MEASUREMENTS AND CONTROLS ON MID-WINTER ALPINE GROUND THERMAL REGIME IN THE PURCELL MOUNTAINS, BRITISH COLUMBIA by Kevin Michael Ostapowich B.Sc., Thompson Rivers University, 2020 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN THE NATURAL RESOURCES AND ENVIRONMENTAL STUDIES GRADUATE PROGRAM – ENVIRONMENTAL SCIENCE THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA April 2022 ©Kevin Ostapowich, 2022 ii ABSTRACT Alpine snow is an important water reservoir for mountain hydrology, climate, ecosystem functions, and has substantial economic value. Snowmelt during spring and summer is driven primarily by incoming shortwave and longwave radiation fluxes, and the ground heat flux is considered to be negligible during this time. However, during the accumulation phase, the ground heat flux may contribute to snowpack thermal conditions and midwinter melt, though this subject has not been studied extensively. The objective of this study is to quantify the alpine ground thermal regime and its relation to topographic setting and the overlying snowpack. The effects of elevation, maximum winter snow depth, snow cover duration, slope, ruggedness, aspect, total potential solar radiation, proximity to glacial ice, and depth of thermistor were evaluated. Four transects consisting of 29 temperature data loggers at the ground-snow interface and one meteorological station collected data from 16 August 2020 to 6 August 2021 at an alpine site in the Purcell Mountains in British Columbia . Snow cover duration, onset, and end-of- winter depths were found to have the greatest influence on the ground thermal regime. Total potential solar radiation had an inverted relation with ground temperatures, however, this was likely related to snow cover duration. Modeled ground heat flux scenarios revealed that snow depth or onset is the most influential of the variables tested. Snow thermal conductivity has the second greatest influence on total ground heat flux, however, true snow thermal conductivity likely varies throughout the winter season and was not measured in this study. Wind has the greatest influence on snow distribution within the Conrad basin with wind scoured slopes coincidentally sharing the same aspect as slopes that receive the greatest total potential solar radiation. There is little evidence to suggest that the ground thermal regime has any influence over the overlying snowpack. iii TABLE OF CONTENTS Abstract Table of contents ... ii List of Tables v List of Figures vi Acknowledgements . xi Chapter 1: Introduction 1 1.1 Research Statement and Objectives Chapter 2: Background 2 4 2.1 Snowpack Energy Balance 4 2.2 The Ground Heat Flux 4 2.2.1 Thermal conductivity 2.3 Thermal Effects of Snow 2.4 Potential controls on alpine ground thermal regimes Chapter 3: Study Area 7 9 11 12 3.1 Study Area Description 12 3.2 Climate Setting 13 3.3 Study Area Geology . 14 Chapter 4: Instrumentation and Data Collection 17 4.1 Instrumentation Overview 17 4.2 Transect layout. 21 4.3 Supporting Data 27 4.3 .1Digital Elevation Models - DEMs 28 4.3 .2 BC Hydro 28 iv 4.3 .3 Thermal Conductivity and Petrographic analyses 28 4.4 Meteorological Stations 30 4.5 GNSS Survey 32 Chapter 5: Analyses 34 5.1 Data Processing 34 5.1.1 HOBO processing 35 5.1.2 Met 1 Processing 35 5.1.3 GNSS Survey 36 5.1.4 DEM processing 36 5.2 Supporting Data Analyses 37 5.2.1Potential Solar Radiation 37 5.3 Collected Data Analyses 37 5.3 .1Snow -on/ off dates 38 5.3 .2 Thermal regime analyses 39 5.3 .3 Heat flux 43 5.4 Quality Control 45 5.4.1 HOBO dual temperature logger quality control 45 5.4.2 Field quality control at Met 1 47 5.4.3 SR 50A wind checks 50 Chapter 6: Results and Discussion 52 6.1 Supporting Data 52 6.1.1 BC Hydro Meteorological Station 52 6.1.2 Potential Solar Radiation 53 6.1.3 LiDARDEM 55 V 6.2 Collected Data 56 6.2.1 Ground Thermal Regime 60 6.2.2 Heat Flux 81 Chapter 7: Conclusion 87 7.1 Ground temperature regime descriptions 87 7.2 Ground thermal regime analysis 88 7.2.1 Snow cover days 88 7.2.2 Ground thermal regime 88 7.2.3 Ground-snow temperature gradients . 89 7.2.4 Effects of wind 90 7.3 Ground heat flux 7.3 .1 7.4 Midwinter ground water recharge and permafrost Summary 90 91 92 References 94 Appendices . 101 Appendix A - Thermal Analysis Labs 101 Appendix B - Site information table 105 Appendix C - Hourly temperature profiles for all sites 111 vi LIST OF TABLES Table 1: Thermal conductivities of snow per snow type ( Sturm et al., 1997 ) 8 Table 2: Thermal conductivities of rocks tested by TAL from various sites in Conrad Basin . 30 Table 3 : ClimaVUE 50 and SR 50A measurement specifications as per the manufacturers.... 31 Table 4: Results from quality control check at Met 1. 49 Table 5: BC Hydro meteorological station statistics for period of record. Note that the record started in October of 2018. Statistics for this period may be skewed 53 Table 6: Calculated SOD and SDD for each site. Number of SCD determined from difference between SDD and SOD. Number of SFD determined by subtracting SCD from 365 . Table 7: Met 1 summary. 54 Air values based on 15 - minute increment instantaneous measurements and wind values based on 15- minute average measurements recorded at Met 1. 58 Table 8: Annual ground ( approximately 15 cm below ground) and interface ( surface ) temperatures for each site. Mean and standard deviation values for all sites listed at bottom . 59 Table 9: OLS multiple linear regression results for ground temperatures at each period of interest , P-values < 0.05 highlighted in light blue . 68 Table 10: OLS regressions for gradient sites with solar radiation and snow cover days removed, p-values < 0.05 are highlighted in blue . 74 Table 11: Average heat flux ( HF) and total energy transfer ( ET) for deep snow site ( 200810101) from 1January, 2021 to 15 April, 2021. 85 Table 12: Average heat flux ( HF ) and total energy transfer ( ET) for shallow snow site ( 200811100) from 1January, 2021 to 15 April, 2021. 85 vii LIST OF FIGURES Figure 1-1: Conrad Glacier site location for field work. Background cartography by Stamen Design ( 2022 ) Figure 3-1: Folding in Mt. Thorington indicating metamorphism. 3 15 Figure 3 - 2: Rock samples from Conrad Glacier collected in August 2020. a ) is low grade metamorphic fine -grained sandstone/mudstone, b ) is quartzite, c) is fissile finegrained sandstone/ siltstone. d ) is low grade metamorphic fine-grained banded sandstone/ metapelite. All rocks identified by Tombe ( 2021). 16 Figure 4-1: Typical arrangement of HOBO thermistor installation in rock. Drawing not to scale. 18 Figure 4-2: Typical ground temperature site as installed in the field . 19 Figure 4-3: Typical arrangement for HOBO thermistor installation in glacial ice. PVC conduit used to keep thermistor at bottom of hole. Drawing not to scale , Figure 4-4: HOBO thermistor installation in glacial ice as installed in the field. 20 21 Figure 4-5 : Overview of transect layout. Individual HOBO sites shown as points. Meteorological station shown as purple triangle . 22 Figure 4- 6: Transect 1. Numbered points indicate installed ground temperature monitoring sites . 23 Figure 4-7: Transect 2. Numbered points indicate installed ground temperature monitoring sites . 24 Figure 4-8: Transect 3. Numbered points indicate installed ground temperature monitoring sites. Background imagery from drone flight August 2021, courtesy of Joseph Shea 25 Figure 4-9: Site 200814100. Left image is install date in August 2020, right image is how instrumentation was found during retrieval in August 2021. Instrumentation viii was likely disturbed by ice or debris movement in flow direction as visible in image on right. 26 Figure 4-10: Transect 4. Numbered points indicate installed ground temperature monitoring sites . 27 Figure 4-11: Rock sample locations. Sample sites shown as blue dots with transect alignment shown for reference. Figure 4-12: Met 1 as installed in field. 29 Conrad Glacier and Mt Thorington visible in background Figure 4-13 : Survey control locations. Triangles denote control location . 32 33 Figure 5 -1: Met 1 ground temperature data with SR 50 snow depth data, and calculated days with snow cover using a modified method ( 2.0° C range over 48 hours) described by Raleigh et al. ( 2013 ). Some discrepancies still exist but this is the method that most closely matches SR 50A snow depth records. 38 Figure 5-2: Three typical ground thermal regimes. Hourly ground ( 0.15 m below surface ), interface ( surface ), and snow ( 0.10 m above surface ) temperature profiles shown for each typical regime. Deep snow ( a ) likely has early season established snow. Shallow snow ( b) likely has a snowpack established late in the season. Intermittent snow ( c) has little to no snow established throughout winter. ... 41 Figure 5 -3: Ice bath test. Left image shows thermistors in bath after ice melt. Image on right shows running experiment with towel to block solar radiation and prevent differential heating. 46 Figure 5 -4: Results from ice bath HOBO dual temperature logger quality control check. Figures b ) and c) plotted after water bath temperature stabilization. All temperature data are hourly. 47 Figure 5-5 : Met 1 quality control arrangement. Site " Met 1" is the actual ground thermal regime monitoring site. Other numbers indicate the serial numbers of HOBO dual temperature loggers used as quality control checks. 48 ix Figure 5 -6: Met 1 interface vs quality control interface hourly temperatures over a sample period (1 June 2021 - 20 June 2021). Met 1 interface temperature is from a HOBO Pendant and quality control interface temperature is from 20171152 HOBO dual temperature thermistor. Note that the pendant consistently records higher daily maximum temperatures than the HOBO dual temperature thermistor. 50 Figure 5-7: Met 1 snow depth and wind speed at 15 minute intervals from 28 January 2021 to 5 February 2021. 51 Figure 6-1: BC Hydro meteorological station hourly record of air temperature and snow depth for 31 October 2018 - 1 September 2021. Horizontal line indicates 0°C air temperature. 52 Figure 6- 2: Total potential solar radiation based on SFD at each site as per Table 6. Note that the HOBO dual logger at site 200813104 failed. SFD were not able to be determined for this site. 55 Figure 6-3: Median LiDAR snow depths in a 5 m radius around each site for 29 May 2020 and 2 May 2021. Note that sites 200810102- 200810105 were outside the extents of the LiDAR DEM for both years, however, the ground temperature profiles for these sites indicate little to no snow accumulation. 55 Figure 6-4: Met 1summary ( a ) Air temperature, ( b) incoming solar radiation, ( c ) snow depth, , ( d) wind speed, ( e ) precipitation, and ( f ) atmospheric pressure 15 minute data from 9 August 2020 - 6 August 2020. 57 Figure 6 - 5: Linear regressions between ground temperatures and ( a ) elevation, ( b) snow depth, ( c ) total potential solar radiation, ( d ) ruggedness, ( e) slope, ( f ) sensor depth, and (g) snow cover days for all non-ice sites during the Winter period of interest. 62 Figure 6 - 6: Linear regressions between ground temperatures and ( a ) elevation, ( b ) snow depth, ( c ) total potential solar radiation, ( d ) distance to ice, ( e ) ruggedness, (f ) X slope, (g) sensor depth, and ( h ) snow cover days for Transect 3 sites during the Winter period of interest. 64 Figure 6-7: Comparison between November, February, April, and WET periods of interest individual regressions, ( a ) Elevation, ( b ) snow depth, ( c) total potential solar radiation, ( d ) ruggedness, ( e ) slope, and ( f ) ground temperature depth ( sensor depth ) are examined . 67 Figure 6-8: Hourly thermal gradient regimes for typical sites: ( a ) Deep, ( b ) shallow, ( c ) no/intermittent snowpacks are shown. Calculated snow cover days are shown as gray shading in background. Note that the y-axis scales are not consistent over the three figures. Positive values indicate heat transfer from ground to snow . 71 Figure 6-9: Individual regressions of temperature gradient versus ( a ) elevation, ( b ) snow depth, ( c ) total potential solar radiation, ( d ) ruggedness, ( e ) slope, and (f ) number of snow cover days for each period of interest . 73 Figure 6-10: Snow distribution in Conrad Basin for May 2020 ( left image) and May 2021 ( right image). Red shading indicates snow depth 75 Figure 6-11: Polar plots of ( a ) maximum winter snow depth ( m ) versus aspect and ( b) median hourly ground temperature (°C) over the Winter period versus aspect 76 Figure 6-12: Area of interest for snow depth versus aspect raster analysis . Magenta dashed line indicates boundary. Snow depth map overlaid as red shading. 77 Figure 6-13 : Lag analysis between hourly ground and air temperatures at Met 1. Hourly ground and air temperatures for ( a ) November, ( c) February, ( e) April, (g) July periods of interest. Corresponding lag analysis with Pearson Correlation Coefficients between ground and air temperatures for ( b) November, ( d ) February, ( f ) April, and ( h ) July at an hourly time scale , 79 Figure 6-14: Lag analysis between hourly ground temperatures and hourly median incoming solar radiation measured at Met 1. Hourly ground temperatures and incoming xi solar radiation for ( a ) November, ( c) February, ( e) April, and ( g) July periods of interest. Corresponding lag analysis with Pearson Correlation Coefficients between ground temperatures and incoming solar radiation for ( b ) November, ( d) February, (f ) April, and ( h ) July at an hourly time scale. 80 Figure 6-15 : Hourly heat flux calculated for ( a ) phyllite and ( b ) quartzite thermal conductivity scenarios and three snow thermal conductivity scenarios at deep snow site 200810101 during the Winter period. expressed in W nr1 K 1. Thermal conductivities shown are 83 Figure 6-16: Hourly heat flux calculated for ( a ) phyllite and ( b ) quartzite thermal conductivity scenarios and three snow thermal conductivity scenarios at shallow snow site 200811100 during the Winter period. Thermal conductivities shown are expressed in W nr1 K 1. 84 xii ACKNOWLEDGEMENTS Firstly, I would like to thank my supervisor Dr. Joseph Shea for all his guidance and camaraderie through this project, for without whom, none of this would have come to fruition. I would also like to thank my committee, Dr. Marten Geertsema and Dr. Stephen Dery, for providing excellent feedback and contributions. Thank you to Dr. Brian Menounos and Dr. Ben Pelto for providing invaluable LiDAR data, Sean Tombe for providing geological expertise in identifying rock samples, and Sara Darychuk and Alex Bevington for providing supporting data and coding assistance in the project. Special thanks go to Mischa Fisher for providing help and direction on all things statistics related. Big thanks to the entire UNBC cryosphere crew who have provided an excellent platform for discussion, feedback, and learning in everything snow and ice related. Last but most importantly I would like to thank Jen, Cole, and Logan for providing support and having infinite patience throughout the entire MSc experience. This work was funded through research grants provided by NSERC and Mountain Water Futures. 1 CHAPTER 1: INTRODUCTION Alpine snow plays important roles in the hydrology, climate, and ecosystem functions of mountain basins, and also provides a large economic value ( Pomeroy and Brun, 2001; Sturm et al., 2017 ). During spring and summer, mountain snowpacks reflect a large proportion of the incoming shortwave radiation and modify the local climate. Snow cover also insulates the ground surface and limits energy exchange between the ground and the atmosphere. As a seasonal freshwater reservoir, snow stores water through the accumulation season and releases it during the melt season, and can be a critical source of streamflow and groundwater recharge ( Pomeroy and Brun, 2001). While mountain hydrology is highly sensitive to climate change, the snow regime ( the pattern of snow accumulation, ablation, redistribution, and duration of snow-covered and snow-free periods) is more sensitive than streamflow ( Rasouli et al., 2014). Snow hazards, such as avalanches, pose not only a threat on human safety but also pose a threat on infrastructure. Snow avalanches cost on average 1.25 million Euros per year due to damage to infrastructure in the canton of Grisons, Switzerland alone (Fuchs and Brundl, 2005 ) . Snow -glide avalanches are particularly hazardous due to the mass of snow usually transported and are associated with free water at the base of the snowpack ( Clarke and McClung, 1999 ). A better understanding of the hydrological processes occurring at the base of the snowpack will aid in understanding processes involved in natural snow hazards. Spring and summer snowmelt regimes are dominated by net shortwave and incoming longwave radiation and turbulent fluxes ( Marks and Dozier, 1992; Cline, 1997; Lapo et al., 2015; Bilish et al., 2018 ). However, heat fluxes from the ground into the snowpack can contribute to substantial basal melt during winter ( Marks and Dozier, 1992; Dingman, 2002). There is much published work regarding shortwave and longwave contributions to snowmelt as well as the other variables of the snowmelt energy budget and the effects that the overlying snowpack has on the ground temperature regime ( Marks and Dozier, 1992; Cline, 1997; Zhang, 2005; Bilish et al., 2018). However, there is little information available in the published literature regarding the effect that ground thermal regime has on the overlying 2 snowpack. The heat flux from the ground into overlying snowpacks is often neglected in snow energy balance calculations ( Cline, 1997 ). However, the influence of the ground thermal regime in alpine environments may affect midwinter alpine hydrological processes, alpine permafrost processes and extents, and glacial retreat due to advection from nearby bare soils and rock (Zhang, 2005; Jiskoot and Mueller, 2012; Gruber et al., 2015 ). Neglecting the ground thermal regime may potentially disregard a significant contribution on a range of mountain processes. For the purposes of this study the ground thermal regime, or ground temperature regime, will be considered as temporal patterns of ground temperatures. Diurnal and seasonal variations in ground temperatures will be assessed in this study. 1.1 Research Statement and Objectives Midwinter basal melt is potentially an important contribution to midwinter streamflow. To understand potential melt contributions an understanding of the alpine ground thermal regime and the magnitude of the ground heat flux is critical. This research quantifies the alpine ground thermal regime and evaluates the environmental factors that determine winter ground temperatures and temperature gradients between the ground and the base of winter snowpacks. I also examine the sensitivity of the modeled ground heat flux to physical and thermal properties of snow and the underlying bedrock. My hypothesis is that the alpine ground thermal regime will be affected by elevation, slope, and aspect, and sites that absorb more solar radiation during the summer will release more heat into the base of the winter snowpack. Specific objectives include the following: I. Conduct fieldwork and collect one year of ground-snow interface temperature data II. Describe alpine ground temperature regimes III. Analyze the factors that control variations in the alpine ground thermal regime IV. Assess the sensitivity of the modeled ground heat flux to thermal conductivity assumptions and snow depth. 3 This study involves field work at Conrad Glacier, located approximately 50 km south of Golden, British Columbia ( BC) in the Purcell Mountains ( Figure 1-1). The study site is described in greater detail in Section 3. o Figure 1-1: Conrad Glacier site location for field work. Background cartography by Stamen Design ( 2022). 4 CHAPTER 2: BACKGROUND 2.1 Snowpack Energy Balance Snowmelt is driven by an energy exchange between the snowpack and its environment ( DeWalle and Rango, 2008 ). The net energy balance of snow can be calculated as = + + + + + + + [ Eq. 1] where SN = net energy balance of snow, K = net shortwave radiation, L = net longwave radiation, H = turbulent exchange of sensible heat, LE = turbulent exchange of latent heat, R = heat input by rain, G = conductive exchange of sensible heat with the ground, AM = the loss of latent heat storage term due to melting or refreezing within the snowpack, and A S = change in heat storage within the snowpack ( Oke, 1987; Dingman, 2002; DeWalle and Rango, 2008). Fluxes are given in units of W nr 2 and fluxes into the snowpack are positive. Net shortwave and incoming longwave radiation and turbulent fluxes dominate the spring and summer snowmelt regimes of temperate snowpacks ( Marks and Dozier, 1992; Cline, 1997; Lapo et al., 2015; Bilish et al., 2018 ), but shortwave radiation provides the majority of the energy flux into the snow ( Pomeroy and Brun, 2001). The contribution that the ground provides to the energy budget of the snowpack during these times is often considered negligible or ignored ( Marks and Dozier, 1992; Cline, 1997; Dingman, 2002; DeWalle and Rango, 2008). This is because soil, while usually a much better heat conductor than snow, is generally considered a poor conductor of heat. Ground temperatures are low as they are less influenced by solar radiation due to the snow cover. However, during the accumulation season the relative contribution from ground heat flux increases as the contribution from the other terms in the net surface energy balance decrease ( DeWalle and Rango, 2008). 2.2 The Ground Heat Flux Ground heat flux has been found to significantly contribute to the snowpack energy balance and midwinter melt ( Lafaysse et al., 2011). Snow with higher internal temperatures suppress heat transfer from the ground upwards into the snow thus permitting more of the heat to be 5 used for basal melt. Snowpacks with colder internal temperatures create stronger basal temperature gradients where more heat is transferred away from the base leaving little heat available for ground melt ( Smith, 2011). Understanding heat storage in the ground, and how this energy interacts with the overlying snowpack has not been well documented and is often ignored ( Dingman, 2002; Lafaysse et al., 2011). The reverse, or the influence the overlying snow has on ground temperatures, is discussed in further detail in section 2.3. A simple, one-dimensional ground heat flux through Fourier' s law can be defined as = × d d [ Eq. 2] where G = the rate at which heat is conducted upward to the base of the snowpack ( W nr 2), kg = thermal conductivity of the soil or bedrock ( W nr1 K 1), TG = soil temperature ( K ), and z = ' the distance below the ground surface ( m ) ( Dingman, 2002). This equation is overly simplistic as it does not account for the properties of the overlying snowpack. Only the thermal properties of the ground up to the interface between the ground and the snow are considered. The ground heat flux when including thermal properties of snow as well the ground can be defined as = 2 , ( , + − , , ) [ Eq. 3 ] where G = heat transfer by conduction and diffusion between snow and soil or bedrock ( W rrr 2 ), Kes,l = effective thermal conductivity of basal snow ( W nr1 K 1), Keg = effective thermal ' conductivity of soil or bedrock ( W m 1 K 1), Tg = soil or bedrock temperature ( K ), Ts,l = basal ' ' snow temperature ( K ), zs,l = thickness of basal snow layer ( m ), zg = depth below surface of ground temperature measurement ( m ) ( Marks and Dozier, 1992 ). 6 The effective thermal conductivity can be defined as = +[ ] [ Eq. 4] where Ke = effective thermal conductivity of the ground or snow, K = thermal conductivity of the ground or snow, Lv = latent heat of vaporization or sublimation (J kg 1), De = effective ' vapour diffusion coefficient ( m 2 s 1), q = specific humidity of the ground or snow (g kg 1, ' ' dimensionless) ( Marks and Dozier, 1992). Marks and Dozier (1992) found that when corrected for vapour diffusion, the thermal conductivity of snow increased by a factor of 20. Alternatively, Calonne et al. ( 2011) and Sturm et al. (1997 ) both considered the effective thermal conductivity as a function of snow density. In this study however, the measurements required to calculate the effective thermal conductivity, either using vapour diffusion or density of snow, are outside of the scope of the project. For all further heat flux calculations in this study, it is assumed that the effective thermal conductivity equals the thermal conductivity ( Ke=K ). To investigate the significance that ground heat flux has on an overlying snowpack, Mickschl ( 2016) compared calculated ground heat flux values from on site measurements to modeled ground heat flux values. Large variations in the calculated heat flux, due to the range of methods employed, led to low confidence in the results ( Mickschl, 2016), but the ground heat flux was found to contribute substantially to the energy balance during the accumulation season by supplying enough energy to base of the snowpack to cause basal melt. One limitation of ground heat flux observations is that ground temperatures at shallow depth thermistor locations ( 5 - 20 cm ) may not vary significantly but the variations in measurement depth can greatly affect the calculated ground heat flux ( Mickschl, 2016). Thermistor measurements may be similar regardless of whether the sensors are 5 cm, 10 cm, or 20 cm below the surface. The value for d or ( − , ) used in Equation 2 and Equation 3 respectively, will not vary substantially. However, if d , or zg, is 20 cm or 10 cm then the calculated value of G will vary substantially. 7 Ground heat flux is difficult to quantify because of variability in soil moisture content and soil thermal conductivity. It is often considered negligible during the snowmelt season but can be hydrologically significant during the accumulation season ( Cline, 1997; Dingman, 2002; Lafaysse et al., 2011). Meltwater generated by heat in the ground in the accumulation season can create substantial groundwater recharge which can contribute substantially to the annual water budget of a drainage basin. While winter streamflow can be sustained primarily by melt at the base of the snowpack ( Federer, 1965 ), heat conduction from the ground into the snowpack is a minor source of energy for melt is because soil is generally a poor conductor of heat ( DeWalle and Rango, 2008). This may not be true of extensive exposed bedrock in alpine environments. Exposed bedrock will be targeted for the thermistor installation in this study and is discussed in greater detail in section 4. 2.2.1 Thermal conductivity The thermal conductivity of bedrock and snow are important to consider in ground heat flux studies as it is a substantial component of the ground heat flux calculation ( Equations 2 and 3 ). The thermal conductivity of rock is highly variable between rock types ( igneous, sedimentary, and metamorphic ) and also variable within each rock type. It is dependant on mineral composition, sediment texture, porosity, anisotropy, and properties of pore filling fluids (Midttpmme and Roaldset, 1999; Balkan et al., 2017 ). Thermal conductivity in published literature often varies linearly with porosity ( Robertson, 1988). Variations in sedimentary rock composition are so great that Midttpmme and Roaldset (1999) and Balkan et al. ( 2017) conclude a universal model of thermal conductivities is impossible to develop. For example, sandstone can range in thermal conductivity from 0.38 to 6.50 W m 1 K 1 ( Schon, 2015 ). Thermal conductivity of snow is also significantly variable. Pomeroy and Brun ( 2001) state a typical thermal conductivity for dry snow is 0.045 W nr1 K 1, however, this value may not be ' representative because thermal conductivity has been shown to vary logarithmically with snow density ( Sturm et al., 1997; Calonne et al., 2019 ). Sturm et al. (1997 ) reviewed 13 regression equations to estimate thermal conductivity of snow and also developed their own. The authors state that the previous regression equations were weak because they describe 8 thermal conductivity as a function of snow density alone. However, other snow attributes such as grain size, bonding, and temperature also influence snow thermal conductivity (Sturm etal., 1997 ). There is substantial variation of snow thermal conductivities in the published literature. Table 1 is a summary from Sturm et al. (1997 ) showing thermal conductivities and snow densities by snow type. It is important to note that these "snow types" are a generalization and somewhat subjective classifications based on interpretations of the authors of the study. Table 1: Thermal conductivities of snow per snow type (Sturm et al., 1997). Snow Type Thermal Conductivity ( W nr1 K 1) Density (g cm 3 ) 0.452 0.359 0.237 0.167 0.250 0.188 0.183 0.072 0.153 0.163 0.169 0.128 0.070 0.488 0.444 0.379 0.348 0.422 0.290 0.345 0.225 0.321 0.345 0.320 0.254 0.135 ' very hard slab hard drift hard slab moderate slab melt grain clusters rounded melt grains indurated depth hoar chains of depth hoar mixed forms large rounded small rounded recent snow new snow ' While snow density is generally less spatially variable than snow depth, it can still account for 5 - 25% of error in snow water equivalent ( SWE ) estimations and has been found to significantly change throughout the day during the melt cycle ( Dingman, 2002; Lopez -Moreno et al., 2013 ). Incorporating properties such as grain size and bonding type to achieve higher accuracy estimates of snow thermal conductivity, as proposed by Sturm et al. (1997 ), is challenging without additional field observations and measurements throughout the winter. For these reasons, calculating the ground-snow heat flux, while important, will not be the 9 primary focus of this study. The variability of the parameters used to calculate the heat flux cannot be accurately assessed within the scope of this project. The focus of this study is to quantify the influence the ground thermal regime has on alpine snowpacks. There are numerous published works describing the influence the snowpack has on the ground temperature and the thermal regime (Goodrich, 1982; Zhang, 2005; Li et al., 2016), however, there are fewer works describing the influence that the ground thermal regime has on the overlying snowpack. This study will help fill that void in the published literature. 2.3 Thermal Effects of Snow Zhang ( 2005 ) provides a comprehensive study of the influence that the snowpack has on the ground thermal regime. Ground thermal regimes can be substantially affected by the presence or absence of snow. Snow acts as an excellent thermal insulator between the atmosphere and ground. The effect that snow has on the ground thermal regime depends on timing and duration of snow accumulation and melt processes. Thickness, density, and structure of seasonal snow cover, micrometeorological interactions, local microrelief and vegetation, and geographic location also affect the ground thermal regime ( Zhang, 2005 ). The thermal influence winter snow cover has over the ground thermal regime is primarily attributed to low thermal conductivity of snow, high surface albedo, and the energy sink created by latent heat of fusion during snowmelt ( Luetschg et al., 2008). Snow can have a significant effect on ground temperatures when snow depth is less than 0.5 m ( Zhang, 2005 ). Thin snow cover can result in ground surface temperatures cooling throughout the winter (Morse et al., 2012). Li et al. ( 2016 ) found that a snowpack depth of 0.4 m was sufficient to insulate the ground surface from the changes in air temperatures above the snowpack. This is greater than the 0.20 - 0.25 m snow depth suggested by DeWalle and Rango ( 2008) as the maximum snow depth at which shortwave radiation can transmit through the snowpack to influence the ground temperature. Other studies have shown that snow depths greater than 0.8 m are critical for thermal insulation ( Keller and Gubler, 1993; Rodder and Kneisel, 2012). These variations in snow depth in the published literature suggest that the morphological differences of the snow are important in controlling how much 10 radiation may be transmitted through the snowpack. These snow depth values may also vary due to differences in density, snow structure ( such as grain size ), or snow morphology such as a homogeneous snowpack or a blocky, debris laden, heterogeneous snowpack ( Zhang, 2005; Rodder and Kneisel, 2012). This phenomenon can also be influenced by geomorphic characteristics of the underlying ground. Farbrot et al. ( 2011) found that ground temperatures were significantly cooler under snow covered block fields than smooth ground surfaces. They hypothesized that the air voids in the block fields allowed colder air to circulate beneath the snow causing the ground surface to cool. Snow cover has also been found to lead to cooler ground surface temperatures. Shallow snow on steep terrain, in some instances, results in cooler ground surface temperatures due to increased albedo, emissivity, latent heat consumption, and by acting as a barrier to solar radiation ( Hasler et al., 2011). Shallow snowpacks in alpine environments where mean annual ground temperatures are below 0°C may also facilitate permafrost formation /continuation. Permafrost is defined as soil or rock remaining below 0°C for at least two consecutive years (Gruber et al., 2017 ). Including highly variable topography in alpine basins, other important factors influencing alpine permafrost distribution are air temperature, potential direct solar radiation, and snow cover ( Hoelzle, 1996; Etzelmuller et al., 2001; Stocker-Mittaz et al., 2002). Alpine permafrost is highly influenced by climatic warming and as its occurrence is often in areas with high relief, degradation often results in significant hazards to human lives and infrastructure ( Etzelmuller et al., 2001; Gruber et al ., 2017 ) . Snow distribution also has a significant impact on alpine and tundra flora and fauna . Deep snow aids in plant survival due to the insulating properties of snow. Some plant species have adapted to shallow or intermittent cold snow cover by trapping snow in dead tissues to use as insulation, altering snow cover distribution patterns. Environments with deeper snowpacks, where ground surface temperatures are near or at 0°C, provide refuge and habitat for vertebrates and invertebrates active on, in, and under the snow (Jones, 1999 ). The temporal distribution of snow accumulation also has significant impacts on short- and long-term ground temperatures ( Zhang, 2005 ). Snowpack establishment in autumn is likely 11 to result in a snow insulating effect where ground temperatures remain warmer. Snowpack establishment in late winter can have the opposite effect and may result in cooling of the ground thermal regime ( Zhang, 2005 ). One implication of this is that under a warming climate, delayed snowpack establishment could actually lead to reduced ground surface temperatures, and a reduction in the ground heat flux to the base of the snowpack. Conversely, earlier snow disappearance in spring could lead to increased heat storage in the bedrock. 2.4 Potential controls on alpine ground thermal regimes While snow cover, depth, and duration are considered to have a large impact on the ground thermal regime (Goodrich, 1982; DeWalle and Rango, 2008; Li et al., 2016), other controls include topographical characteristics. Aspect and slope will both determine how much of the ground surface is available for heating by direct solar radiation. Shading from vegetation would also be important to consider, however, as this study is focused on alpine terrain, effects from vegetation are negligible and will be accounted for in instrumentation layout. Terrain shading on the other hand is very important to consider in these high relief mountain environments and will also affect how much direct solar radiation is available to heat the ground surface. As air temperatures typically change between 0.5°C and 1.0°C per 100 vertical meters ( Dingman, 2002), elevation and atmospheric conditions will also affect the ground thermal regime. 12 CHAPTER 3: STUDY AREA 3.1 Study Area Description The study site is located at Conrad Glacier in the Purcell Mountains in British Columbia ( BC) ( Figure 1-1). The Conrad Glacier basin is a sub- basin of the Columbia River basin, which covers 668 000 km2 through portions of BC and seven western states in the United States ( US) ( Payne et al., 2004; Pelto et al., 2019 ). Conrad Glacier is 11.45 km2 ( Pelto et al., 2019 ). Water from the Columbia River is used for hydropower production, irrigation, and navigation. There are 214 dams on the system with the capacity to generate 36,400 MW of electricity ( Payne et al., 2004). Approximately 1.4 million hectares of agricultural land are irrigated with the water from the Columbia River and its tributaries ( Payne et al., 2004). The Conrad Glacier basin consists of glacierized alpine terrain with exposed till, bedrock, and sparse alpine forest. Conrad Creek flows from the terminus of Conrad Glacier which joins Vowell Creek before joining the Spillimacheen River and ultimately the Columbia River. The study site falls within a portion of Conrad Glacier Basin that ranges in elevation from 2252 m to 2596 m, ranges in slope from 4° to 34° with a median slope of 15 °, and ranges in slope aspect from 46° to 311° from with a median of 135 ° from North. Conrad Glacier has been studied extensively in recent years. Previous studies in the area include Pelto et al. ( 2019) seasonal mass balance studies utilizing airborne geodetic surveys. Menounos et al. ( 2020) included Conrad Glacier in a study of Columbia River Basin ice and snow response to climate change. Fitzpatrick et al. ( 2019 ) included Conrad Glacier in an investigation of glacier surface roughness lengths. Menounos ( 2021) is continuing to study the mass balance of Conrad Glacier. Ongoing monitoring at the site includes a BC Hydro Automatic Weather Station ( AWS) on the nunatak surrounded by Conrad Glacier as well as ongoing alpine permafrost studies with continuous data collection led by Dr. Marten Geertsema and Alex Bevington of MFLNRORD. Three sites near the existing BC Hydro station on the nunatak currently have temperature loggers installed and have been collecting hourly ground temperature data since August 2016. The Water Survey of Canada operates a hydrometric gauging station on the 13 Spillimacheen River near the confluence of the Columbia River ( FLNRORD, 2021). The University of Northern British Columbia ( UNBC) has been collecting Light Detection And Ranging ( LiDAR ) data in Conrad basin since 2014. LiDAR has been collected every winter for the purpose of snow depth monitoring and mass balance calculations ( Pelto et al., 2019; Menounos et al., 2020). Other relevant works near Conrad Glacier include Fitzpatrick et al. ( 2017 ) surface energy balance study on Nordic Glacier in the Purcell Mountains which used field collected data to generate a surface energy balance model. The Canadian Avalanche Association ( CAA ) has InfoEx snow monitoring stations in nearby basins as does the BC Snow Survey Program (Government of BC, 2020; Menounos et al., 2020). Canadian Mountain Holidays ( CMH) has an operational lodge near the site works. CMH has a via ferrata route used in glacier tours adjacent to our work site near the terminus of Conrad Glacier. CMH has been contacted regarding the nature of the work near their operational area and is supportive of the project. The study area falls within the traditional territory of Ktunaxa First Nations. 3.2 Climate Setting The Conrad Glacier basin falls within the Columbia Mountains and Highlands ecoregion in the Montane Cordillera ecozone ( Ecological Framework of Canada, 2020). The site also falls within the Engelmann Spruce - Subalpine Fir and Interior Mountain - Heather Alpine zones of the Biogeoclimatic Ecosystem Classification ( BEC ) ( B.C. Ministry of Forests and Range, 2020 ) . The mean annual temperature at the lower meteorological station ( 2319 m) during the study period (10 August 2020 - 10 August 2021) was -0.8° C while the upper meteorological station ( 2599 m ) recorded a mean annual temperature of -2.9°C. This was warmer than the previous year where the mean annual temperature for the upper elevational extent was -4.5 °C for 10 August 2019 - 10 August 2020. Total precipitation as rain measured at the lower elevational extent during the study period was 217 mm. 14 ClimateBC ( version 7.10, using normal 1991-2020 ) shows the mean annual air temperature at the lower elevational extent to be -0.7°C and -2.2° C at the upper elevational extent which compare relatively well with the measurements from onsite meteorological stations during the study period ( Wang et al., 2016). ClimateBC shows the mean summer precipitation to be 366 mm and 402 mm, precipitation as snow to be 951 mm and 1159 mm, summer mean maximum temperature to be 12.3°C and 10.0°C, and winter mean minimum temperature to be -10.5°C and -11.4°C for lower elevational extents and upper elevational extents, respectively ( Wang et al., 2016). Permafrost likelihood in the study area ranges from "only in very favourable conditions" to "in most conditions" as per modeling by Hasler and Geertsema ( 2013 ). 3.3 Study Area Geology Bedrock geology at the study site consists of coarse clastic sedimentary rock belonging to the Horsethief Creek Group, part of the Windermere Assemblage. Notably, and consistent with plutonic structures found throughout the Omineca belt, approximately 6 km southeast of the study site lies an igneous intrusion consisting of granite, granodiorite, and monzonite that dates from the early to late Cretaceous ( Hoy et al., 1994). There have been no published geological studies done in the immediate area of the study site at Conrad Glacier. However, through synthesizing the available published information the geological history of the area can be inferred as follows: The northern Purcell mountains were a site of continental rifting and deposition on the supercontinent of Rodinia during the late Proterozoic coinciding with a period of extensive glaciation as interpretations of the Windermere Assemblage would suggest ( Miller et al., 1973; Brennan et al., 2020) . More locally around what would become Conrad Glacier, was a site consistent with near-shore, subaqueous deposition during the Neoproterozoic ( Poulton and Simony, 1980). During the formation of the Omineca belt, the rocks making up the Horsethief Creek group were deeply buried as the North American continent was driven over the oceanic plate. Sediment accumulations on the cratonal margin were thrust eastward 15 ( Poulton and Simony, 1980 ). Due to the great depth at which these rocks were buried, various grades of metamorphism are common among the Windermere Assemblage ( Price, 2000; Evenchick et al., 2007 ). This is consistent with what was observed in the field. Figure 3 -1 shows significantly folded layers in the peak of Mt. Thorington at the head of Conrad basin. Figure 3-1: Folding in Mt. Thorington indicating metamorphism. Figure 3 - 2 shows samples collected in August 2020 from the Conrad Glacier site which show various grades of metamorphism in predominantly fine-grained sandstone/ siltstone. 16 Figure 3- 2: Rock samples from Conrad Glacier collected in August 2020. a ) is low grade metamorphic finegrained sandstone / mudstone , b ) is quartzite , c ) is fissile fine-grained sandstone/ siltstone. d ) is low grade metamorphic fine-grained banded sandstone/metapelite. All rocks identified by Tombe ( 2021). The Windermere Assemblage rocks deformed and accumulated in a series of thrust faults as the compression due to tectonic plate motions continued until 59 Ma ( Price, 2000; Simony and Carr, 2011). After 59 Ma, crustal extension occurred, causing the formation of the Rocky Mountain Trench, located approximately 30 km east of Conrad Glacier, as well as facilitating exhumation of the deeply buried rocks ( Price, 2000; Brown and Gibson, 2006; Simony and Carr, 2011). The Purcell anticlinorium ( a large anticline in which a series of minor folds are superimposed) was formed as these orogenic processes occurred. Uplift and erosion exposed the Horsethief Creek group at the Conrad Glacier site ( Price, 2000). Continual erosion from glaciations have shaped the area into the dynamic topography that is currently visible today. 17 CHAPTER 4: INSTRUMENTATION AND DATA COLLECTION This section describes how and where the meteorological station and ground temperature instruments were installed in the field, what data are collected, and data collection interval. Individual logger and transect locations are discussed. Supporting data and their sources are discussed as well. 4.1 Instrumentation Overview Ground and surface temperature data were collected over an elevation range from 2266 to 2610 Meters Above Sea level ( MASL), a difference of 344 m, to quantify the thermal regime between the ground and the overlying snowpack. HOBO Pro v2 U 23-003 2 x ( dual ) External Temperature Data Loggers were used to measure ground ( approximately 15 cm below ground surface ) and near-surface ( approximately 10 cm above ground surface ) temperatures. HOBO Pendants UA-001-64 were used to measure ground surface, or ground-snow interface, temperatures. These sensors were chosen because they are cost -effective and have proven to be reliable in similar alpine studies ( Bevington, 2020). HOBO Pro v2 U 23 data loggers have an operating range between -40 to 70°C and achieve an accuracy of ±0.21°C from 0 to 50°C according to the manufacturer's specifications. HOBO Pendants UA -001-64 have an operating range of -20 to 70°C and achieve an accuracy of ±0.53 °C from 0 to 50 °C according to the manufacturer' s specifications. HOBO data loggers sampled and recorded temperatures at one-hour intervals. Instrumentation took place at the study site from 10 August to 15 August 2020. Ground temperatures were measured approximately 15 cm below the surface by drilling into bedrock with a handheld rock drill and installing one of the thermistors from the HOBO dual temperature logger. Basal snow temperatures were measured by installing the other thermistor approximately 10 cm above the ground surface on wood dowelling anchored into the bedrock ( Figure 4-1). Wood dowelling was chosen as an anchor over rebar as it is easier to work with and biodegradable despite being weaker and more prone to breaking when shear stress is applied. Aluminum tape and PVC conduit was used to protect exposed data logger cables from wildlife. Installation of thermistors in sediments or soils was avoided to 18 minimize the effect that variable soil moisture would have on ground temperatures and heat transfer. Bedrock sites, as free from cracks and fractures as possible, were preferentially selected along the transects to minimize the thermal effect of infiltrating water on ground temperature measurements. Rock samples were collected from three specific ground temperature sites and one general site for petrologic and thermal conductivity analysis. These data were used for heat flux calculations. THERMISTOR 1/ 2" WOOD DOWELLING ANCHOR % GROUND ' SURFACE sir * SILICONE PLUG THERMISTOR Figure 4-1: Typical arrangement of HOBO thermistor installation in rock. Drawing not to scale. Figure 4- 2 shows a typical ground temperature monitoring site as installed in the field. Each site was given a unique label consisting of the date of installation and a unique identifier. Using Emlid Reach RS GNSS receivers with horizontal and vertical accuracies of 7 mm + 1ppm and 14 mm + 2 ppm, respectively, a GNSS survey point was collected at each site giving an accurate global coordinate. 19 Figure 4-2: Typical ground temperature site as installed in the field. Two sites were also installed in Conrad Glacier. One of the HOBO dual temperature thermistors was installed at approximately 50 cm below the ice surface, one thermistor at approximately 100 cm below the ice surface and a HOBO pendant on the surface of the ice. This is an ancillary study to observe thermal gradients and interactions at and near the ice surface. The typical arrangement is shown in Figure 4-3. 20 r HOBO Pro ICE SURFACE THERMISTOR ATTACHED TO PVC CONDUIT Figure 4-3: Typical arrangement for HOBO thermistor installation in glacial ice. PVC conduit used to keep thermistor at bottom of hole. Drawing not to scale. Figure 4- 4 shows the thermistors as installed in the field. 21 SP"^ v‘ w m % /* ' ;, > • - -^ l '• Figure 4-4: HOBO thermistor installation in glacial ice as installed in the field. 4.2 Transect layout In August 2020, 29 HOBO Pro v 2 U 23-003 2 x External Temperature Data Loggers were installed over four transects ( Figure 4- 5 ) . The sites range in elevation from 2253 MASL to 2596 MASL. These sites were chosen to maximize consistency of bedrock outcrop that sensors were installed into while achieving variation in elevation and slope aspect. HOBO sites along Transects 1 and 2 were spaced out as near to 50 vertical meters as possible given field conditions with competent exposed bedrock outcrops. Transect 3 was established to determine the ground thermal regime of recently deglaciated terrain. Transect 4 added northwest facing sites on a nunatak in the middle of Conrad Glacier and also two sites in the surface of Conrad Glacier. Figure 4- 5: Overview of transect layout. Individual HOBO sites shown as points. Meteorological station shown as purple triangle. Transect 1 is shown in Figure 4- 6. The sites were chosen based on 50 vertical meter spacing surveyed with the Emlid Reach RS GNSS RTK receiver and the presence of competent bedrock outcrops. Transect 1 was originally planned to go over the ridge and descend the northwest slope to include a larger sample representation of the northwest aspect, however, there was little to no bedrock on the majority of the northwest facing slope which led to the decision to abandon that leg of Transect 1. 23 Figure 4-6: Transect 1. Numbered points indicate installed ground temperature monitoring sites. Transect 2 is shown in Figure 4-7. Similar to Transect 1, sites along Transect 2 were chosen based on 50 vertical meter spacing surveyed with the Emlid Reach RS GNSS RTK receiver and the presence of competent bedrock outcrops. The alignment of the transect was chosen to increase sample points along a similar elevational gradient while increasing slope aspect diversity. Nine HOBO sites are spread along the transect which trends in a southwest direction from the meteorological station parallel to the glacier to the upper extents of the basin. Figure 4-7: Transect 2. Numbered points indicate installed ground temperature monitoring sites. Transect 3 is shown in Figure 4-8. The HOBO sites installed along Transect 3 will be to determine how proximity to glacial ice affects the ground thermal regime, and has a smaller elevational spacing to achieve a high spatial resolution of the ground thermal regime versus elevation gradient. Seven HOBO sites are spaced at approximately two-meter vertical increments along Transect 3, beginning at the ice margin where the first site was installed in bedrock approximately 2 m below the surface of ice as of August 2020. Figure 4-9 shows the extent of glacial ice loss at this site over the study period. . . Figure 4-8: Transect 3 Numbered points indicate installed ground temperature monitoring sites Background imagery from drone flight August 2021, courtesy of Joseph Shea. 26 Figure 4-9: Site 200814100. Left image is install date in August 2020, right image is how instrumentation was found during retrieval in August 2021. Instrumentation was likely disturbed by ice or debris movement in flow direction as visible in image on right. Transect 4 is shown in Figure 4-10. Two HOBO sites are installed in Conrad Glacier. These are a supplementary study to observe thermal patterns in annual surface and near-surface glacial ice. Two HOBO sites are also installed on the nunatak to provide additional slope aspects to the primary study. Note that the background imagery in Figure 4-10 displays the glacier at greater extents than is current. Site 200811106 is installed in rock near the ice margin, not in the ice itself. Figure 4-10: Transect 4. Numbered points indicate installed ground temperature monitoring sites. While the intention of this study is to quantify the ground thermal regime and its effects on snow across a variety of parameters, elevation has been chosen as the parameter of focus. The transects were chosen to best maximize the elevation gradient available while still achieving sites of exposed bedrock for instrument installation, sufficient variation in slope aspect, and areas with minimal avalanche hazard to the instrumentation. 4.3 Supporting Data The supporting data are provided by third parties in this study and not directly collected in Conrad Basin. These include LiDAR DEMs provided by colleagues, publicly available satellite data, and publicly available climate data. 28 4.3.1 Digital Elevation Models - DEMs Remote sensing data were utilized to calculate slope, aspect, ruggedness, and total potential solar radiation utilizing LiDAR DEM data provided by Menounos and Pelto ( 2021) and Shuttle Radar Topography Mission (SRTM ) DEM data retrieved from EarthExplorer published by the US Geological Survey ( EROS, 2018). The LiDAR DEM has a resolution of 1m with vertical and horizontal uncertainties of ± 0.15 m ( Pelto et al., 2019 ). The SRTM DEM has a horizontal resolution of 1 Arc-Second or 30 m and a minimum vertical accuracy of 16 m ( Mukul et al., 2017 ). Total potential solar radiation was calculated using the SRTM DEM data due to greater coverage of Conrad Basin than the LiDAR DEM. All other DEM analyses were conducted using the LiDAR DEM as it is higher resolution than the SRTM DEM. Snow depth DEMs, provided by Menounos and Pelto ( 2021), provided maximum winter snow depths for winter 2019-2020 and 2020- 2021 based on LiDAR measurements for bare earth in August and snow depths in May. The bare earth DEM was captured on 26 August 2020. Winter DEMs were captured 29 May 2020 and 2 May 2021. All LiDAR DEMs were coregistered. Snow depth was determined by DEM differencing of the August DEM from the May DEMs for each year. 4.3.2 BC Hydro Meteorological data from a BC Hydro weather station located on the nunatak in Conrad Glacier are available on the web and were accessed to provide additional climate context in the upper elevations of the study area for the duration of the study period (Government of BC, 2020) . This weather station provides snow depth and air temperature data . 4.3.3 Thermal Conductivity and Petrographic analyses Three rock samples were collected from the Conrad Glacier study area for petrographic and thermal conductivity analyses ( Figure 4-11). One additional sample ( dubbed "Conrad General") was taken from a location that approximates well banded rock, commonly seen across Conrad Basin. Figure 4-11: Rock sample locations. Sample sites shown as blue dots with transect alignment shown for reference. Thermal conductivity testing was performed by Thermal Analysis Labs (TAL). They employed the Modified Transient Plane Source ( MTPS) method. Of the four samples sent to TAL, samples Conrad General and 200811100 were tested at -20, 0, and 20° C. Samples 200810102 and 200813105 were tested at 0°C. The test results are summarized in Table 2 and the full test report from TAL is available in Appendix A. Average values were calculated for sites with three tests. 30 Table 2: Thermal conductivities of rocks tested by TAL from various sites in Conrad Basin Sample Geology Conrad General banded quartzite & phyllite 200811100 quartzite 200810102 phyllite w /minor 200813105 phyllite Temperature tested (°C) Thermal conductivity ( W nr1 K 1) -20 0 20 - 20 0 20 5.076 5.132 4.993 4.658 4.54 4.282 0 2.856 2.856 0 2.297 2.297 quartz ' Average conductivity ( W nr11C1) 5.067 4.493 A visual petrographic analysis was performed by Sean Tombe, Regional Geologist, NW Region at BC Ministry of Energy, Mines and Petroleum Resources. These rock thermal conductivity data were used primarily for determining heat flux and also by providing indicators to better describe the geologic history and setting for the Conrad Basin study area. 4.4 Meteorological Stations The installed meteorological station ( Met 1) consists of a ClimaVUE50 compact digital weather sensor and a SR 50A sonic distance sensor both from Campbell Scientific. Meteorological variables measured by the ClimaVUE 50 include air temperature, relative humidity, vapor pressure, atmospheric pressure, wind speed and direction, total incoming solar radiation, and precipitation as rain. Snow depth data were collected by the SR 50A. The measurement specifications are described in Table 3. 31 Table 3: ClimaVUESO and SR50A measurement specifications as per the manufacturers. Measurement Air temperature Relative humidity Vapour pressure Atmospheric pressure Wind speed Wind direction Solar radiation Precipitation Snow depth Range Resolution Accuracy -50 to 60°C 0 to 100% 0 to 47 kPa 500 to 1100 hPa 0 to 30 m s 1 Oto 359° 0 to 1750 W m 2 0 to 400 mm hr1 0.5 - 10 m o .rc ±0.6° C ±3% RH ±0.2 kPa ±5 hPa Greater of 0.3 m s 1 or 3% ±5 ° ±5% of measurement ±5% of measurement Greater of ±1cm or 0.4% 0.1 % 0.01 kPa 0.1 hPa 0.01 m s 1 1° 1W m 2 0.017 mm 0.25 mm ' The location of the Met 1 is shown in Figure 4-5 and is installed at 2319 MASL. Data were recorded by a Campbell Scientific CR 300 datalogger with 10 second sampling and 15 minute averaging in Pacific Standard Time. All instruments were powered by a 20 Watt solar panel from Campbell Scientific ( model number: SP 20-L10 ) and mounted on Campbell Scientific's UT10 3 m tower. The ClimaVUE50 was installed on the top of the mast at approximately 3.5 m above ground and the SR 50A was installed on a cross arm at approximately 3.2 m above ground to ensure sufficient clearance from the maximum expected snowpack depth ( Figure 4-12 ). A HOBO temperature gradient site was also installed at Met 1 to record ground, surface, and snow temperatures, with two additional HOBO dual temperature loggers installed for quality control. The purpose of these sensors is to test how much variation is likely to occur in similar rock with similar attributes ( slope, aspect, elevation ). This is discussed further in section 5.4.2. Two additional HOBO dual temperature loggers were also installed at the bottom of the tower to measure snow temperatures. However, this site did not consistently have snow accumulation on the ground due to wind redistribution so these measurements were not used in analyses. 32 Figure 4-12: Met 1 as installed in field. Conrad Glacier and Mt Thorington visible in background. 4.5 GNSS Survey Emlid Reach RS GNSS receivers were used to collect coordinates for all sites in the field. These are Real Time Kinematic ( RTK ) receivers that when used in conjunction with a base station can achieve centimeter positional accuracy ( EMLID, 2022 ). Two receivers were used. One was set as a base station and one was set as a RTK rover that was used to survey individual HOBO sites as they were installed. Data collection at the base station occurred while the RTK rover collected coordinates for each HOBO site. Control was initially established by placing a nail in the ground at control site UNBC 01, located near camp, and occupying with the base station on 10, 13, and 14 August, 2020. Control point locations are shown in Figure 4-13. The base was moved to UNBC 02 on 11 August, 2020 and occupied the point for 10 hours and 31 minutes while HOBO site installation and RTK survey took place. All survey data were post processed after returning from the field. This is discussed further in section 5.1.3. Figure 4-13: Survey control locations. Triangles denote control location. 34 CHAPTER 5 : ANALYSES Data analyses occurred as a two-step process. The first step was to process the raw data from the field to get them into a useable form. The second was the analyses of the processed data. This included calculations such as gradients, regressions, and heat flux. 5.1 Data Processing Climate and ground temperature data downloaded from the field were mostly in good condition with few errors or missing data. One HOBO Dual temperature logger failed completely where no data could be retrieved. Three HOBO dual temperature loggers failed in the spring after snow disappearance, likely due to water infiltrating the data logger. As this study is concerned with the ground thermal regime and the interaction with the overlying winter snowpack, the data during periods of no snow cover are of less interest. Hence the data in the three loggers that failed in the spring are still quite useful. After an initial look through ground temperature profiles, date ranges were established as periods of interest. These consist of (1) 15-day ranges that generally had consistent snow cover at most HOBO sites, ( 2 ) an overall winter period, and ( 3 ) a Winter Equilibrium Temperature ( WET) period defined as a period when constant ground surface temperatures develop at the end of winter due to significant snow cover decoupling the ground from the atmosphere ( Bosson et al., 2015; Sattler et al., 2016 ). In previous studies, WET has also been used as an indicator of local permafrost occurrence ( Sattler et al., 2016). In this study, I define WET as the mean ground surface temperature over the first two weeks in April. The five periods of interest in this study are defined as: a) 15- 30 November 2020 b) 1-15 February 2021 c) WET: 1-15 April 2021 d) 15 - 30 April 2021 e) Winter: 14 October 2020 - 10 June 2021 All meteorological and temperature data were truncated to these ranges for analyses. 35 5.1.1 HOBO processing Once all the data were downloaded in the field, all processing was performed using Python. Little was done to the raw data collected from the field. By visual inspection, much of them were intact, with few to no anomalies. However, HOBO logger condition varied from site to site when retrieved from the field, for example, some sites had broken dowels. This meant that the "snow" temperature sensor had been lying directly on the ground surface since an unknown date and did not reflect temperatures 10 cm above ground surface nor could they be used in heat flux calculations. Site specific conditions are summarized in Appendix B. To minimize errors due to site condition, HOBO logger data were separated into the following categories based on the condition of the logger and inferred winter snow depth regime. a) snoWET - Sites to be used in WET period of interest looking primarily at ground temperatures. Sporadic snow cover or sites with no snow removed. On glacier sites removed. Includes sites with broken dowelling. b) snoGRD - Sites to be used in gradient analyses and calculations. Sites on glacier and with broken dowels removed. Sporadic snow cover or sites with no snow removed. c) snoGLC - Includes only sites on glacier. d) sitesALL - All ground temperature sites excluding sites on glacier. These site categories, combined with periods of interest were then used for analyses described in section 5.3 . 5.1.2 Met 1 Processing Meteorological data from Met 1 required little cleanup. Non numbers and infinites were removed and replaced with Not -a -Number ( NaN) values. The data from the SR 50A snow depth sensor, however, were noisy. Periods of snowfall seemed to interfere with the acoustic sensor giving false snow depth readings appearing as spikes in the data. This was cleaned up by removing data points greater than a quality value of 300 in the raw data as per the manufacturer's specifications. Quality values from 300 to 600 are defined as "high measurement uncertainty" as per the manufacturer. Any measurements above 3.0 cm 36 before 1 September 2020 were also changed to NaN. Gaps in the data were filled in by interpolating between the last good quality point to the next using the cubic spline method of interpolation. 5.1.3 GNSS Survey The raw static data from each base station was sent into Natural Resources Canada ( NRCAN ) Canadian Spatial Reference System Precise Point Positioning online tool ( CSRS- PPP ). A static processing mode using the International Terrestrial Reference System and Frame ( ITRF) was chosen as it closely aligns with the World Geodetic System ( WGS), the latest revision being WGS 84. This system was chosen because other data used in this study, such as SRTM or lidar provided by Menounos and Pelto ( 2021), were in WGS 84. Coordinates were then projected into Universal Transverse Mercator (UTM) to simplify mapping and utilize linear units, in this case meters . This project lies within UTM zone 11 North. The RTK surveys were then imported into QGIS and shifted into place to match the processed UTM coordinates of the static control survey. Point UNBC 02 was used as primary control as the errors outlined in the CSRS-PPP report are the smallest. RTK survey points were shifted to this coordinate and checks on the other control points were all within the expected errors as stated in the CSRS- PPP report: the horizontal accuracy of the survey is ±0.195 m for the northing and ±0.326 m for the easting. All surveys were completed using instrument and rod heights to measure elevation, however, only horizontal measurements were used in data analyses. All snow depth and other elevation data were derived from DEM raster calculations where elevation data were measured directly from DEM raster surfaces. 5.1.4 DEM processing SRTM DEM data were reprojected into WGS 84/UTM zone 11N after they were downloaded using the EarthExplorer web service. A cubic resampling method was chosen to remove striping that is inherent with reprojecting SRTM DEM data . Whitebox Geospatial Analysis Tools version 3.4 was used to generate a watershed boundary for the Conrad basin. A buffer of 500 m was applied outside the boundary and the SRTM DEM was clipped to this buffer . This allowed some working room for the total potential solar radiation algorithms to run in 37 the basin while greatly reducing the run time compared to using the entire SRTM DEM as downloaded from EarthExplorer. All LiDAR data were processed by Brian Menounos and Ben Pelto who provided the data at a 1 m grid resolution. The bare earth LiDAR DEM was used to generate slope, aspect, and ruggedness rasters. Ruggedness is defined as the heterogeneity of terrain, quantified by calculating the elevation difference between a center cell and the eight surrounding cells ( Riley et al., 1999 ). These rasters were used to extract specific characteristics at each site. 5.2 Supporting Data Analyses This section consists of all analyses performed on third party data . 5.2.1 Potential Solar Radiation Total potential solar radiation was calculated using SRTM DEM data in QGIS running the GRASS command sun.insoltime. This function generates daily total potential solar irradiance and irradiation sums based on topographic shading, slope, aspect, geographic location, and sun location in the sky as a function of day of the year. The generated values do not account for cloud cover or changes in atmospheric transmissivity. This function was run over every day of the year to generate yearly totals. However, each site receives different amounts of solar radiation at the ground surface due to differing snow cover regimes. Snow cover duration was based on the ground temperature profile at each site and is discussed in greater detail in section 5.3.1. Seasonal potential solar radiation totals were then calculated for each site based on the snow-free duration. These values are used in correlative analyses discussed in section 5.3.2. 5.3 Collected Data Analyses Data were recorded at all sites between 16 August 2020 and 6 August 2021. None of the data were retrieved before the end date. This section describes the analyses of all data collected on site at Conrad Basin . 38 5.3.1 Snow-on/off dates Snow cover duration, or alternatively snow free days, is necessary to determine days available for solar radiation to directly add heat energy to the bare ground. This is necessary at each individual site as snow accumulation and disappearance occurs at different times throughout the basin. Shallow ground temperature diurnal ranges can be used to determine the presence of snow given that snow has a low thermal conductivity making it an excellent thermal insulator ( Lundquist and Lott, 2008). Raleigh et al. ( 2013 ) identified the presence of snow by considering days of snow cover as those that fall within a 48 hour period where the diurnal range in shallow ground temperature did not exceed 1.0°C. When testing this method against Met 1, the 1.0°C range did not capture all days with snow cover the SR50A picked up. A 48-hour period with a temperature range of less than or equal to 2.0°C matched much more closely with collected data at the meteorological station as shown in Figure 5 -1. 40 - 20 30 - u 20 - £ 10 - E o- _ - 15 ? - 10 -g £ Q Ground T ( ° C) i -i o i- Snow cover -20 - Snow depth - 5 § £ - 30 - 0 2020-09 2020-11 2021- 01 2021- 03 2021-05 2021-07 Figure 5-1: Met 1 ground temperature data with SR 50 snow depth data, and calculated days with snow cover using a modified method (2.0°C range over 48 hours) described by Raleigh et al. ( 2013 ). Some discrepancies still exist but this is the method that most closely matches SR 50A snow depth records. Other methods of determining snow cover days were also investigated but were less successful at matching the snow cover days recorded by the SR50. For example, Danby and Hik ( 2007 ) defined the Snow On Date ( SOD ) as the beginning of the start of a period of three consecutive days where the 24 hour ground temperature variance was less than or equal to 1.0°C. Using daily variance did not achieve as close of a match with the SR 50 as using a 48- hour range less than or equal to 2.0°C of ground temperatures. 39 After determining the range of snow cover days using Raleigh et al. ( 2013 ), SOD was determined as the first day of the first two- week period of continual snow cover for the season following Staub and Delaloye ( 2017 ). The Snow Departure Date ( SDD ) was determined as the last day of snow cover in the spring of 2021. However, this method did not calculate SOD for all sites. Sites with intermittent snow cover that did not have a two - week period with snow cover resulted in no SOD or SDD calculated. All dates were visually checked against corresponding plots and adjusted if necessary. Potential solar radiation at each site was then re-calculated for the snow free period by summing all the daily total potential solar radiation rasters for each snow free period. It is important to realize that this method results in an approximation of the snow free days in 2020, but is based on site conditions from August 2020 to August 2021. As no ground temperature data exist for spring of 2020, calculating the true number of snow free days for summer 2020 is not possible. The total potential solar radiation at each site for the site -specific snow free period was then used in regression analysis described in section 5.3.2. 5.3.2 Thermal regime analyses The ground thermal regime and its interactions with the overlying winter snowpack is the primary focus for this study. This section describes the analyses used to describe the ground thermal regime and factors controlling it. Discussed below are typical thermal regimes, gradient calculations, regression analyses, and lag analyses. All statistical analyses were performed in Python. 5.3 .2.1 Typical Regimes Three typical thermal regimes presented themselves upon initially plotting the HOBO data . I have labelled them as : a ) Deep snowpack: a site with early season established snow, deep enough to insulate from atmospheric conditions. b ) Shallow snowpack: a site with either shallow snow throughout the season where ground temperatures are still affected by atmospheric conditions throughout most of 40 the winter season or a site with late developed snowpack where ground temperatures dropped substantially in early winter. c ) Intermittent snowpack: a site with either no snow throughout the winter or intermittent snowpack development, likely redistributed away from site via wind. As there are snow depth data at each site only on the date of the LiDAR flight, some inferences have been made based on the temperature profiles of the sites and snow depth at the time of the LiDAR flight. The three typical profiles are shown in Figure 5 - 2. Appendix C shows all other site ground temperature profiles. 41 Deep snowpack - 2.6m a) 40 - u CL> 20 :s 2fD &E 11 0 0) - 20 - Site: 200810101 40 - b) Shallow snowpack - 1.4m L : U ~ G -4 0.2 - 0.0 2021-01- 28 2021-01-30 2021- 02 - 01 2021- 02 - 03 5 0 2021- 02 - 05 Figure 5-7: Met 1snow depth and wind speed at 15 minute intervals from 28 January 2021 to 5 February 2021. 52 CHAPTER 6: RESULTS AND DISCUSSION Results from all data collection and analyses are described and discussed below. Interpretations and inferences are also included in this section. 6.1 Supporting Data Results from all third-party data analyses are discussed in this section. 6.1.1 BC Hydro Meteorological Station BC Hydro' s meteorological station on the nunatak has been in operation since 31 October 2018. The mean annual temperature and snow depth for the study period of August 2020 to August 2021 at - 2.9°C and 132.0 cm, respectively (Table 5 ). It is important to note that as data collection began in October 2018, it is likely that the statistics for period 2018-2019 are skewed. A visual check of the data ( Figure 6-1) indicates, however, that the pattern of median snow depth decrease and mean air temperature increase are accurate. 4- Snow depth Air temperature - 20 - 10 u § 3£ T3 0 2 - - 10 g. E § - S £ 2 1 — 20 I— 0- - -3 0 2019-03 2019-11 2020-07 2021- 03 Figure 6-1: BC Hydro meteorological station hourly record of air temperature and snow depth for 31 October 2018 - 1 September 2021. Horizontal line indicates 0°C air temperature. 53 Table 5: BC Hydro meteorological station statistics for period of record. Note that the record started in October of 2018. Statistics for this period may be skewed. Annual periods Mean temp Median winter snow depth ( cm ) October 2018 - August 2019 ( incomplete ) August 2019 - August 2020 - 5.0 185 -4.5 149 August 2020 - August 2021 - 2.9 132 6.1.2 Potential Solar Radiation The results from the total potential solar radiation accounting for snow free days ( Figure 6- 2 ) show sites with a large variation in total potential solar radiation. Note that the calculations for total potential solar radiation at site 200813104 could not be performed as there is no ground temperature data for that site due to a logger failure. Snow free days are based on site conditions from August 2020 to August 2021 which are used to estimate likely snow free duration for summer of 2020. SOD, SDD, Snow Cover Days (SCD), and Snow Free Days ( SFD ) are shown in Table 6. 54 Table 6: Calculated SOD and SDD for each site. Number of SCD determined from difference between SDD and SOD. Number of SFD determined by subtracting SCD from 365. Site 200810100 200810101 200810102 200810103 200810104 200810105 200811100 200811101 200811102 200811103 200811104 200811105 200811106 200813100 200813101 200813102 200813103 200813104 200813105 200813106 200813107 200813108 200814100 200814101 200814102 200814103 200814104 200814105 200814106 Metl SOD 31/12/ 2020 11/ 10/ 2020 11/ 10/ 2020 11/ 10/ 2020 11/ 10/ 2020 11/ 10/ 2020 13/10/ 2020 11/ 10/ 2020 11/ 10/ 2020 11/ 10/ 2020 11/ 10/ 2020 11/ 10/ 2020 01/ 11/ 2020 11/ 10/ 2020 11/ 10/ 2020 11/ 10/ 2020 11/ 10/ 2020 NAN 11/ 10/ 2020 09/10/ 2020 09/10/ 2020 11/ 10/ 2020 11/ 10/ 2020 08/11/ 2020 11/ 10/ 2020 11/ 10/ 2020 11/ 10/ 2020 16/12/ 2020 11/ 10/ 2020 11/ 10/ 2020 SDD 01/01/ 2021 24/06/ 2021 02 / 02/ 2021 02 / 02/ 2021 20/ 02/ 2021 28/ 02/ 2021 10/ 06/ 2021 20/ 06/ 2021 29 /05/ 2021 04/07/ 2021 26/ 06/ 2021 12 /06/ 2021 02 / 02/ 2021 19 /05/ 2021 28/06/ 2021 02/ 02/ 2021 09 /05/ 2021 NAN 29 /04/ 2021 24/06/ 2021 02/ 02/ 2021 25/05/ 2021 04/07/ 2021 29 /05/ 2021 24/06/ 2021 02 / 02/ 2021 24/06/ 2021 12 /06/ 2021 20/ 06/ 2021 02 /02/ 2021 SCD 0 256 114 114 132 140 240 252 230 266 258 244 93 220 260 114 210 NAN 200 258 116 226 266 202 256 114 256 178 252 114 SFD 365 109 251 251 233 225 125 113 135 99 107 121 272 145 105 251 155 NAN 165 107 249 139 99 163 109 251 109 187 113 251 55 „ 1750000 - E 1500000 .c g 1250000 O 1000000 - ' S =5 750000 - fD 500000 - £ $ 250000 0 O - invooiHrsim - invoi OrH ^ o o o r H r M m O- iOn vO ^O- iOn oO r HO r Os i m O ^ O O O O O O O ^ O O O O O O O O O ^ - i i OrHrMm f t — — — — — — — —I rH rH rH —I —I —I —I —I —H —I —I —I —I —I —I —I — I —I rH —I —I —I «—I —I —I —H «—I —I —I —I —I —I —H —I «—H —I —I —I —I —I «—li—I «—I —I rH — oocococooooooocooooocococococooooooocooocococooococoooooco O O i i O rHrHrHi li li li li li li I i i i i i t i i i i ( i i i i i i i i i i i i i i i i i i i i i o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o (N I rMfN (N f \ r M f N f N f N f N f\ (N f \ f\ f N f N f\ fNtN (\ (N (N f N f N J fMfNfNfN (\ Figure 6-2: Total potential solar radiation based on SFD at each site as per Table 6. Note that the HOBO dual logger at site 200813104 failed. SFD were not able to be determined for this site. 6.1.3 UDARDEM End of winter snow depths for May 2020 and May 2021 show similar spatial patterns of accumulation ( Figure 6 - 3 ). Generally, sites receive similar amounts of snow each year. _ 5- 2021 2020 4 E = 3- + % ~o !2< 55 i HI - 0 O r H r s j m 's f- L n o i-H f N m ^ O L n t D O i-i r s i m O O O O O O ^ O L n t D i^ O O O o o o i-i f N j m O D ^ LOn tOD i-i — — — — — — —— — — — — — — — — — — — — — — — — — — — — — — .S O i i o O i i o O i i o O i i o O i i o O i i o O i i O i i O i i O i i i i i i i i i i i i i i i i i i i i i i i i O i i O i i O i i O i i O i i i i i i i i " i —I i—I i—I i—I i—I i—I rH i—I i—I i—I i—I i—I i—I i—I i—I i—I i—I i—I i—I i—I i—I i—I i—I i—I i—I rH i—I i—I rH oooococococooococococococooocooocooocococooooooocococoooco i o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o (M (N f N (N (\ (N (M (N (N f N f N (N t N f M f N f N (N f N f N (N f N (N (N f N (N f N (N f M f N Figure 6-3: Median LiDAR snow depths in a 5 m radius around each site for 29 May 2020 and 2 May 2021. Note that sites 200810102- 200810105 were outside the extents of the LiDAR DEM for both years, however, the ground temperature profiles for these sites indicate little to no snow accumulation. 56 6.2 Collected Data This section presents and discusses all data and analyses that were collected on site. Data logger information for all transects is presented in Appendix B. Meteorological data from Met 1 ( Figure 6-4) show a mean annual air temperature of -0.8°C, mean annual wind speed of 5.2 m s 1, and total liquid precipitation of 217 mm, summarized in Table 7. " The precipitation measured is likely an undercatch due to wind and the limitations of the ClimaVUE meteorological station. Mean annual ground surface temperatures and mean annual air temperatures have both been used as indices to predict permafrost occurrence ( Etzelmuller et al., 2001; Staub et al., 2017 ) . Annual air temperatures below 0.0°C have been recorded both at Met 1 and the BC Hydro AWS. Subfreezing mean annual ground temperatures and subfreezing mean annual ground surface temperatures were also recorded throughout the study area (Table 8). Ground ice and permafrost are not only the product of climate but can be influenced by other biophysical factors and result from long term energy and mass exchange between the atmosphere and the ground surface ( Shur and Jorgenson, 2007; Gruber et al., 2017 ). That being said, these subfreezing values for mean annual air temperature and mean annual ground temperature likely indicate the presence of alpine permafrost in Conrad Basin, although additional study is needed to confirm this. HOBO dual temperature loggers placed on the glacier surface were both found with water inside. Data were recovered from both of them, however, both loggers were found lying on top of the ice. As it is uncertain as to when they melted out of the ice, it is assumed that all temperature measurements represent the ice surface, and gradients were not calculated. 57 20 - u a ) Air temperature 0- -2 0 - 7 1.0 E 1 0.5 o.o 4 c ) Snow depth 100 c u 50 jJ\ o -1 Uwju J L Jl I 1 x d ) Wind speed — i 20 - i U7 E 10 0 -1 3E E e) Precipitation 21- I 0 -1 . i - ill Jl jli l 780 1 £ _c 760 740 - f ) Atmospheric pressure 2020 - 09 2020 -11 2021- 01 2021 - 03 2021- 05 2021- 07 Figure 6-4: Met 1 summary, (a ) Air temperature, ( b ) incoming solar radiation, ( c ) snow depth, ( d ) wind speed, ( e ) precipitation, and (f ) atmospheric pressure 15 minute data from 9 August 2020 - 6 August 2020. 58 Table 7: Met 1 summary. Air values based on 15- minute increment instantaneous measurements and wind values based on 15- minute average measurements recorded at Met 1. Solar Average daily solar flux ( W nr 2) 152.4 Air temperature Minimum recorded temp (°C ) Maximum recorded temp (°C) Mean annual temp (°C ) Median annual temp ( ° C) - 29.9 25.4 -0.8 - 2.4 Wind Maximum wind speed ( m s 1) Minimum wind speed ( m s-1) _ Mean annual wind speed ( m s 1) Median annual wind speed ( m s 1) 24.5 0.1 5.2 4.7 Precipitation Total liquid annual precipitation ( mm ) 217 ' ' 59 Table 8: Annual ground (approximately 15 cm below ground) and interface ( surface) temperatures for each site Mean and standard deviation values for all sites listed at bottom . Site 200810100 200810101 200810102 200810103 200810104 200810105 200811100 200811101 200811102 200811105 200811106 200813100 200813101 200813102 200813103 200813104 200813105 200813106 200813107 200813108 200814100 200814101 200814102 200814103 200814104 200814105 200814106 Met 1 Mean Std. dev. . Mean annual ground temp ( °C) Median annual ground temp (°C ) 1.4 3.6 1.3 - 1.8 2.1 0.5 2.7 4.1 1.9 2.9 0.9 3.4 3.5 1.3 1.8 No Data 1.9 3.2 - 0.1 -1.3 2.1 3.6 2.8 1.1 3.2 2.5 3.7 1.6 -0.5 2.0 1.5 Mean annual interface temp (°C) 2.7 Median annual interface temp (°C) -1.2 0.3 4.3 0.2 -1.2 -1.1 -4.0 -3.2 - 0.4 -3.3 0.1 -1.0 -0.4 0.4 -1.6 -4.8 - 2.0 - -0.7 -0.1 -0.8 - 0.1 0.1 -1.5 -0.4 No Data -0.7 0.1 - 2.0 - 2.7 0.0 0.0 0.0 -0.9 0.2 -1.3 -0.2 -0.6 -0.6 0.9 1.7 3.3 4.2 1.8 3.5 0.4 3.7 3.8 1.4 2.0 3.9 1.9 3.6 - 0.2 - 4.0 1.8 3.6 3.1 1.9 3.2 2.2 3.8 1.8 2.0 2.0 0.8 0.0 -1.1 - 0.2 -1.5 - 0.1 0.0 - -1.8 -1.8 0.2 - 2.0 0.1 - 2.5 - 6.9 0.1 0.1 0.0 -1.5 0.1 -1.5 - 0.2 -1.3 -1.3 1.7 Ground temperature data collected in this study may contribute to our understanding of alpine permafrost extents and the hydrological influences that permafrost has on hillslope and ground water processes. As permafrost has been shown to impede water infiltration and flow through the ground, the presence of discontinuous permafrost, indicated by these observations from Conrad Basin, may have an impact on the ground water recharge and 60 landform evolution (Sjoberg et al., 2021). Ground water infiltration may be reduced by the presence of permafrost thus erosional landform development may be altered. This is a field that warrants additional study. 6.2.1 Ground Thermal Regime This section focuses on the relationships between ground temperatures and topographic and climatic variables including elevation, snow depth, total potential solar radiation, ruggedness, slope, and depth of ground temperature sensor ( sensor depth ). Slope aspect is not included as it is not a linear relationship and is accounted for in total potential solar radiation. Relations during each period of interest are explored through statistical analyses such as linear regression, ordinary least squares multiple linear regressions, and lag analyses. 6.2 .1.1 Winter The Winter period is defined here as the entire snow season from 14 October 2020 to 10 June 2021 at Conrad Basin. Examining this period gives a good understanding about what variables control the overall ground thermal regime while snow cover is present. Snow depth, potential solar radiation, and snow cover days are the three variables that correlate the strongest with ground temperatures ( Figure 6-5 ). It is important to reiterate that snow depth measurements were taken in May of 2021, and the temporal evolution of the snowpack is an important, but unknown, consideration. Early or late snow onset, with depth to sufficiently insulate from atmospheric conditions, will affect the ground thermal regime profile. Despite the "shallow" site measuring 1.4 m in snow depth in May, the ground temperature profile is below 0°C throughout much of the winter period ( Figure 5 - 2 ). The disconnect between the ground thermal regime and the measured snow depths in May will possibly skew relations between the ground thermal regime and snow depths. Snow depth versus ground temperature shows the strongest correlation with a R2 value of 0.78 and a p-value of 0.000. The next strongest correlation is total potential solar radiation with a R2 value of 0.65 and p-value of 0.000 followed by snow cover days with a R2 value of 0.54 and a p-value of 0.000. The other variables show little to no correlation or significance 61 for the Winter period. Section 6.2.1.3 discusses the individual periods of interest that make up the overall Winter period. 62 S 5 o- rp E OJ -5 - a 8 *- - 10 a ¥ 5 2300 2400 Elevation ( m ) 2500 0 d) 0- 1 4 3 2 Snow depth ( m ) •••• % •• rp (U -5 - % 8 -10 - R 2 = 0.00 E) Q •• o o 600 k 800 k 1M 1.2 M 1.4 M Total potential IR ( W h rrT 2 ) 1.6 M e) S rp E) 0 f) ••i 0- 0.5 p- value = 0.845 1.0 1.5 Ruggedness ( m ) : % •i . 9 •. * •• -5 - D " -8 O O) 5 2 - 10 - R = 0.00 10 p- value = 0.908 20 Slope ( ° ) 30 R 2 = 0.02 10 p- value = 0.491 12 14 Sensor depth ( cm ) 16 0- rp E; a -5 - o 8>- - 10 - to 0 50 100 150 200 Snow cover days 250 Figure 6-5 : Linear regressions between ground temperatures and (a ) elevation, ( b ) snow depth, ( c ) total potential solar radiation, ( d ) ruggedness, ( e ) slope, (f) sensor depth, and (g) snow cover days for all non-ice sites during the Winter period of interest. 63 6.2 .1.2 Recently deglaciated terrain Transect 3 consists of short elevational steps between sites to evaluate the ground thermal regime in recently deglaciated terrain at a high elevational resolution. This terrain has likely become ice free within the last 100 years in the upper extents to ice free within the last year in the lower extents. Simple individual linear correlations of median ground temperatures during the winter period and the predictor variables ( Figure 6- 6 ) revealed similar relations between Transect 3 and all sites as shown in Figure 6-5. Snow cover days, total potential solar radiation, and snow depth show the greatest relationship with median ground temperatures for the winter period with R2 values of 0.90, 0.79, and 0.69 and p-values of 0.001, 0.008, and 0.021, respectively. These are consistent with the pattern of results from all other sites. Interestingly, there is a very strong relationship to snow cover days for this transect. This may be in part due to a small sample size and coincidental results. However, the overall pattern of which variables are significant remains the same compared to all other sites for the winter period. Distance to edge of glacial ice, unique to this transect, revealed no correlation with ground temperatures over the winter period. 64 u 2 5 0- a) •• • £ Q_ E aj -5 - D “ 2 o -10 - R = 0.00 ^ 2280 p- value = 0.898 2300 2320 Elevation ( m) R 2 = 0.69 2340 p- value = 0.021 1 3 2 4 Snow depth (m) u 2 5 d) 0- . •• 2 Q_ E v a -5 - 2 o -10 - R = 0.79 u 600 k p- value = 0.008 800 k 1M 1.2 M 1.4 M Total potential IR ( W h m -2 ) R 2 = 0.00 0 20 p- value = 0.948 40 60 80 Distance to ice ( m ) 100 u 2 5 e) 0 - •• f) 3 t 2 _ Q E Si -5 - O 3 2 o -10 - R = 0.06 (J p- value = 0.594 1.0 0.5 R 2 =0.10 10 1.5 p- value = 0.498 15 Ruggedness (m) 20 Slope ( ° ) 25 u £ 0- 9) i 2 _ Q E 2 -5 - a 2 o -io - R = 0.18 5 13.0 p- value = 0.340 14.0 13.5 14.5 Sensor depth ( cm ) 15.0 R 2 = 0.90 125 p-value = 0.001 150 175 200 225 Snow cover days 250 Figure 6-6: Linear regressions between ground temperatures and (a ) elevation, ( b ) snow depth, ( c ) total potential solar radiation, (d) distance to ice, (e ) ruggedness, ( f) slope, ( g) sensor depth, and ( h ) snow cover days for Transect 3 sites during the Winter period of interest. 65 6.2 .1.3 Periods of interest The remaining periods of interest individual regressions are compared in Figure 6- 7. Each sub-figure shows the individual regressions for the November, February, April, and WET periods of interest for each variable. The resulting coefficient of determination and p-values are shown for each period of interest. Snow depth, solar radiation, and snow cover days are the three variables that have the most influence over the ground thermal regime among the variables tested. Snow depths show the strongest predictive power for the periods of November and WET with R2 equal to 0.67 and 0.75, respectively. The p-values for these periods are both 0.000. Total potential solar radiation ( annual ) shows a strong inverse relation to February temperatures with a R2 value of 0.83 and a p-value of 0.000. This is counter to the original hypothesis that states sites that receive greater solar radiation are expected to have increased ground temperatures. Snow cover days show a strong correlation to ground temperature in November and February with R2 values of 0.59 and 0.72, respectively and p-values of 0.000 for both. Similar results were found in the Winter period individual regressions. I surmise that as the variable snow free days, derived from snow cover days, was utilized in calculating total potential solar radiation, the relation between total potential solar radiation and ground temperatures is actually picking up the relation between snow free days and ground temperatures instead of the annual quantity of solar radiation and ground temperatures. These snow free days likely allow the ground to be cooled by atmospheric conditions in these high alpine environments . The measurement depth for ground temperature loggers (sensor depth ) shows the greatest correlation and significance for the April period among the variables tested with a R2 value of 0.53 and a p-value of 0.000. Sensor depth was not significant in the Winter period regressions. The relation between the depth of measurement and ground temperature may be a result of long-term energy storage in the rock. All logger depths in this study were relatively shallow and varied only slightly so these results may be picking up the near -surface temperature regime of a vertical ground temperature profile. Additional study with deeper boreholes and 66 more temperature measurements at various depths at some of these sites would be necessary to confirm. 67 u OJ 3 2 0- -6 - . Q I-1 2 i -i s - ' D 2 0 2400 2500 Elevation ( m ) 2300 1 2 3 Snow depth ( m) 4 u £ o- £ -6 - 3 c) > d) r ; •• • Q. |~1 2 ' rvalue = i§ -i8 - 6R. = <5 J 000 2 0.000 0.030 o.ooo 0.61 0.83 0.20 0.46 p- value = 0.788 R2 = 0.00 600 k 800 k 1 M 1.2 M 1.4 M 1.6 M Total potential IR ( W h m -2 ) * 0.732 0.127 0.879 0.01 0.11 0.00 0.5 1.0 1.5 Ruggedness ( m) u £ 0- 2 -6 - e) ~ r i Q. •• I p0.803 value• • i -i8 ft = 0.737 0.139 0.923 -1 2 ‘ - - 2 0.00 2 = 13 0.01 10 0.10 0.00 30 20 Slope ( ° ) 10 12 14 16 Sensor depth ( cm) u Q) B 2 0- Si Nov Feb Apr WET -6 - Q. I-1 2 ' Py 1 -18 - R2 ^=0 g 0.59 6 0 0.000 0.019 0.008 0.72 0.23 0.29 100 150 200 50 Snow cover days 250 Figure 6-7: Comparison between November, February, April, and WET periods of interest individual regressions, ( a ) Elevation, ( b ) snow depth, (c) total potential solar radiation, ( d ) ruggedness, ( e) slope, and (f) ground temperature depth ( sensor depth ) are examined. 68 6.2.1.4 OLS multiple linear regression Multiple linear regressions were performed to determine which variables were the most significant when compared to ground temperature observations ( Table 9 ). Solar radiation was removed as it is highly correlated with snow cover days and due to the inverted relations with ground temperatures. The results are consistent with what was found with individual regressions. Snow cover days and snow depth are the two primary factors likely controlling the ground thermal regime. Sensor depth also appears significant for the April period. Table 9: OLS multiple linear regression results for ground temperatures at each period of interest , P-values < 0.05 highlighted in light blue. Period: No. Observations: Dependent variable: R-squared: Adj. Rsquared: Independent variables Elevation Snow depth Aspect Slope Ruggedness Sensor depth Snow cover days Annual ( med ) Annual ( mean) Winter November February April WET 23 23 23 23 23 23 23 med T meanT med T med T medT med T medT 0.693 0.731 0.874 0.839 0.851 0.738 0.799 0.550 0.606 0.815 0.764 0.782 0.616 0.565 0.585 0.020 0.791 0.275 0.229 0.966 0.005 0.839 0.134 0.326 0.162 0.124 0.006 0.015 0.182 0.008 0.977 0.153 0.157 0.420 0.337 P>|t| 0.277 0.135 0.225 0.415 0.330 0.123 0.223 0.076 0.401 0.899 0.146 0.097 0.064 0.032 0.246 0.002 0.774 0.206 0.170 0.256 0.034 0.474 0.024 0.306 0.159 0.158 0.188 0.009 Observations of Figure 6- 7 and Figure 6-5 indicate the following: Ground temperatures generally increase as snow depth and snow cover days increase, indicating the insulating properties of snow. Interestingly, ground temperatures generally decrease with an increase in total potential solar radiation. Solar radiation and snow cover days show very similar, albeit inverted, trends. The more solar radiation a site receives does not translate into warmer 69 ground temperatures but actually results in the opposite effect. From this I infer that the sites that receive more solar radiation are not being heated enough by the additional solar load to increase the ground temperatures but are actually exposed for greater duration to the atmosphere due to shorter snow cover duration. This allows for greater cooling to occur and hence the negative slope in this relation. Snow cover timing plays a vital role in the ground thermal regime. Sites with snow cover late in winter that are subject to cold atmospheric temperatures are likely to have subfreezing ground temperatures for much of the winter season despite how much snow accumulates. This is why snow cover days are important. The days with snow cover, or conversely, the days without snow cover, are important at each site because this affects how atmospheric temperatures are able to directly influence the ground thermal regime. In this regard, the ground thermal regime is not primarily controlled by large seasonal trends, i.e., how much heat storage occurs in summer, but controlled more greatly by atmospheric conditions directly before sufficient snow cover insulates the ground from changes in atmospheric temperatures or solar radiation. Future studies may want to look at longer term ground temperature measurements to properly assess the influence that seasonal trends have on the ground thermal regime over a multi year period. 6.2 .1.5 Gradients To calculate temperature gradients between the ground and the base of the snowpack, sites with broken dowel, no snow cover were excluded, as were glacier sites. As a result, only ten sites remained for gradient calculations. This small sample size was not anticipated when designing this study, and it limits the implications of the gradient analyses. Depth of ground temperature thermistor ( sensor depth ) as a variable was not included in these analyses as it was used directly in gradient calculations and was thought to lead to confounding results if it was included. 70 Gradient profiles reveal certain characteristics unique to each typical regime ( Figure 6-8 ). The time series of deep snowpack sites shows a well - defined logarithmic decay in temperature gradient. Profiles for shallow snowpack sites and intermittent to no snow sites are much less defined. In general, more extreme temperature gradients were observed at sites with less snow. Deep snow sites show consistent heat transfer from the ground to the overlying snowpack whereas sites with less snow to no snow show periods of heat transfer from the snow, or atmosphere, into the ground as values become negative. This has important implications for the ground thermal regime as well as midwinter basal snow melt. Midwinter basal snow melt, driven by positive heat flux from the ground into the snowpack will occur in deep snow sites where ground temperatures remain positive all winter. It is also important to note that while Figure 6-8 shows snow cover days, these were determined from ground temperatures. Snow cover may not cover the thermistors mounted on the dowel substantially enough to insulate from atmospheric conditions, hence the extreme variation observed in snow covered periods of Figure 6-8b. In this study, shallow snow sites show more extreme ranges in temperature gradients than deep snow sites ( Figure 6-8), which indicates greater heat transfer between the ground and snow at these shallow sites. However, this does not necessarily indicate that there is greater basal melt occurring at these sites. Figure 5- 2b shows negative values for both the ground and snowpack temperatures. While this study did not directly measure basal melt, mid winter ground water recharge will likely occur at a much lower rate in shallow snow, cold sites, than deep snow, warm ground sites. Intermittent snow sites show the greatest values and variation in thermal gradients ( Figure 6-8 c ). This is likely due to the absence of a thermally moderating snowpack. Without this thermal barrier, the ground thermal regime is directly affected by atmospheric conditions leading to large, fluctuating gradients. 71 30 - b) Shallow snowpack - 1.4m i E u 20 - 0) . L oi 10 - OJ 3 03 -10 - 200811100 2020- 11 2021-01 2021- 03 2021-05 Ground temperature gradient Snow cover Figure 6-8: Hourly thermal gradient regimes for typical sites: ( a ) Deep, ( b ) shallow, ( c) no/intermittent snowpacks are shown. Calculated snow cover days are shown as gray shading in background. Note that the y-axis scales are not consistent over the three figures. Positive values indicate heat transfer from ground to snow. 72 Each subfigure in Figure 6-9 illustrates regressions between the median gradient versus elevation, snow depth, total potential solar radiation, aspect, ruggedness, and slope. Squared Pearson Correlation Coefficients and p-values are shown for each period of interest. Snow depth shows the greatest relationship with median temperature gradients for the November period with an R2 value of 0.78 and a p-value of 0.001. Solar radiation also shows significant results albeit with a lesser R2 of 0.52 and p-value of 0.023. Solar radiation shows the greatest correlation with median temperature gradients for the April period with an R2 of 0.67 and a p-value of 0.004. Snow depth also shows significant results with a R2 value of 0.54 and pvalue of 0.016. Snow cover shows R2 values of 0.40 and 0.43 with p-values of 0.048 and 0.040 for the November and April periods, respectively. All other time periods and variables show insignificant results. As in section 6.2.1.3, snow depth and snow cover days are likely two variables controlling ground-snow temperature gradients with the number of snow cover days directly influencing the amount of cooling possible on the ground temperatures. Again, total potential solar radiation and snow cover days show inverted trends suggesting that they are connected. As stated in section 6.2.1.4, this is likely due to the amount of time the ground is exposed to the atmosphere. Heating from solar radiation is insufficient to counteract cooling driven by atmospheric conditions. This seems to be the case for early and late season temperature gradients. Controlling factors on gradients mid winter are not well defined in this study. This may be due to a small number of observations or perhaps by overlooking a controlling variable. Additional study is warranted. 73 rH i E u . 0) TO fD a) p- value = is - 0.639 0.189 10 - R2 = 0.03 0.21 0.461 0.104 0.07 0.30 e s - •• •• ro (U 9E |t| 0.508 0.753 0.488 75 6.2 .1.6 Snow redistribution via wind Met 1 recorded maximum wind speeds of 24.5 m s 1, a mean wind speed of 5.2 m s 1 and a ' mean wind direction of 247° ( WSW), calculated as per Stull ( 2017 ). Threshold values for dry _ snow redistribution via wind range from 4 to 11 m s 1 ( Li and Pomeroy, 1997). These are well within the measurements taken at Met 1. This is also consistent with the variable snow depth measurements at Met 1. Wind is one of the primary factors affecting snow distribution within Conrad Basin, as snow distribution patterns are similar year to year ( Figure 6-10). Figure 6-10: Snow distribution in Conrad Basin for May 2020 (left image ) and May 2021 (right image ). Red shading indicates snow depth. The reason this may be important is the relation between potential solar radiation and wind in Conrad Basin. The predominant wind direction throughout the study period was west south west. Snow is generally scoured on windward southwestern slopes and deposited on the leeward northeastern slopes ( Figure 6-10). Snow depths versus aspect ( Figure 6-lla ) shows similar results, calculated from snow depth and aspect rasters, although with a more 76 pronounced west -east trend. However, there may be sampling bias as northwest to northeast slopes were not sampled as frequently as other aspects due to constraints in DEM and ice margin limits on the area of interest. Area surrounding Transects 1, 2, and 3 were covered from ice margin to DEM limit. The cliff band east of Transect 1 was the easternmost limit ( Figure 6-12). Southwestern slopes will, however, receive the most solar radiation. The wind is transporting snow off the slopes that receive the most solar radiation also making them available for atmospheric cooling. While no relation necessarily exists within the data, I surmise that within this basin on a large scale overall, there is an amplification in the relation between solar radiation and ground temperatures because of the direction of the prevailing wind and given that the likelihood of wind driven snow redistribution is high. The aspects that receive the most direct solar radiation are also those that face the windward direction with wind speeds in the basin more than capable of redistributing snow. This will also lead to greater cooling driven by colder atmospheric conditions and shallower snowpacks on windward sides. Figure 6-11: Polar plots of ( a ) maximum winter snow depth (m ) versus aspect and (b ) median hourly ground temperature ( °C) over the Winter period versus aspect. Figure 6-12: Area of interest for snow depth versus aspect raster analysis. Magenta dashed line indicates boundary. Snow depth map overlaid as red shading. 6.2 .1.7 Lag Analyses Met 1 had sporadic snow accumulation over the study period. Early snow accumulations reduced the variability of ground temperature profiles in November ( Figure 6 -13 a and Figure 6-14a ) . Correlation coefficients were calculated between observed temperature gradients and ( a ) air temperature and ( b ) solar radiation lagged at one- hour intervals (+48 and -48 hours) for each period of interest. The maximum Pearson Correlation Coefficient and corresponding lag is noted in each plot. The greatest absolute lag between ground temperature and air temperature or solar radiation was observed in periods with greater snow depth. Negative lag indicates ground 78 temperature response subsequent to changes in air temperature. Correlation coefficients between solar radiation and ground temperatures are very low during snow - covered periods ( Figure 6-14b and d). This is likely due to snow cover acting as a barrier and reflector that prevents solar energy from reaching the ground surface. The relationship between air temperature and ground temperature during snow covered periods is much stronger ( Figure 6-13 b and d). The lag between air temperature and ground temperature is the shortest in the July period with a lag of -1 hour. This is also the period with the shortest lag between solar radiation and ground temperatures where lag is -4 hours. These results reaffirm the importance that snow plays as an insulator and as a barrier for the ground thermal regime. Air temperatures show a stronger relationship with ground temperature than solar radiation, however, there is still a strong relationship with solar radiation and ground temperatures in snow free periods. 79 G 20 - 1.0 a ) Nov b) 0.5 QJ 3 ra 0 0.0 cu E ^ -0.5 -20 2020 -11-18 2020-11- 24 2020 -11- 30 C max r = 0.73 -1.0 20 - -100 1.0 c ) Feb at Lag = -10 T i -50 0 T 50 100 d) 0.5 QJ a 03 0 0.0 cu E -0.5 £ - 20 - - 1.0 2021-02 -05 U 20 - 2021-02 -11 max r = 0.93 atl ag = -4 T i -100 -50 0 50 100 1.0 e ) Apr 0.5 03 3 03 0 0.0 - E -0.5 £ -20 - -1.0 2021- 04-18 2021-04 - 24 2021- 04- 30 maxr = 0.60 -100 atLag = -3 0 - 50 50 100 1.0 u 20 0.5 QJ a fO 0 0.0 E -0.5 £ - 20 - g) July - 1.0 2021- 07 -18 2021-07 - 24 2021- 07- 30 MetairT( ° C ) Ground T( ° C ) max r = 0.80 at Lag = -1 T -100 -50 0 50 ( ) Lag hrs r vs Lag 100 Figure 6-13: Lag analysis between hourly ground and air temperatures at Met 1. Hourly ground and air temperatures for ( a ) November, (c ) February, (e ) April, (g) July periods of interest. Corresponding lag analysis with Pearson Correlation Coefficients between ground and air temperatures for ( b ) November, ( d ) February, ( f) April, and ( h ) July at an hourly time scale. 80 u a B rp 20 - 1.0 a ) Nov r f Q . -A A A A A A A A A A A H_A A_A. _ Q |-20 - 2 50 ( 0- 0.0 - -0.5 - S? 2020-11-16 20 - 0.5 - J5 -1 < CL u 5 - 1.0 2020-11- 24 r f AJLLAJLAJ\JL/UUIAAA | .0 I J5 0 -25 0 25 50 d) 0.0 - -0.5 - -l £ 2021-02-03 2 3 rp -50 at Lag = - 28 0.5 - -1.0 20 - max r= 0.15 1.0 c ) Feb |-20 - u b) 2021-02 -11 max r= -0.33 -50 -25 \at Lag= -43 0 25 50 1.0 e) Apr 0.5 r fE 0 1 - 0.0 > Q. ID ^ o -0.5 - - -1 ui — 20 - -1.0 2021-04-16 2021-04- 24 max r= |0.72 -50 -25 at Lag= - 4 0 25 50 1.0 u 20 - 1 fE o o S

( accessed May 31, 2020 ). Bevington, A., 2020, Personal Communication. Bilish, S. P., H. A. McGowan, and J. N. Callow, 2018, Energy balance and snowmelt drivers of a marginal subalpine snowpack: Hydrological Processes, v. 32, no. 26, p. 3837-3851, doi:10.1002/ hyp.13293. Bosson, J.- B., P. Deline, X. Bodin, P . Schoeneich, L. Baron, M. Gardent, and C. Lambiel, 2015, The influence of ground ice distribution on geomorphic dynamics since the Little Ice Age in proglacial areas of two cirque glacier systems: Earth Surface Processes and Landforms, v. 40, no. 5, p. 666-680, doi:10.1002/ esp.3666. Brennan, D. T., D. M. Pearson, P. K. Link, and K. R. Chamberlain, 2020, Neoproterozoic Windermere Supergroup Near Bayhorse, Idaho: Late-Stage Rodinian Rifting Was Deflected West Around the Belt Basin : Tectonics, v. 39, no. 8, doi:https:// doi.org/10.1029/ 2020TC006145. Brown, R . L., and H. D. Gibson, 2006, An argument for channel flow in the southern Canadian Cordillera and comparison with Himalayan tectonics: Geological Society, London, Special Publications, v. 268, no. 1, p. 543-559, doi:10.1144/GSL.SP. 2006.268.01.25. Calonne, N., F. Flin, S. Morin, B. Lesaffre, S. R. du Roscoat, and C. Geindreau, 2011, Numerical and experimental investigations of the effective thermal conductivity of snow: Geophysical Research Letters, v. 38, no. 23, doi:10.1029/ 2011GL049234. Calonne, N., L. Milliancourt, A. Burr, A. Philip, C. L. Martin, F. Flin, and C. Geindreau, 2019, Thermal Conductivity of Snow, Firn, and Porous Ice From 3- D Image- Based Computations: Geophysical Research Letters, v. 46, no. 22, p. 13079-13089, doi:10.1029/ 2019 GL085228. Clarke, J., and D. McClung, 1999, Full - depth avalanche occurrences caused by snow gliding, Coquihalla, British Columbia, Canada : Journal of Glaciology, v. 45, no. 151, p. 539-546, doi:10.3189/ S0022143000001404. Cline, D. W., 1997, Snow surface energy exchanges and snowmelt at a continental, midlatitude Alpine site: Water Resources Research, v. 33, no. 4, p. 689-701, doi:10.1029/97 WR00026. 95 Danby, R. K., and D. S. Hik, 2007, Responses of white spruce ( Picea glauca ) to experimental warming at a subarctic alpine treeline: Global Change Biology, v. 13, no. 2, p. 437451, doi:10.1111/ j.1365-2486.2006.01302.x. DeWalle, D. R., and A. Rango, 2008, Principles of Snow Hydrology: Cambridge, Cambridge University Press, doi:10.1017/ CB09780511535673. Dingman, S. L., 2002, Physical hydrology: Upper Saddle River, N.J, Prentice Hall. Ecological 2020, of Framework Canada, Cordillera Montane Ecozone: ( accessed May 30, 2020). EMUD, RTK GPS receiver for high 2022, Reach RS ( accessed January 5, 2022). precision surveying: EROS, E. R. 0. and S. ( EROS) C., 2018, USGS EROS Archive - Digital Elevation - Shuttle Radar Topography Mission ( SRTM ) 1 Arc -Second Global: United States Geologic Survey. Etzelmuller, B., M. Hoelzle, E. S. Flo Heggem, K. Isaksen, C. Mittaz, D. V. Muhll, R. S. pdegard, W. Haeberli, and J. L. Sollid, 2001, Mapping and modelling the occurrence and distribution of mountain permafrost: Norsk Geografisk Tidsskrift - Norwegian Journal of Geography, v. 55, no. 4, p . 186-194, doi:10.1080/00291950152746513. Evenchick, C. A., M. E. McMechan, V. J. McNicoll, and S. D. Carr, 2007, A synthesis of the Jurassic-Cretaceous tectonic evolution of the central and southeastern Canadian Cordillera : Exploring links across the orogen, in Special Paper 433 : Whence the Mountains ? Inquiries into the Evolution of Orogenic Systems: A Volume in Honor of Raymond Geological Society A. 117-145, of Price: America, p. doi:10.1130/ 2007.2433 (06). Farbrot, H., T. F. Hipp, B. Etzelmuller, K. Isaksen, R. S. 0degard, T. V. Schuler, and 0. Humlum, 2011, Air and Ground Temperature Variations Observed along Elevation and Continentality Gradients in Southern Norway: Air and Ground Temperatures along Gradients in Southern Norway: Permafrost and Periglacial Processes, v. 22, no. 4, p. 343-360, doi:10.1002/ppp.733. Federer, C. A., 1965, Sustained winter streamflow from groundmelt: Research Note NE -41. Upper Darby, PA: U.S. Department of Agriculture, Forest Service, Northeastern Forest Experiment Station. 1-4, v. 41. Fitzpatrick, N., V. Radic, and B. Menounos, 2019, A multi-season investigation of glacier surface roughness lengths through in situ and remote observation: The Cryosphere, v. 13, no. 3, p. 1051-1071, doi:10.5194/tc-13 -1051-2019. 96 Fitzpatrick, N., V. Radic, and B. Menounos, 2017, Surface Energy Balance Closure and Turbulent Flux Parameterization on a Mid-Latitude Mountain Glacier, Purcell in v. p. Earth Canada : 5, Mountains, Frontiers 67, Science, doi:10.3389/ feart.2017.00067. FLNRORD, 2021, BC Water Tool: ( accessed October 14, 2021). Fuchs, S., and M. Brundl, 2005, Damage Potential and Losses Resulting from Snow Avalanches in Settlements of the Canton of Grisons, Switzerland: Natural Flazards, v. 34, no. 1, p. 53-69, doi:10.1007 /sll069 -004-0784-y. Goodrich, L. E ., 1982, The influence of snow cover on the ground thermal regime: Canadian Geotechnical Journal, v. 19, no. 4, p. 421-432, doi:10.1139 /t82-047. Map: of Interactive 2020, BC, Snow Stations ( accessed May 11, 2020). Government Gruber, S., C. R. Burn, S. Harris, S. L. Smith, P. Bonnaventure, L. Arenson, M. Geertsema, and B. Benkert, 2015, Permafrost in mountainous regions of Canada, in GEO Quebec 2015: Challenges from North to South: Unpublished, doi:10.13140/ RG.2.1.4091.4405. Gruber, S., R. Fleiner, E. Guegan, P. Panday, M .-O . Schmid, D. Stumm, P . Wester, Y . Zhang, and L. Zhao, 2017, Review article: Inferring permafrost and permafrost thaw in the mountains of the Hindu Kush Himalaya region: The Cryosphere, v. 11, no. 1, p. 81-99, doi:10.5194/tc-ll-81-2017. Hasler, A., and M. Geertsema, 2013, Provisional permafrost map of British Columbia: Government of British Columbia . Hasler, A., S. Gruber, and W. Haeberli, 2011, Temperature variability and offset in steep alpine rock and ice faces: The Cryosphere, v. 5, no. 4, p. 977-988, doi:10.5194/tc -5-9772011. Hock, R., 2005, Glacier melt: a review of processes and their modelling: Progress in Physical Geography: Earth and Environment, v. 29, no. 3, p. 362-391, doi:10.1191/0309133305 pp453 ra. Hoelzle, M ., 1996, Mapping and modelling of mountain permafrost distribution in the Alps: Norsk Geografisk Tidsskrift - Norwegian Journal of Geography, v. 50, no. 1, p. 11-15, doi:10.1080/00291959608552347. Hoy, T., B. N. Church, A. Legun, K. Glover, G. Gibson, B. Grant, J. O. Wheeler, K. P. E. Dunne, J. Cunningham, and P. J. Desjardins, 1994, Bedrock Geology - Data Catalogue: ( accessed May 31, 2020 ) . 97 Jiskoot, H., and M. S. Mueller, 2012, Glacier fragmentation effects on surface energy balance and runoff : field measurements and distributed modelling: Hydrological Processes, v. 26, no. 12, p. 1861-1875, doi:10.1002/hyp.9288. Jones, H. G., 1999, The ecology of snow-covered systems: a brief overview of nutrient cycling and life in the cold: Hydrological Processes, v. 13, p. 13. Keller, F., and H. Gubler, 1993, Interaction between snow cover and high mountain permafrost at Murtel/Corvatsch, Swiss Alps, in Proceedings of the 6th International Conference on Permafrost, Beijing: South China University of Technology Press, p. 332-337. Lafaysse, M., B. Hingray, P. Etchevers, E. Martin, and C. Obled, 2011, Influence of spatial discretization, underground water storage and glacier melt on a physically-based hydrological model of the Upper Durance River basin: Journal of Hydrology, v. 403, no. 1-2, p. 116-129, doi:10.1016/ j.jhydrol.2011.03.046. Lapo, K . E., L. M. Hinkelman, M. S. Raleigh, and J. D. Lundquist, 2015, Impact of errors in the downwelling irradiances on simulations of snow water equivalent, snow surface temperature, and the snow energy balance: Water Resources Research, v. 51, no. 3, p. 1649-1670, doi:10.1002/ 2014WR016259. Li, L., and J. W. Pomeroy, 1997, Estimates of Threshold Wind Speeds for Snow Transport Using Meteorological Data : Journal of Applied Meteorology and Climatology, v. 36, no. 3, p. 205-213, doi:10.1175/1520-0450(1997) 036<0205 :EOTWSF>2.0.CO;2. Li, X., X. Zheng, L. Wu, K. Zhao, T. Jiang, and L. Gu, 2016, Effects of snow cover on ground thermal regime: A case study in Heilongjiang Province of China : Chinese Geographical Science, v. 26, no. 4, p. 527-538, doi:10.1007/sll769 -016-0825 -y. Lopez-Moreno, J. I., S. R. Fassnacht, J. T. Heath, K. N. Musselman, J. Revuelto, J. Latron, E. Moran-Tejeda, and T. Jonas, 2013, Small scale spatial variability of snow density and depth over complex alpine terrain: Implications for estimating snow water equivalent: Advances in Water Resources, v. 55, p. 40-52, doi:10.1016/ j.advwatres.2012.08.010. Luetschg, M., M. Lehning, and W. Haeberli, 2008, A sensitivity study of factors influencing warm/thin permafrost in the Swiss Alps: Journal of Glaciology, v. 54, no. 187, p. 696704, doi:10.3189 /002214308786570881. Lundquist, J. D., and F. Lott, 2008, Using inexpensive temperature sensors to monitor the duration and heterogeneity of snow-covered areas: Water Resources Research, v. 44, no. 4, doi:10.1029/ 2008WR 007035. Marks, D., and J. Dozier, 1992, Climate and energy exchange at the snow surface in the Alpine Region of the Sierra Nevada: 2. Snow cover energy balance: Water Resources Research, v. 28, no. 11, p. 3043-3054, doi:10.1029/92WR 01483. 98 Menounos, B., 2021, Continuing works at Conrad. Personal communication. Menounos, B., and B. M. Pelto, 2021, Conrad Glacier LiDAR data. Menounos, B., B. M. Pelto, S. W. Fleming, D. Moore, F. Weber, D. Hutchinson, and J. Brahney, 2020, Glaciers in the Canadian Columbia Basin Technical Report : Canadian Columbia Basin Glacier and Snow Research Network. Mickschl, C., 2016, Effects of the ground heat flux on snowpack ablation in a semi-arid mountain climate: Albuquerque, New Mexico, The University of New Mexico, 78 p . Midttdmme, K., and E. Roaldset, 1999, Thermal conductivity of sedimentary rocks: uncertainties in measurement and modelling: Geological Society, London, Special Publications, v. 158, no. 1, p. 45-60, doi:10.1144/GSL.SP.1999.158.01.04. Miller, F. K., E. H. McKee, and R. G. Yates, 1973, Age and Correlation of the Windermere Group in Northeastern Washington: Geological Society of America Bulletin, v. 84, no. 11, p. 3723, doi:10.1130/0016-7606(1973 ) 84<3723 :aacotw>2.0.co;2. Morin, S., F. Domine, L. Arnaud, and G. Picard, 2010, In -situ monitoring of the time evolution of the effective thermal conductivity of snow: Cold Regions Science and Technology, v. 64, no. 2, p. 73-80, doi:10.1016/ j.coldregions. 2010.02.008. Morse, P. D., C. R. Burn, and S. V. Kokelj, 2012, Influence of snow on near-surface ground temperatures in upland and alluvial environments of the outer Mackenzie Delta, Northwest TerritorieslThis article is one of a series of papers published in this CJES Special Issue on the theme of Fundamental and applied research on permafrost in Canada .: Canadian Journal of Earth Sciences, v. 49, no. 8, p. 895-913, doi:10.1139/ e 2012-012. Mukul, Manas, V. Srivastava, S. Jade, and Malay Mukul, 2017, Uncertainties in the Shuttle Radar Topography Mission ( SRTM ) Heights: Insights from the Indian Himalaya and Peninsula : Scientific Reports, v. 7, p. 41672, doi:10.1038/srep 41672. Oke, T. R., 1987, Boundary layer climates: Routledge, 460 p. Payne, J. T., A. W. Wood, A. F. Hamlet, R. N. Palmer, and D. P. Lettenmaier, 2004, Mitigating the effects of climate change on the water resources of the Columbia River basin : Change, no . 62, v. 1-3, 233-256, Climatic Pdoi:10.1023 / B:CLIM.0000013694.18154.d 6. Pelto, B. M., B. Menounos, and S. J. Marshall, 2019, Multi-year evaluation of airborne geodetic surveys to estimate seasonal mass balance, Columbia and Rocky Mountains, Canada : The Cryosphere, v. 13, no. 6, p. 1709-1727, doi:10.5194/tc -13 -1709- 2019. Pomeroy, J . W., and E . Brun, 2001, Physical properties of snow. 99 Poulton, T. P., and P. S. Simony, 1980, Stratigraphy, sedimentology, and regional correlation of the Horsethief Creek Group ( Hadrynian, Late Precambrian ) in the northern Purcell and Selkirk Mountains, British Columbia: Canadian Journal of Earth Sciences, v. 17, no . 12, p. 1708-1724, doi:10.1139/ e80-179. Price, R. A., 2000, The Southern Canadian Rockies: Evolution of a Foreland Thrust and Fold Belt: Geological Association of Canada-Mineralogical Association of Canada, v. Joint Annual Meeting (GeoCanada 2000 ), p. 252. Raleigh, M. S., K. Rittger, C. E. Moore, B. Henn, J. A. Lutz, and J. D. Lundquist, 2013, Groundbased testing of MODIS fractional snow cover in subalpine meadows and forests of the Sierra Nevada : Remote Sensing of Environment, v. 128, p. 44-57, doi:10.1016/ j.rse.2012.09.016. Rasouli, K., J. W. Pomeroy, J. R. Janowicz, S. K. Carey, and T. J. Williams, 2014, Hydrological sensitivity of a northern mountain basin to climate change: Hydrological Processes, v. 28, no. 14, p. 4191-4208, doi:10.1002/hyp.10244. Riley, S. J., S. D. DeGloria, and R. Elliot, 1999, A terrain ruggedness index that quantifies topographic heterogeneity: Intermountain Journal of Sciences, v. 5, no. 1-4, p. 23-27. Robertson, E., 1988, Thermal Properties of Rocks, Open-File Report: United States Department of the Interior Geological Survey, Open -File Report. Rodder, T., and C. Kneisel, 2012, Influence of snow cover and grain size on the ground thermal regime in the discontinuous permafrost zone, Swiss Alps: Geomorphology, v. 175176, p. 176-189, doi:10.1016/j.geomorph.2012.07.008. Sattler, K., B. Anderson, A. Mackintosh, K. Norton, and M. de Roiste, 2016, Estimating Permafrost Distribution in the Maritime Southern Alps, New Zealand, Based on Climatic Conditions at Rock Glacier Sites: Frontiers in Earth Science, v. 4, p. 4, doi:10.3389/ feart.2016.00004. Schon, J., 2015, Physical properties of rocks: fundamentals and principles of petrophysics: Amsterdam, Netherlands, Elsevier . Shur, Y. L., and M. T. Jorgenson, 2007, Patterns of permafrost formation and degradation in relation to climate and ecosystems: Permafrost and Periglacial Processes, v. 18, no. 1, p. 7-19, doi:10.1002/ppp.582. Simony, P. S., and S. D. Carr, 2011, Cretaceous to Eocene evolution of the southeastern Canadian Cordillera: Continuity of Rocky Mountain thrust systems with zones of "insequence" mid-crustal flow : Journal of Structural Geology, v. 33, no. 9, p. 1417-1434, doi:10.1016/j.jsg.2011.06.001. 100 Sjoberg, Y., A. Jan, S. L. Painter, E. T. Coon, M. P. Carey, J. A. O'Donnell, and J. C. Koch, 2021, Permafrost Promotes Shallow Groundwater Flow and Warmer Headwater Streams: Water Resources Research, v. 57, no. 2, doi:10.1029/ 2020WR027463. Smith, R., 2011, Space-time dynamics of runoff generation in a snowmelt- dominated montane catchment: University of British Columbia . Stamen Design, 2022, Stamen Terrain, Terrain: worldwide, OpenStreetMap. Staub, B., and R. Delaloye, 2017, Using Near -Surface Ground Temperature Data to Derive Snow Insulation and Melt Indices for Mountain Permafrost Applications: Permafrost and Periglacial Processes, v. 28, no. 1, p. 237-248, doi:10.1002 /ppp.l890. Staub, B., A. Hasler, J. Noetzli, and R. Delaloye, 2017, Gap-Filling Algorithm for Ground Surface Temperature Data Measured in Permafrost and Periglacial Environments: Permafrost and Periglacial Processes, v. 28, no. 1, p. 275-285, doi:10.1002 /ppp.l913. Stocker -Mittaz, C., M. Hoelzle, and W. Haeberli, 2002, Modelling alpine permafrost distribution based on energy-balance data : a first step : Permafrost and Periglacial Processes, v. 13, no. 4, p. 271-282, doi:10.1002/ppp.426. Stull, R., 2017, Practical Meteorology: An Algebra -based Survey of Atmospheric Science: Vancouver, B . C., University of British Columbia, 940 p. Sturm, M., M. A. Goldstein, and C. Parr, 2017, Water and life from snow: A trillion dollar science question: Water Resources Research, v. 53, no. 5, p. 3534-3544, doi:10.1002/ 2017 WR020840. Sturm, M., J. Holmgren, M. Konig, and K. Morris, 1997, The thermal conductivity of seasonal v. Glaciology, 43, of no. 26-41, p. 143, snow: Journal doi:10.3189/ S0022143000002781. Tombe, S., 2021, Pers Comm: Rock Sample Identification. Wang, T., A. Hamann, D. Spittlehouse, and C. Carroll, 2016, Locally Downscaled and Spatially Customizable Climate Data for Historical and Future Periods for North America : Public Library of Science ( PLoS), e0156720 p., doi:10.1371/journal.pone.0156720. Zhang, T., 2005, Influence of the seasonal snow cover on the ground thermal regime: An Geophysics, 43, of no. Reviews RG 4002, overview : p. v. 4, doi:10.1029/ 2004RG000157. 101 APPENDICES Appendix A - Thermal Analysis Labs Thermal Conductivity Evaluation Report MTPS TAL THERMAL ANALYSIS LABS * 494 Queen Street Fredericton, N.B., E3B 1B6 Tel: (506 ) 457-1515 Fax: (506 ) 454-7201 www.ThemialAnalysisLabs.com Contact Information Company Address Contact University of Northern BC Meaghan Fielding, BSc Sarah Ackermann, MSc September 3 , 2021 Final 1.0 Prepared by Reviewed by Date Report Type Rev. 3333 University Way Prince George, BC V2N 4Z9, Canada Kevin Ostapowich Test Details and Example Setup Sample(s) Submitted: 4 Sample(s) Tested: 4 Test Method( s): MTPS Accessory(s): N/A Calibration(s): Ceramics Temperature Condition(s): -20°C, 0°C, 20°C Atmospheric Condition(s): Ambient Contact Agent(s): Wakefield Thermal Grease . Figure 1 Example Test Setup Result Highlights 6 5 2 4 I > 1 3 I, II I 1 0 Conrad General 200811-100 200810-102 200813-105 -20"C iO'C i20“C . ' Figure 2 Thermal Conductivity Results I Error bars represent 2x standard deviation. 1 102 TAL THERMAL ANALYSIS UBS 494 Queen Street Fredericton , N.B. , E36 1B6 Tel: (506 ) 457-1515 Fax: (506 ) 454-7201 www.ThermalAnalysisLabs.com * Detailed Results Table 1. Thermal Conductivity Results 2 Sample ID and Reference Name Conrad General (CG) 200811-100 ( 100 ) 200810-102 ( 102 ) 200813-105 ( 105) Sensor Temperature (°C )5 Thermal Conductivity k (W /mK ) -20 -18.5 0 20 -20 0 20 0.7 19.8 -18.5 0.5 19.6 0 0 Set Temperature RSD Thermal Effusivity e RSD (%) ( Ws VK ) (% ) 5.076 5.132 4.993 4 658 4.540 4282 1.0 16 21 1.1 2.6 1.0 3408 3433 3371 3219 3165 3047 0.7 1.1 1.4 0.1 2.856 2.0 2358 1.2 -0.2 2.297 3.5 2072 2.0 ' -1 0.7 1.7 0.7 Procedure Details on the operation of the Trident, the test method, and the analysis of results are provided in Appendix 1. All samples were tested using the Ceramics calibration, and Wakefield 120 thermal grease was used as a contact agent. Samples CG and 100 had multiple useable pieces, and thus samples were interchanged to test for uniformity. No significant difference in thermal conductivity was observed between pieces of sample. Samples 102 and 105 had only 1 useable piece each (due to thickness/flatness requirements), and thus all measurements were taken on the same piece of sample. Sample was placed on sensor and a 500g weight was placed on top of the sample to aid in reducing contact resistance. All testing was performed in a Tenney Thermal Chamber to enable temperature control. : Reported thermal conductivity (k) and associated relative standard deviation (RSD) are the average of 15 measurements taken over 3 test sets, with the exception of -20*C, for which the 15 measurements were taken consecutively 3 Temperature was determined as the average MTPS sensor temperature recorded during each test set. 2 103 494 Queen Street Fredericton , N B , E36 1B6 Tel: ( 506 ) 457-1515 Fax: (506 ) 454-7201 www.ThermalAnalysisLabs.com TAL THERMAL ANALYSIS LABS Discussion 4 samples provided by UNBC were tested using the MTPS (Modified Transient Plane Source ) method. All samples were tested at 0°C, and samples CG and 100 were additionally tested at -20°C and 20°C. At 0°C, sample 105 had the lowest thermal conductivity (2.297 W /mK) and sample CG had the highest thermal conductivity (5.132 W/mK ). For sample 100, thermal conductivity decreased slightly with increasing temperature. For sample CG, thermal conductivity stayed relatively constant with increasing temperature. Thermal effusivity is also reported. Thermal effusivity represents the touch perception' of the sample. Samples with high thermal effusivity values feel cool to the touch (metal), and samples with low thermal effusivity values feel warm to the touch (blanket). Figure A 7 shows the typical thermal effusivity range for textiles, as this is the field that most commonly tests for this property. 3 104 494 Queen Street Fredericton , N.B. , E36 1B6 Tel: (506 ) 457-1515 Fax: (506 ) 454-7201 www.ThermalAnalysisLabs.com TAL THERMAL ANALYSIS UBS * Appendix 1: C -Therm Trident Thermal Conductivity Analyzer Figure A1. C-Therm’s Trident Controller & MTPS Sensor The C-Therm Trident is a modular system that uses different sensor configurations to accommodate a wide range of sample types. When operated with a Modified Transient Plane Source (MTPS) sensor, it is an integrated solution for thermal conductivity and thermal effusivity measurement and operates in compliance with ASTM Standard D798416. The MTPS sensor is a one-sided, interfacial heat reflectance device that applies a constant current heat source to the sample. The interfacial sensor heats the sample by approximately 1-3°C during the testing The sample's thermal properties define the dissipation of the heat, and causes a distinct temperature rise at the sensor interface. The rate of the increase in temperature at the sensor surface is inversely proportional to the ability of the sample to transfer heat. Thermal conductivity and effusivity are measured directly, providing a detailed overview of the thermal nature of the sample material. For solid samples: The sample is simply placed onto the sensor with a 500g weight applied on top of the sample to ensure sufficient contact. For samples weighing more than 150g the weight is not used, unless otherwise specified. i 105 Appendix B - Site information table HOBOjend Chjjigt HOBO dual 20Rug mean 20Rug medi 20Asp mean 20Asp medi 20Slp mean Chl depth 20862559 10 147 20880638 1.00 148 14 21 1.10 2008 K>103 200810104 200810105 Geology UTM X UTM Y 504017 5631233 metapelite/ quartzite 503892 5631353 metapelite 1 503603 5631789 metapelite w/quartz vein 1 503589 5631807 metapelite 1 503670 5631755 metapelite 1 503726 5631626 metapelite 1 200811100 1 503769 200811101 200811102 200811103 1 1 4 504058 504014 503701 200811104 4 503908 200811105 4 200811106 4 504477 504393 200813100 2 2 503562 5630980 metapelite 5630497 glacier 5630342 glacier 5630155 metapelite 5630357 metapelite w/ minor quartzite 5630942 metapelite 20880637 20862553 14.5 14 14.5 503257 5630836 metapelite 20880612 20862576 14 200813102 200813103 200813104 2 2 503053 502599 2 502238 5630661 metapelite 5630609 metapelite 5630588 metapelite w / minor quartzite 20863987 20880635 20880639 20862579 20862568 20862560 14.5 14 15 200813105 2 502151 5630562 metapelite 20880632 20862567 200813106 2 501955 20880634 20862571 200813107 200813108 200814100 200814101 200814102 200814103 2 2 3 3 3 3 501815 501671 20781200 20880627 20863988 20863990 20880607 20862529 20862570 20862527 20862566 20862564 20862530 15.5 9 14.5 15 13 15 200814104 3 503628 20880610 20862558 200814105 200814106 Met Metl 3 3 503581 503519 5630395 metapelite w / minor quartzite 5629809 metapelite 5629564 quartzite 5630738 metapelite w / minor quartzite 5630746 metapelite w / minor quartzite 5630742 metapelite w/ minor quartzite 5630766 quartzite w/ minor pelite 5630789 metapelite w/ minor quartzite 5630788 metapelite w/ minor quartzite 5630787 metapelite w / minor quartzite 5631009 metapelite 20880614 20880611 20880640 20862569 20862557 20862555 Site Transect 200810100 200810101 200810102 200813101 1 503685 503681 503684 503674 503877 5631469 quartzite 5631105 metapelite 20880630 20862577 20880605 20862575 10 20862552 20862572 20862551 14.5 14.5 14.5 14 14.5 20880628 20863992 20880602 20880606 20862554 14.5 10 0.81 20880609 20880603 20880604 20862556 20862565 20862578 13 13.5 10 10 50 0.98 0.85 0.53 0.86 20781197 20862573 100 20863991 20862528 20880615 20862531 20880608 100 1.02 10 140 221 311 121 116 118 21 0.77 146 147 17 0.97 144 144 89 143 142 90 20 16 11 8 I 10 10 10 -50 0.98 0.53 140 221 311 121 21 9 32 28 0.44 0.43 66 46 10 1.84 1.81 264 266 33 10 0.71 0.70 263 257 13 10 0.68 0.49 0.63 128 123 13 10 0.48 102 105 10 10 10 10 0.79 0.25 1.26 0.68 0.22 1.24 105 202 152 112 219 152 15 5 25 14 10 0.48 0.46 130 125 9 14.5 M 10 0.43 1.15 1.57 M 62 92 9 10 10 10 0.44 1.07 1.74 1.55 1.51 1.91 2.07 1.21 1.79 0.82 86 132 119 133 150 60 87 127 118 130 161 14.5 10 0.56 0.51 148 154 21 32 23 27 30 25 12 14.5 14.5 10 10 0.62 0.58 139 128 113 139 11 11 15.8 10 0.51 0.52 0.57 0.49 152 152 9 10 10 1.24 106 200810105 Tot pot sol Conductiv Removed. 20Sno mean 20Sno medi 21Sno niean 21Sno medi 21Sno min 21Sno max 21Sno rang 0.17 0.19 0.14 0.10 0.82 0.22 1.04 1534949 4.5 yes 2.17 2.09 2.69 2.61 1.91 3.51 1.60 689050 2.3 yes 9 2.856 no 1822584 1840498 2.9 yes 32 1307836 2.3 yes 28 1256858 21 2.3 yes 200811100 17 0.51 0.48 1.44 1.45 0.87 2.15 1.28 783794 4.49 yes 200811101 200811102 200811103 1.49 0.12 4.22 1.47 0.09 4.20 2.36 0.76 3.31 2.35 0.73 3.30 1.83 0.42 2.95 - 2.76 1.80 3.72 0.93 2.22 0.77 679358 780678 612629 2.9 no 2.3 yes yes 200811105 20 18 11 8 34 200811106 12 200813100 12 Site 200810100 200810101 200810102 200810103 200810104 200811104 20Slp medi 20 20 - - - - - 4.06 4.06 3.58 3.59 3.18 3.96 0.78 732215 yes 0.55 -0.10 0.18 0.52 1.54 0.07 0.05 1.60 -0.07 3.44 938230 0.05 -0.34 0.42 0.00 -0.35 0.68 3.51 0.76 1.03 2.9 no 5.1 no 5.1 yes 2.9 yes 5.1 yes 2.9 no 2.3 yes 2.297 yes 5.1 yes 2.9 yes 4.5 no 5.1 yes 5.1 yes 5.1 yes 4.5 yes 5 1 yes 5.1 yes 5.1 yes 2.9 yes - -0.11 -0.17 1549559 1025901 200813101 9 2.26 2.30 2.77 2.81 2.09 3.38 1.29 680634 200813102 14 4 -0.06 -0.07 -0.07 0.27 0.30 -0.46 1.44 1531799 0.16 0.17 -0.16 0.98 0.44 0.60 1171457 25 1.48 2.47 2.55 1.67 3.53 1.86 200813105 9 -0.12 1.46 0.14 - 0.17 0.11 -0.28 0.77 200813106 200813107 200813108 9 1.77 - -0.06 0.54 1.71 -0.05 0.16 1.74 -0.06 0.02 1.19 0.51 0.72 2.05 0.55 1.44 1.05 0.86 22 31 1.75 0.05 0.69 1.06 2.16 717864 1555384 827570 200814100 20 4.43 5.04 3.89 4.05 2.04 5.59 3.55 622090 200814101 200814102 200814103 24 28 14 11 9 1.38 3.03 0.08 1.23 1.76 0.75 3.20 2.86 0.20 4 72 2.45 4.44 6.24 2 10 989883 709577 3 00 11 1.19 1.65 1.08 1.62 1.71 2.32 1.60 2.26 0.77 -4.55 2 62 0.64 5.21 1.69 3 05 1.83 3.01 0.03 3 57 1.87 3.01 2.95 2.37 1.08 849741 714460 9 -0.16 -0.15 0.00 -0.02 -0.31 0.31 0.62 1539852 200813103 200813104 200814104 200814105 200814106 Metl -0.09 2.70 0.12 - 3 52 - 1118918 1484717 687335 107 Site Dowel 200810100 good 200810101 cracked but still standing at angle 200810102 good 200810103 good 200810104 good 200810105 good Annual T Mean G Annual T Med G Annual T Med l 10.1 2.7 -1.2 3.6 0.3 -1.2 -3.2 6.1 0.2 -4.0 -3.3 -4.8 1.3 -1.8 200811100 good 2.7 200811101 good 200811102 completely broken 200811103 na 4.1 1.9 0.4 0.7 200811105 good 200811106 good 200813100 good 200813101 good 200813102 good 200813103 good 200813104 completely broken - HOBO Dual failed Annual T Mean l -0.5 -0.1 -1.0 -0.4 200811104 na Annual T SD G 1.4 2.1 0.5 -0.7 -0.2 2.9 0.9 3.4 8.3 4.3 -1.1 -0.4 -1.6 7.9 1.7 6.7 3.3 7.0 8.6 4.2 1.8 9.2 8.8 -0.9 -1.5 -0.1 6.9 3.5 -0.8 9.3 0.4 2.3 0.4 4.0 0.4 Annual T SDJ 13.7 9.2 9.5 12.0 8.7 Winter T Mean G 3.5 - 0.3 -3.8 -6.6 -2.4 -3.9 -1.2 -2.0 -0.8 12.3 0.0 8.8 0.3 -1.1 -0.7 -1.0 -0.2 10.4 -3.3 -1.5 -1.6 -1.2 -4.0 -1.8 -0.1 -4.4 -2.8 -1.5 4.3 4.1 9.4 10.8 11.1 -0.1 9.0 3.7 3.5 0.1 6.8 3.8 0.0 8.9 1.3 1.8 -1.5 -0.4 10.5 8.3 1.4 2.0 3.9 -1.8 -1.8 12.4 11.0 0.2 8.2 200813105 good 1.9 -0.7 1.9 -2.0 200813106 good 200813107 good 200813108 good 200814100 completely broken (thermistor almost pulled from hole) 200814101 good 200814102 good 200814103 completely broken 200814104 cracked but standing on angle 200814105 completely broken 200814106 completely broken good Metl 3.2 0.1 1.3 2.1 0.1 2.0 -2.7 9.3 6.2 10.3 7.8 3.6 0.2 4.0 - 0.1 2.5 6.9 - 11.7 8.5 11.6 9.0 -3.2 -0.1 -5.5 0.0 5.2 1.8 0.1 5.0 3.6 0.0 7.1 3.6 0.1 8.6 5.9 9.1 -0.4 -0.6 -0.4 -3.9 - 3.1 0.0 7.1 1.9 3.2 2.2 -1.5 11.5 2.8 0.0 1.1 -0.9 3.2 2.5 0.2 5.2 -1.3 8.9 7.3 3.8 -0.6 9.8 1.8 3.7 1.6 -0.2 -0.1 9.3 -5.7 0.1 6.8 0.5 -1.5 9.8 -1.3 12.3 -2.8 -0.3 -3.7 -0.2 8.6 108 Site Winter T Med G Winter T SD G Winter T Mean l Winter T Med l Winter T SD l November T Mean November T Med G November T SD G November T Mean. November T Med l 5.7 5.0 6.5 3.2 10.0 6.7 7.1 1.7 7.2 7.3 200810100 - - - 0.2 6.3 6.1 5.1 0.2 -6.1 7.7 5.5 -6.1 0.1 3.7 8.0 4.8 8.4 -5.3 1.6 0.2 1.3 0.0 0.0 3.3 -3 7 -1.4 -1.7 -1.6 -1.9 -1.3 -4.6 -1.7 -0.2 -4.4 -1.6 -6.0 -3.2 200810101 200810102 200810103 200810104 200810105 0.2 5.4 7.6 3.8 -5.2 0.2 5.2 5.5 4.7 4.3 200811100 1.4 0.8 200811101 0.3 200811102 -3.5 - -3.7 -3.8 200811105 200811106 200813100 -1.8 -2.0 -1.5 -5.2 -3.1 200813101 -0.1 200813102 -5.6 0.9 1.0 0.9 6.5 3.9 0.2 6.2 200813103 -4.1 3.9 200813105 4.8 5.3 3.5 200813106 0.4 6.8 4.3 -0.2 -5.6 -9.0 -7.3 200813108 -0.1 -6.9 -6.7 -9.0 200814100 -0.6 0.4 -0.4 200814101 -0.7 2.0 200814103 200814104 200814105 200814106 200811103 200811104 200814102 Metl 0.4 -6.6 8.1 5.7 1.0 1.0 1.0 0.4 0.2 0.4 0.4 0.1 4.4 -4.2 -0.8 -1.5 -1.3 -5.3 -1.6 -4.3 1.0 0.9 1.0 1.0 8.4 5.2 0.2 8.3 - -5.5 -0.8 -1.5 -1.2 -5.5 -1.8 0.0 0.0 -7.1 0.0 1.1 1.4 1.4 0.9 0.1 0.0 0.3 1.1 0.5 0.1 1.4 0.2 -8.6 8.4 7.5 6.8 - 0.2 -9.3 - -8.7 -8.2 -7.1 1.5 0.0 -1.3 -5.2 -0.7 -1.4 -1.6 -6.7 -1.9 -0.1 -7.5 -6.4 0.1 -5.2 -0.7 -1.5 -1.5 -6.9 -2.0 -0.1 6.2 -4.4 0.1 7.5 0.4 6.4 6.6 1.2 7.6 7.9 0.2 0.2 0.0 0.1 -8.2 -7.6 -8.4 -7.9 1.5 3.8 -0.5 0.4 -0.4 -0.2 0.4 -0.7 0.5 -1.3 -0.6 -6.5 - 1.8 0.1 0.1 0.1 0.2 1.4 0.1 -1.5 -0.5 -6.8 0.1 7.6 0.1 -1.8 -0.6 -6.2 1.0 -4.6 -0.2 3.1 0.5 5.7 -0.8 -0.3 -3.6 0.1 -8.6 8.7 -0.4 2.2 -0.6 6.5 0.0 -2.4 -0.3 -5.0 2.5 0.3 6.1 -3.1 -0.4 -3.9 -2.5 -0.4 -5.7 3.2 0.6 8.7 -6.2 -0.4 -5.8 -6.3 -0.3 -6.0 1.3 0.2 1.2 -7.1 -0.5 -7.1 -7.1 -0.3 -7.3 -3.2 0.2 -0.2 -6.1 -4.9 - 0.4 6.4 7.8 5.4 -7.3 -4.5 200813104 200813107 -0.2 - 0.1 5.5 0.2 - -5.1 8.4 0.9 0.2 1.2 0.2 - -7.7 -6.6 0.2 -9.0 -9.0 -0.2 0.1 109 Site November T SD l April T Med G February T Mean G February T Med G February T SD G February T Mean l February T Med l February T SD l April T Mean G 12.4 11.4 4.3 14.1 12.7 6.1 2.5 - - - - 200810100 2.3 200810101 0.0 0.2 0.2 0.0 0.1 0.1 0.0 200810102 2.7 -11.0 -10.8 -15.0 -4.7 -8.8 2.9 -6.1 -6.4 200810103 2.0 200810104 2.7 200810105 1.7 200811100 0.7 200811101 200811102 200811103 0.2 1.6 0.0 200811104 0.0 200811105 200811106 0.5 2.0 200813100 0.6 200813101 0.1 200813102 2.0 200813103 200813104 1.9 200813105 2.4 200813106 -15.4 -4.8 -9.1 -1.9 5.3 0.3 2.3 -16.8 -5.1 -11.1 -1.8 0.1 -2.0 0.3 0.3 0.0 -7.8 -2.1 -2.3 -1.9 -7.1 -2.2 -2.3 -1.9 0.0 3.0 0.1 -9.4 -2.1 -2.2 -2.0 0.1 0.1 5.1 -13.5 -13.1 -4.7 -0.3 11.9 -4.6 -0.3 0.3 3.7 0.1 0.1 1.0 0.8 0.1 -16.1 -5.0 -10.9 -1.9 6.7 -2.6 -2.5 0.4 3.4 1.5 0.2 0.0 2.1 1.5 2.7 -0.4 -0.2 -0.1 -0.1 1.4 0.0 9.1 2.1 0.0 0.1 0.2 0.9 2.4 - 0.2 0.5 2.4 - 0.0 0.9 0.0 0.1 -2.6 -2.6 0.0 0.2 8.3 -0.6 -0.5 -0.8 -0.2 0.6 -0.5 0.3 3.0 -0.8 -0.5 -2.3 -2.0 0.2 4.2 -16.7 -15.9 -4.6 -0.4 0.3 0.0 -4.7 -0.4 -11.4 3.8 -14.0 5.9 -6.1 -5.2 1.4 -7.4 -13.9 -6.6 0.1 2.4 2.2 -7.9 -0.5 -7.0 -0.5 -15.2 -9.3 - April T SD G 0.9 0.0 1.4 0.4 -0.2 -0.1 0.4 0.9 0.0 2.6 0.1 0.1 2.2 0.0 -9.4 -0.6 -8.8 -0.5 3.4 0.9 0.1 2.6 0.1 -0.5 -0.5 0.0 -16.2 -10.7 -0.7 -13.6 -9.0 -0.7 7.5 1.0 0.3 3.2 3.8 0.1 0.7 12.4 0.1 0.8 11.8 - 0.0 0.1 4.6 -2.6 -0.7 -0.1 -0.7 -2.7 -0.7 -0.1 -0.7 0.5 0.0 2.3 1.3 200814100 0.4 -0.8 -15.1 -8.0 -0.8 200814101 200814102 200814103 1.3 0.2 1.9 0.1 0.8 11.8 0.1 0.8 10.7 0.1 5.4 2.4 0.0 0.0 0.0 4.0 0.3 1.7 0.1 0.2 0.1 1.6 200814104 0.1 0.1 0.1 0.0 0.1 0.1 0.0 0.1 0.1 0.0 200814105 200814106 2.1 -2.7 -0.4 -2.6 -0.4 0.2 -2.7 -0.5 0.2 0.0 -2.8 -0.5 0.1 -0.8 -0.2 -0.8 -0.2 -12.3 4.4 -14.3 -12.1 7.4 1.1 0.1 200813107 200813108 Metl 0.3 2.1 - - -11.9 - - - - - 0.0 0.4 0.1 2.8 110 Site April T Mean [ April T MedJ April T SD l WET Med G WET Mean G WET Med l WET Mean l WET SD G WET SD I 200810100 4.7 0.9 10.9 -2.5 -2.6 3.4 -1.0 -3.7 9.9 200810101 0.2 0.2 0.1 0.1 0.1 0.0 0.1 0.1 0.0 -5.1 -6.8 -3.2 -4.8 2.1 3.1 1.8 -5.5 -1.9 -2.7 -6.5 -4.6 -5.2 6.5 7.9 8.7 200810104 200810105 2.6 -0.7 9.4 -4.7 -6.9 -2.9 -4.5 200811100 0.0 0.1 0.4 -1.5 -1.4 0.1 -1.4 -1.5 0.1 200811101 200811102 0.0 0.4 0.0 0.0 0.0 0.2 3.0 0.0 - 0.1 0.0 0.0 3.0 0.0 - 0.2 3.0 - -3.0 0.0 0.1 -2.6 -1.5 -2.6 -1.5 0.0 -2.5 -1.4 -2.5 -1.5 200810102 200810103 200811103 0.0 - -2.4 200811105 -2.5 -0.3 200811106 1.8 200811104 -1.9 -2.4 7.5 1.1 0.0 -2.5 -0.2 0.0 0.6 6.9 0.4 -2.5 -3.3 -2.5 -3.4 1.7 0.1 2.9 -2.5 -3.3 -2.5 0.1 0.1 0.1 -3.7 6.6 200813100 -0.3 0.0 0.9 -3.8 -3.8 0.3 -3.8 -3.8 0.5 200813101 -0.2 -0.2 0.1 -0.3 -0.3 0.0 -0.3 -0.3 0.0 200813102 200813103 200813104 1.7 -0.3 -0.7 -4.1 -3.9 -4.2 -3.9 1.4 0.4 -3.4 -4.3 -4 5 0.1 0.1 8.7 6.8 0.0 0.1 C1 5.2 1.9 0.0 200813105 1.8 -0.1 8.4 -3.9 -3.9 0.7 -4.2 -4.4 2.3 200813106 200813107 -0.5 1.3 -0.4 -0.3 0.1 8.0 -0.5 -4.7 0.0 2.9 -0.6 -4.3 -0.5 -5.6 0.0 7.6 200814100 -0.7 -0.7 0.1 200814101 0.0 0.1 0.2 200814102 200814103 -0.6 -0.5 -0.8 200814106 -0.1 - 0.1 4.8 0.0 0.4 -0.9 -0.5 -0.8 -2.8 200814105 0.2 0.1 0.7 -0.9 -0.6 -0.8 -2.6 200814104 0.6 0.1 -0.5 -4.4 -5.6 -1.0 -0.6 -0.9 -2.9 Metl 2.2 0.3 0.4 200813108 -0.1 0.1 8.0 -5.6 0.9 0 0.0 -0.6 -0.8 -2.9 0.0 -1 0.1 0.1 -1.6 -1.6 -0.4 -3.7 -0.4 -3.6 0.0 1.5 0.0 0.1 0.0 2.4 -4.5 0.1 0.1 -1.6 -1.6 -2.9 -4.3 -0.3 -0.3 0.0 0.1 0.0 4.1 0.1 0.1 0.0 7.4 Ill Appendix C - Hourly temperature profiles for all sites C) 200810102 e) Site 200810104 g) h) : !i :• TC7sj =' T57 ^ Site: 200811101 Site 200811100 ,'s H - 25 - Site : Site : 200811103 200811102 2020 - 10 2021 - 02 2021-06 2020-10 Ground Snow 2021-02 Interface 2021-06 Snow cover 112 50 v l) k) u 25 - 3IQ 11 _ E 0. l 25 - Site: 200811104 CD Q CD 50 U V Site: 200811105 m) .cjsS n) ; II $ CD 3 TO CD ;i Q. E H 50 ID= 200813100 200811106 u OJ Site: - 25 - Site: 71 0) p) 'il • i 25 j:; i 5 CD CL ! JU o i E I CD Site: 25 - Site: 200813101 50 U 200813102 q) r) i CD 25 : .3 il 5 CD CD CL m _: \C 0 E l CD 25 - Site: 200813103 50 u Site: 200813105 s) CD 5 CD CD _ E Q l CD 25 - Site: 200813106 2020-10 2021-02 2021-06 2020-10 Ground Snow 2021-02 Interface 2021- 06 Snow cover 113 u) v) i 200813108 I 200814100 x) : Ik. Site: 200814102 z) ; 200814104 ab ) nr Site: 200814106 2020-10 2021-02 Snow Interface Snow cover