N ., Fliigel, W .-A., Yan, Y ., Behrawan, H., & Huang, J. (2010). Relationship between temperature trend magnitude, elevation and mean temperature in the Tibetan Plateau from hom ogenized surface stations and reanalysis data. G lobal an d P lanetary Change, 77(1-2), 124-133. doi:10.1016/j.gloplacha.2010.01.020 Zhang, T., Osterkamp, T. E., & Stamnes, K. (1996). Some characteristics o f the Climate in Northern Alaska, U .S.A . A rctic and A lpine Reserach, 28(4), 509-518. Zhang, X ., Vincent, L. a., Hogg, W. D., & Niitsoo, A. (2000). Temperature and precipitation trends in Canada during the 20th century. Atm osphere-O cean, 38(3), 395^429. doi: 10.1080/07055900.2000.9649654 Zwiers, F. W., Schnorbus, M. A ., & Maruszeczka, G. D. (2011). H ydrologic Im pacts o f Clim ate Change on B C Water Resources. Pacific Climate Impacts Consortium, University o f Victoria, Victoria BC. | 127 | 8. Appendices Appendix A Validation of gridded data with observed data (daily frequency). C C C & C A M n e t d a ily T m in . in t h e C M R 2 0 - .............. N SE- 0.71 . RMSEb 4 2 7 'C . r« 0.89 (p - 0 ) .......................................................................................... ....................................................................................................... Browntop 0-2 0 -4 0 N SE- 0.87 , RMSE*2.87 *C , r» 0.96 (p« 0 ) E 20' ^ o- | |- 2 0 £ ® -40~ e 20' i o - NSE. 0 6 7 ,R y S E U » 2 X , r. 0 8 » ( P - ° ) t . I t . .1 j4 jA A il .i J j , . w |.2 0 - re q -4 0 20- N SE- 0.79 . RMSE- 3.71 *C . r* 0.92 (p « 6 ) o8 o -2 0 -4 0 £ & ^ <£> # . ^ o & S ’ c? O ^ ^ S> c? ^ 0 /■ ^ ^ 0 T im e [ D a y s ] CCC & CAMnet daily Tmax. in the CMR — Canadian Climate Coverage — Cariboo Alpine Mesonet f 20o- y -20- NSE* 0.36 ,RMSEa 6.96 ®C, r- 0.95 (p- 0 ) ^ J 0.97 (p* 0 ) 4 ill O ^ 20- | ° |- 2 0 - NSE- 0 .9 4 . RMSE- 2.59 "C , i WtH E i| l| 1 20- i s o>» « -2 0 - NSE- 0.7 9 . RMSE* 4.5 'C , rX .9 6 ( p - 0 ) ||| 20o-20- NSE* 0 .3 3 . RMSE" 6.86 *C , r* 0.94 ( p * 0 1 ^ ^ ^ * ^ o ^ Vs ^ 1 ^ ^ ^ o Time[Days] | 128 | $ O ^ ^ ^ O'5- CCC & CAMnet daily Precip. in th e CMR ^ C a n a d i a n Climate C o v e r a g e C a r ib o o Alpine Mesonet 30- NSE- -0.94 . RMSE- 3 2 2 mm , r- 0 2 7 (p« 0 ) 20* 10 - 30* N SE- 0 2 3 , RMSE- 2.67 mm . r- 0.55 (p - 0 ) 20- 30- NSE- -0.16 . RMSE- 3.69 m r . r- 0 2 9 (p - 0 ) 30- MSE- -0.98 , RMSE- 3.58 m n . r» 0 2 2 (p* 0 ) 10 - TimeJDays] I 129 | A ppendix B Spatial plots o f m onthly minimum and m axim um air temperatures. Monthly Tmin. (average 950-2010) over the CMR Mar Feb Apr I 53 I 54 Jan 52 I 'W ;'’ : \ i 0 Jut Jun Aug 54 May Latitude 53 I Temp.{°C] \ -5 52 I I I | : Oet rV -- Nw ...... Dm 52 I I 53 I 54 Sap ......... -122 -121 -120 -119 -122 -121 -120 -119 -122 -121 -120 -119 -122 -121 -120 -119 Longitude | 130 | Monthly Tmax. (average 1950-2010) over the CMR 122 -121 -120 -119 -122 -121 -120 -119 -122 -121 -120 -119 -122 -121 -120 -119 L ongitude A ppendix C Spatial plots o f m onthly precipitation. Monthly Prep, (average 1950-2010) over the CMR Prep.[mm] ■ 600 122 -121 -120 -119 -122 -121 -120 -119 -122 -121 -120 -119 L ongitude I 132 | A ppendix D M inim um and m axim um annual temperature and trend. Annual Tmin. trend in the CMR,1950-2010 - 2- Trend= 0.322 °C/decade p-value= 8.13e-06 -3 - M - 6 5 yrs moving average ~ Median B ased Trend - ^ YriyTmin. -7-\ Time[Year] Annual Tmax. trend in the CMR, 1950-2010 9 Trend= 0.193 °C/decade p-value=y.00732 7 H >» — >- 6 YrlyTmax. 5 yrs moving a \era g e “ Median B ased Trend 5 & Time[Year] |133| A ppendix E M ean annual air temperature anom aly and trend. Annual mean temp, anomaly in the CMR, 1950-2010 2. 0 - T rend- 0.27 ° C/decade p = 5.4e-05 >. 0.5o 0 .0 - £-0.5- 1. 0 - -1.5- 2. 0 - 5 yrs moving average “ Median based Trend -2.5|N eg ativ e Anom aly! Positive Anomaly Time [Year] Annual mean temp, trend in the CMR, 1950-2010 Trend= 0.2696 °C/decade p-value= 5.38e-05 ± A ^ » >» ” 5 yrs moving average “ Median Based Trend A YrlyTmean Time[Year] I 134 | ]' A ppendix F Seasonal minimum , m axim um , and m ean temperatures and trends. S easo n al Minimum trend in the CMR, 1950-2010 Trend= 0.62 "C/D ecade , pval= 0^ - 10- -15- 20- -21 Trend= 0.35 °C /d ecad e , pval= 0 * -4- 65 yrs moMng average — Median Based Trend Trend= 0.12 ° C /d e c a d e , pval= 0.$\ - Temp. 2- * -4- 6- Trend= 0 .1 ‘‘ C /d ecad e , pval= 0.44 Time[Year] I 135 | A * S easo n al maximum Temp, trend in the CMR, 1950-2010 ♦........................................ 4 A ♦ ,4> Trend= 0.39 ’C/decade , pval= 0.01 is >W o Irend= 0.15 ‘Q/decade .BAal=a.t5. E 22 0) H 20 i !l \ 'T . v 3. ■5 yrs moving average — Median Based Trend A Temp. A A 18- A o>i e ■3 3: 16 Trand= 0.11 ’C/decade . pval= 0.28 10.0 * 4 N* 7.5- ; . * * ‘Hi » u 5.0Trend= -0.04 °C/decade, pval= 0.74 2.5 ctf3 nfbN Tlm e[Y ear] | 136 | r»N Winter Spring. Fail Summer O! CM; I^ o m O) a: o C; t-* o> CL a * -■ o <11 ✓ ^ - <( m o in o o oo o in in csi o o in o cvi in A ppendix G Spatial trend o f mean annual air temperature. Annual Tmean trend in the CMR, 1950-2010 ■'3to Trend (°C d ecade 1) Significance * Sig. co m a) 3 co T3 CM in -122 -121 Longitude (°W ) I 138 | -120 -119 A ppendix H Spatial trend o f m ean seasonal temperature in the CM R, 1950-2010. W inter.T m ean -122 -121 S p rin g jm e a n -120 -119 Longitude Longitude S ig n ific a n c e T r e n d (“C d e c a d e 1) S ig . 0.3 0.4 0.5 0.6 0 7 * T re n d T C d e c a d e ) N o t s ig . 0.20 0.25 0.30 0.35 0.8 Fall.Tm ean S um m er.! m ean « ■» « * * * * # * ■ a ar Longitude T re n d ( X d e c a d e ) Longitude Trend ( " C d e c a d e S i g n i f i c a n c e S i g n if ic a n c e Sig. * Not stg 0.00 | 139 | 0.05 0.10 0.15 A ppendix I Annual and seasonal mean temperature trends vs elevation. Annual mean Temp, trend vs elevation in the CMR 2500 ♦ ♦ 2000 R2 u0 . 56 — ^ « * * * •*—-------- --------------------; h j -------------------------------------------------------------” - " I n n 11 A J mJ wEZ **-----------*— ♦♦♦ * U T T ----------------------------------------------------------------------------------- +W *L . +AP » » — ^ + »>—»------------------------------»-------------------------- - & A -A a i . *. . •. jf gl yS f* r f c ♦ ♦ 1500 I 1000 I Elevation (m) ! ♦ Sig.trend ** • '' V ♦ 1 3.10 1 0.15 1 0.20 a A # • ? * . * #* •• r .* % ? * + * 't ; V f f c & ^ 1 0.25 1 0.30 * * * ♦ * ? * Trend ( Cdecade ) | 140 | V '♦* . *«* ♦ ♦ 1 0.35 1 0.40 • Sig.trend ♦ Not sig. trend Trend [°Cdecade ]