C21C-0550
Sensitivity Analysis of Snow Patterns in Swiss Ski Resorts to Shifts in Temperature, Precipitation and Humidity Under Condition of Climate Change
The value of snow as a resource has considerably increased in Swiss mountain regions, in particular in the context of winter tourism. In the perspective of a warming climate, it is thus important to quantify the potential changes in snow amount and duration that could have large repercussions on the economy of ski resorts. Because of the fine spatial variability of snow, the use of a Surface Energy Balance Model (SEBM) is adequate to simulate local snow cover evolution. A perturbation method has been developed to generate plausible future meteorological input data required for SEBM simulations in order to assess the changes in snow cover patterns. Current and future snow depths have also been simulated within the ski areas themselves. The results show a large decrease of the snow depths and duration, even at high elevation in a warmer climate and emphasize the sensitivity of snow to topographical characteristics of the resorts. The study highlights the fact that not only the altitude of a domain but also its exposure, localization inland and slope gradients need to be taken into account when evaluating current and future snow depths. This method enables a precise assessment of the snow pattern over a small area.
C21C-0551
FASST modeled snow predictions in vegetatated and non-vegetated environments
Numerical experiments of snow accumulation and depletion as well as surface energy fluxes over several
vegetated and non-vegetated sites using FASST (Fast All-season Soil STrength) were conducted. FASST is
a newly developed one-dimensional dynamic state of the ground model. It calculates the ground's moisture
content, ice content, temperature, and freeze/thaw profiles, as well as soil strength and surface ice and snow
accumulation/depletion. Since FASST is relatively new, I wanted to determine its use as a snow model. I
demonstrate that even though FASST is only a single-layer snow model, it does well in both environments,
especially during the accumulation phase.
http://www.crrel.usace.army.mil
C21C-0552 INVITED
Multi-Scale Reconstructions of Snowpack Variability for Key Watersheds in Western North America: Tree-Rings Provide Insights on the Past 500 to 1000 Years
One of the most robust lines of evidence for climate change impacts in the Western US is the decline in
snowpack during the latter half of the 20th century. It is critical to ascertain whether this trend is anomalous
relative to long-term patterns of snowpack dynamics. We are using tree-ring data networks coupled with
NRCS snowcourse based reconstructions of Snow Water Equivalent (SWE) to reconstruct snowpack
variability at multiple watershed scales. Snowcourse records provide the raw data needed to generate
calibration datasets of April 1 SWE. Records span 1930 to present, and were utilized to generate historic
SWE anomalies from the scale of individual USGS level 6 watershed to the entire Upper Colorado River
Basin. Across multiple watershed scales historic SWE records were used in combination with more then 600
existing and recently collected tree-ring chronologies to produce 500 to 1000 year records of April 1 SWE
variability. Initial work targets key high-mountain headwaters for the Upper Colorado, Upper
Yellowstone/Missouri, and Columbia/Saskatchewan Rivers. For the Colorado Plateau region, 9 out of 17 level
6 watersheds achieve quality reconstructions (R2> 0.45) of April 1 SWE, and a reconstruction extending
back to 1181 AD was produced for the entire Upper Colorado River Basin. Preliminary analyses show marked
interannual to multidecadal variability in total April 1 SWE. Comparisons with existing proxy records of Pacific
Basin climate show coupled ENSO and PDO influences on the total amount of mountain snowpack in these
regions. The strength and sign of these relationships is shown to vary over time and on a watershed-by-
watershed basis. These and other results exemplify why long-term records are essential baseline information
for evaluating recent and future changes in mountain snowpack. The overarching goal of this project is to lay
the foundation for snowpack reconstructions that encompass high mountain areas in all of western North
America.
http://www.snr.arizona.edu/project/snowpack
C21C-0553
Controls on Snow Accumulation Variation Between a Healthy and a Dead Pine Stand
Forest disturbance (fire, infestation, disease) has a major effect on the forest canopy, with subsequent impacts on snow accumulation and melt. This study examines the impact of mountain pine beetle infestation on forest canopy structure on the Nechako Plateau, British Columbia, Canada, and the resulting impacts on snow accumulation under varying climatic conditions. Field measurements collected in both a healthy and a dead coniferous stand in the 2007 and 2008 winters are compared with those made in a clearcut (canopy- free control) to quantify inter-stand differences in snow depth, density, water equivalent, and structure. Canopy structure in the dead stand differs from that in the healthy stand largely as a function of needle drop and the loss of fine branches and stems following beetle infestation. Snow structure and density between years is a function of micrometeorology which is also driven by canopy structure. In high snow years, the dead stand behaves similarly to the healthy stand due to the ability of large snowfalls to exceed the interception capacity of the canopy. In low to average snow years, however, distinct differences in snow accumulation in dead stands can be attributed to post-disturbance canopy structure.
C21C-0554
Examining the Representativeness of Current Climate Measurement Site Locations in the McKenzie River Basin, Oregon
Hydrologists and resource managers frequently use measurements from the Natural Resources Conservation Service (NRCS) Snowpack Telemetry (SNOTEL) system to forecast streamflows and direct resource planning strategies across the western United States. In the Columbia River Basin, where 60-70 percent of the region's power is generated through large-scale hydroelectric projects and several species of native anadromous fish are listed as endangered, the accuracy of streamflow forecasts directly impacts ecosystem and economic health. Anecdotal reports suggest that forecasts of late spring flows in alpine basins are frequently incorrect, perhaps due to imprecise measurements of basin-wide snow water equivalent (SWE). Here the spatially distributed SnowModel is used to distribute SWE across the McKenzie River watershed, a sub basin of the Columbia. A binary regression tree classification system is then used to explore relationships between spatially distributed SWE and a range of independent physiographic variables. We use this information to identify optimal locations that are representative of basin-wide SWE, at different times of the snow season. We also explore representative locations of SWE for different climate change scenarios The information provides a basis for identifying present-day and future SNOTEL and climate station locations for monitoring water resources.
C21C-0555
Improving Borehole Optical Stratigraphy (BOS): Modeling, laboratory calibration, and design
Borehole Optical Stratigraphy (BOS) uses a video camera with a white light source lowered into a borehole to record variations in returned brightness. The variations have been shown to correspond to annual layers, though the cause of the brightness changes is not well understood. A combination of radiative transfer modeling and ray-tracing was used to model both the reflectivity of the borehole wall and the multiple reflections within the borehole that produce the received brightness signal. Radiative transfer modeling showed reflectivities between 75% and 95% for most polar firn conditions, with small (< cm) scale lateral energy transfer within the borehole wall. These properties allowed us to simulate BOS measurements using a ray-tracing model with the borehole wall represented as a diffusely reflecting surface. Ray-tracing simulations of different layering configurations show that both the interaction of the multiple reflections of the light in the borehole and the reflectance of the firn directly within the camera's field of view are important in determining the BOS signal; for a 5 cm layer, the former effect is about half of the latter. The presence of bumps or scars on the walls can also produce both local and long-range variations in brightness. The current BOS system has two primary limitations: 1) the single light source cannot distinguish between changes in grain size and density as the cause of brightness variations and 2) the system cannot be used on cores because the multiple reflections in the borehole are necessary to produce noticeable variations. To address these limitations, we have developed an instrument which uses two light sources, yellow (590nm) and near-IR (950nm), to estimate the degree to which scattering in the borehole wall transports photons laterally. An estimate of the scattering density is determined from the spreading distance. This allows us to correct the infrared-visible spectral differences giving an estimate of grain size or bubble density at high spatial resolution. We present measurements on ice cores from 50 meters depth at Summit, Greenland, where our measurements show strong annual-scale layering.
C21C-0556
Characterizing Present-day and Future Snow Water Equivalent in the McKenzie River Basin, Oregon
Snow water equivalent (SWE) is typically measured at one or few points within a basin and is not necessarily representative of snow water equivalent within a basin. In this study, we model the spatial distribution of SWE in the McKenzie River Basin in western Oregon. Using SnowModel, we simulate SWE over the course of two snow years, including accumulation and ablation periods. We validate the modeled spatial distributions of SWE using snow cover extent from the MODIS instrument. A binary regression tree model is then used to identify the portions of the basin that are most representative of the basin-wide average SWE at different times of the snow accumulation and ablation seasons. We also modify the temperature inputs to SnowModel to simulate SWE distributions under a warming climate. This approach will be valuable for selecting locations for current and future monitoring of snow for water resources.
C21C-0557
High-Resolution Borehole/Core Techniques for Measuring Snow and Firn Stratigraphy
The characteristics of snow and firn stratigraphy have been shown to change spatially, making wide spatial sampling important, particularly in remote-sensing ground-truth studies. As traditional snow-pit studies of shallow stratigraphy are both time-consuming and limited in depth, faster methods that can penetrate deeper are desirable. We have developed multiple independent methods for measuring stratigraphy in boreholes and on firn cores. Borehole Optical Stratigraphy uses a video camera to record the scattering properties of the borehole wall, which are associated primarily with density and grain size. The Wallingford Neutron Probe (NP) measures a detailed density profile in the borehole by measuring the rate at which neutrons, emitted from the instrument, return to the detector through multiple-scattering. Complimentary measurements on a core can be made with traditional digital photography and Dielectric Profiling (DEP), both of which are frequently included as standard procedure on core processing lines. In this paper, we describe these methods, and compare them side-by-side at several locations, focusing on the accuracy and spatial resolving power of each technique.
C21C-0558
Observed Climate-Snowpack Relationships in California and Their Implications for the Future
A study of the California Sierra snowpack has been conducted using snow station observations and modeled surface temperature data. First of the month snow water equivalent ("SWE") measurements were combined from two datasets to provide sufficient data for statistical analysis of the evolution of the snowpack during the snow season from 1930 to 2007. The temporal centroid of SWE ("SCD") was used to assess variability in the timing of snow accumulation and melt, from February 1st to May 1st. Since 1930, there has been a trend towards earlier SCD by 0.4 days per decade. During the same time period, regional average March temperature, using the NCEP Reanalysis 1 surface temperature dataset, has increased at a rate of 0.4°C per decade. The trend in SCD can be explained by its sensitivity to March temperature. The SCD is shown to shift earlier in the season by 1.3 days per 1°C increase in March temperature.
C21C-0559
Decadal changes of phenological patterns over Arctic tundra biome
The northern high latitudes have experienced a continuous and accelerated trend of warming during the past 30 years, with most recent decade ranks the warmest years since 1850. Warmer springs are especially evident throughout the Arctic. Meanwhile, Arctic sea ice declined rapidly to unprecedented low extents in all months, with late summer experiences the most significant declining. Warming in the north is also evident from observations of early melting of snow and reducing snow cover. Now a key question is: in the warmth limited northern biome, what will happen to the phenological patterns of tundra vegetation as the global climate warms and seasonality of air temperature, sea ice, and snow cover shift? To answer the question we examined the onset of vegetation greenness, senescence of greenness, length of growing season, and dates of peak greenness along Arctic bioclimate gradients (subzones) to see how they change over years. Here, we combine multi-scale sub-pixel analysis and remote sensing time-series analysis to investigate recent decadal changes in vegetation phenology along spatial gradients of summer temperature and vegetation in the Arctic. The datasets used here are AVHRR 15-day 8 km time series, AVHRR 8-day 1 km dataset, and MODIS 8-day 500m Collection 5 dataset. There were detectable changes in phenological pattern over tundra biome in past two decades. Increases of vegetation greenness were observed in most of the summer periods in low arctic and mid-summer in high arctic. Peak greenness appeared earlier in high arctic and declined slower after peak in low arctic. Generally, tundra plants were having longer and stronger photosynthesis activities, and therefore increased annual vegetation productivities. Field studies have observed early growth and enhanced peak growth of many deciduous shrub species in tundra plant communities. These changes in seasonality are very likely to alter surface albedo and heat budget, modify plant photosynthesis/respiration and soil microbial activities, and even change hydrological patterns in the arctic. Next step, data fusing and assimilation of multi-sensor data with process models will be applied to improve the interpretation.
C21C-0560
Microstructural Studies on Snow from Dome Concordia (Antarctica)
Snow structure and its evolution as a function of depth is an important parameter in several fields of glaciology. In particular it influences the snow thermal properties, the dynamic of snow and firn densification as well as remote sensing signals. We present here results from an intensive study of the snow properties down to 3 m depth at Dome Concordia (Antarctica), a site which experiences very cold conditions (-53.8° C) and very low accumulation rate (2.5 g/cm2.yr). Several methods (near IR photography, coaxial episcopy, snow micro pen, …) have been used to get quantitative information on snow stratigraphy and snow structure. Among the measured properties, we will discuss in particular the relationship between the Snow Specific Area, determined by coaxial episcopy, and near IR pit image. This later technique potentially allows to get continuous profile of the SSA an important snow parameter. Another aspect of our study is linked to the interpretation of the temperature profile down to 20 m depth. We will discussed the evolution of this profile during the year in term of snow thermal conductivity.
C21C-0561
Refining Distributed Snowmelt Models in a Mountain Environment
In the radiation-dominated snowmelt environment of the Tuolumne River basin in Yosemite National Park, California, we have observed that snowmelt models using explicit representations of the snow surface energy balance are superior to bulk temperature-index models for capturing spatial and temporal variations in snowmelt. In comparing a temperature-index model (Snow-17) with two energy-balance models (the Utah Energy Balance model – UEB and the Distributed Hydrology Soil Vegetation Model – DHSVM), we found that the details in melt timing necessary for ecosystem studies are best simulated using the energy balance models. Additionally, two model-derived variables, namely shortwave albedo and snowpack surface and internal temperatures, are critical to accurate estimation of snowmelt onset, spring melt volumes, and the timing of hydrograph recessions, including when ephemeral streams go dry. We examine model representations of these variables as implemented in UEB and DHSVM using comparisons with observed streamflow and MODIS snow-covered area (SCA) and albedo derived from the MODSCAG algorithms.
C21C-0562
Combining MOD10A1 and MYD10A1 Images For Snow Cover Area Monitoring
MOD10A1 and MYD10A1 daily snow cover maps at 500 m resolution are available from MODIS sensors on Terra and Aqua satellites. Aqua obtains the image of same region approximately three hours after Terra over Turkey region. MODIS is an optic sensor and cloud cover degrades the usability of derived snow cover maps. Moreover, spectral similarity between clouds and snow complicates their separability in visible imagery. Fortunately, dynamic behavior of clouds enables their discrimination from snow stationary on the surface. Combined use of MOD10A1 and MYD10A1 images mostly reduces the cloud cover present in one image alone and provides better representation of surface snow cover. Comparison of merged images with in situ data indicated higher hit ratios. The individual comparison of MOD10A1 and MYD10A1 images with ground data each yielded 31% hit ratio whereas, the merged images provided 38%. One-day shifts in comparisons increased hit ratios to 52 % and 46% whereas and two-day shifts gave 77 % and 79 % for MOD10A1 and MYD10A1 respectively. Merged maps yielded 54% and 83% for one and two day shifts. The improvement provided by the merging technique is found to be 7% for the present day, 7 % for one- day and 5% for two-day shifts for the whole season. Monthly decomposition resulted 25% improvement as the maximum. The snow cover product obtained by merging Terra and Aqua satellites provided higher hit ratios, as expected.
C21C-0563
Shades of White: Explaining Variations in Clear-Sky Snow Albedo
Multi-year data from four Antarctic automatic weather stations show that clear-sky snow albedo ranges from 0.77 to 0.88, varying in time and space. The broadband albedo of a snowpack under clear-sky conditions is largely determined by snow grain size, solar zenith angle, and the vertical composition of the atmosphere -- but it is not obvious which process contributes what. A model that describes broadband radiative transfer in both the atmosphere and the snowpack is used to show that the observed variations in clear-sky snow albedo is dominated by strong spatial and temporal variation in snow grain size. Average values of snow grain size range from 22 μm on the Antarctic plateau to 64 μm on the ice shelves, whereas maximum values (December, January) are 40--150% higher. The seasonal cycle in solar zenith angle explains at most 0.02 of albedo variation at a location, whereas spatial variation in the atmospheric optical thickness leads to a 0.01 difference between plateau and ice shelf albedos. The most critical success factor for this method is the quality of the radiation observations, which is most notably affected by the horizontal tilt of either the radiation sensors or the snow surface. In order to tie model and observations together, we carried out an experiment at Summit, Greenland, where we simultaneously measured albedos and snow grain sizes, about which results will be presented.
C21C-0564
Canopy-snow Interaction in a sub-Arctic Shrub Tundra With Shrubs Buried by Snow
The accumulation and ablation of the winter snowpack is strongly affected by the structure and distribution of vegetation. Near the tree-line where the boreal forest transitions to sub-Arctic tundra, wind scours snow from the open tundra and deposits it near the edges of shrub tundra and forests. The Trail Valley Creek (TVC) watershed is located just north of the forest-tundra transition at 68°45'N, 133°30'W in the Northwest Territories and is characterized by gently rolling hills with some deeply incised river valleys. Tundra vegetation dominates much of the upland areas, with shrub tundra and sparse black spruce found on hill slopes and in the valley bottoms. Typically, open tundra and areas with short shrubs are snow-covered for most of the winter and exhibit a high winter albedo (~0.8) characteristic of the snowpack, whereas dense forests are very efficient at trapping solar radiation and can maintain a low albedo (~0.2-0.3) for much of the winter. The winter albedo of sparse forests and unburied shrubs falls between these values. Tall shrubs are interesting because the effect of the shrubs on the snowpack and energy exchange is dependent on whether the shrubs remain standing during the winter, or are bent over and buried by the snowpack, subsequently springing up during the spring melt. In this study the influence of the canopy on snowmelt and energy exchanges is examined at a tall tundra shrub (TTS) site in the TVC basin for the 2003 spring melt. The shrubs were bent over and buried during the winter and sprang up over a short period during the spring melt. Measurements from a nearby tundra upland (TUP) site serve as a basis of comparison and the interpretation is aided using simulations performed with the Canadian Land Surface Scheme (CLASS). Snow accumulation was higher at TTS than at TUP, but lower than the over-winter snowfall, suggesting that transport out of TTS by wind and/or sublimation was significant. The albedo was similar at the two sites (~0.8) at the start of the melt, but decreased quickly by almost 50 percent at TTS as the shrubs sprang up. The smaller albedo at TTS resulted in an additional 6.8 MJ m-2 d-1 of absorbed solar radiation (6.3 MJ m-2 d-1 modelled) during the primary melt period, relative to TUP. According to CLASS simulations, additional shading from the canopy at TTS resulted in a smaller net solar radiation over the snowpack of about 1 MJ m-2 d-1 relative to TUP. However, modelled values of net longwave radiation at the snowpack show an increase at TTS relative to TUP following the shrubs springing up, with daily average values being 1.9 MJ m-2 d-1 larger. The result is an increase in modelled net radiation over the snowpack at TTS relative to TUP of 0.9 MJ m-2 d-1. Both snow surveys and CLASS simulations show that following the shrubs springing up, the melt rate was initially faster at TTS relative to TUP but this lasted only a few days. The measurements and simulations show that much of the additional absorbed solar radiation at TTS was expended through larger turbulent heat fluxes (particularly sensible heat) during this time. Most of the larger initial SWE at TTS was depleted shortly following the shrubs springing up, after which the melt proceeded at a similar rate at both sites which subsequently became snow-free at about the same time.
C21C-0565
Scenario Calculations for Alpine Snow Cover and Runoff
The snow cover in the Alps is heavily affected by climate change. Recent data show that at altitudes below 1200 m a.s.l. a time-continuous winter snow cover is becoming an exception rather than the rule. This will also change the timing and characteristics of river runoff in Alpine catchments. We present an assessment of future snow and runoff in two Alpine catchments, the larger Inn catchment (1945 km2) and the smaller Dischma catchment (43 km2), based on two common climate change scenarios (IPCC A2 and B2). The changes in snow cover and runoff are predicted using ALPINE3D, a model for the high resolution simulation of alpine surface processes, in particular snow, soil and vegetation processes. The predicted changes in snow and runoff are drastic. While the current climate still supports permanent snow and ice at the altitudes of the highest peaks above 3000 m a.s.l., this zone will disappear under the future climate. The changes in snow cover can be summarized by approximately shifting the elevation zones down by 900 m. The corresponding changes in runoff are also severe: While the current climate shows a significant contribution from snow melt until mid to late summer, future climate will feature a much narrower snow melt runoff peak in spring. A further observation is that heavy precipitation events in the fall will change from mainly snow to mainly rain and will have a higher probability to produce flooding.
C21C-0566
VALIDATION OF SPRING SNOWMELT MODELLING IN AN ARCTIC ENVIRONMENT USING SPATIALLY DISTRIBUTED TURBULENT FLUXES DERIVED FROM AIRCRAFT MEASUREMENTS
Snow cover is commonly characterized by strong spatial variability as a result of both the spatial patterns of the meteorological forcings, and, above all, redistribution due to blowing snow events. This is particularly evident for arctic environments, where the landscape during spring snow melt is characterized by a mosaic of snow-covered and bare ground patches. Many efforts to model such variability have been completed using remotely sensed snow covered area maps as calibration and validation tools. However, it is also very important to validate the single components of the surface energy balance in order to ensure that their parameterizations are appropriate, and that good modelling results are not due to the compensation of offsetting errors. The availability of validation maps of energy fluxes is quite limited, but, recently, 2D information on turbulent fluxes has been derived from aircraft measurements. In this work, maps of sensible and latent heat fluxes derived from NRCC Twin Otter aircraft measurements collected during the Mackenzie GEWEX, during the melting season, have been used and compared with corresponding modelled fluxes. In particular, the distributed hydrological model GEOtop has been applied to the highly instrumented arctic catchment of Trail Valley Creek. The aircraft flux data covered almost the entire drainage basin. GEOtop in particular deals with the small-scale spatial variability of topography, and jointly solves the water and energy balance for the snowpack and soil. It also takes into account soil freeze-thaw processes, which can also have implications on snow hydrologic processes. GEOtop has been coupled with other submodels that better represent snow transport processes in arctic environments. Though some problems related to modelling the shrub tundra snow are present, the results are promising and suggest a reasonable correspondence between modelled and observed energy fluxes.
C21C-0567
The Development of the GENESY hydrometeorological model for mountain environments
It is important to understand the local variability in climatic conditions over complex terrain to make accurate assessments of environmental change in fresh water ecosystems. This work presents the continued development and application of a spatial mountain hydrometeorological model, GENESYS (GENerate Earth SYstems Science) using data within St. Mary River Watershed, Glacier Park, Montana. We use SNOTEL and snow survey data to derive a precipitation function that accounts for influences of seasonality and elevation. Following the addition of subroutines for evapotranspiration, soil water modeling, sublimation and canopy interception, GENESYS is used to simulate daily hydrometeorological variables at a fine spatial resolution over the St. Mary River Watershed. Simulated daily snowpack values compared well with observed data for a 10 year trial period. Simulated runoff volumes also compare well with observed annual stream flow for a thirty year period. GENESYS simulations of April 1 snow water equivalent (SWE) and August 31 soil moisture demonstrate the applicability of the model in estimating change in hydrological conditions. We show that GENESYS is a practical tool for detailed assessments of the impacts of environmental change in mountain environments.
C21C-0568
AN IMPROVEMENT OF SNOW PARAMETERIZATION IN ORDER TO BETTER REPRESENT SOIL TEMPERATURES
The snow cover commonly exhibits large vertical gradients in terms of temperature, density, and water content. In order to take into account such variability many models represent the snowpack with multilayer schemes, normally using thinner layers near the surface, and with thicker layers downwards, since the largest gradients occur at the atmosphere interface. However, in deep alpine snowpacks it has been found that such parameterization may not well describe the heat flux exchanged between the snow cover and the soil. In particular, if the thermal conductivity at the layer interfaces is modelled with an harmonic mean, having a deep snow layer and a neighboring thin soil layer may result in a very small thermal conductivity, and, consequently, a strong underestimation of the snow-soil heat flux. This heat flux component has actually a very important implication in the thermal and hydraulic regime of permafrost and seasonally frozen ground, especially during the melting season, when the snow disappearence time should be carefully described, and, therefore, it should be carefully determined. This work shows that a symmetrical discretization of the snow with thin layers located at both extremities of the snowpack considerably improves the soil temperature modelling. For example, representing the snow cover just with three layers, of which the upper and the lower particularly thin, could be a good compromise between computational speed and model precision. This new parameterization has been included in the GEOtop model, a physically based hydrological model that jointly solves the water and soil energy balance for snow and soil. The model has been applied to a heavily instrumented site in the Swiss Alps (Murtel Rock Glacier) showing good results.
C21C-0569
Assessing potential environmental change impacts in the St. Mary River watershed, Montana
Mountain watersheds provide up to 90 percent of the fresh water for many of the world's river systems. Unfortunately, mountainous regions are highly susceptible to even slight changes in climatic conditions. It is important to understand how mountain environments will respond to changes in climate in order to properly manage natural resources. In order to objectively assess potential changes in mountain watersheds and subsequent effects on water supply, we use a combination of hydrometeorological and climate models. This work describes the application of the GENESYS (GENerate Earth SYstems Science) spatial hydrometeorological model to assess possible future changes in watershed processes for the St. Mary River basin, Glacier Park, Montana. Mean temperature and precipitation changes from 5 General Circulation Models (GCMs), representing a range of plausible future climates, are applied using the delta technique. The GENESYS model is used to simulate potential changes in watershed variables including spring snowpack, the onset of melt, and changes in snow extent for the 2020s, 2050s, and 2080s. Our results suggest the St. Mary basin is highly susceptible to altered temperature and precipitation regimes and should be monitored and managed in a way that will enable adaptation to climate change.
C21C-0570
Sensitivity analysis of GENESYS Model simulations in an alpine environment
This work presents a sensitivity analysis of output from the GENESYS (GENerate Earth SYstems Science) mountain hydrometeorological model. GENESYS has been under continued development at the University of Lethbridge for a number of years. This application addresses model response to variations in input variables. Multiple GENESYS runs for the St. Mary River Watershed, Glacier Park, Montana are analyzed to determine model sensitivity to individual input variable and modeled response ranges for output variables to specific ranges of input perturbation. The spatial variation in the hydrometeorology is illustrated using a 140 year time series that includes historical data from 1960-2004 and modelled data for the period 2010-2099. Preliminary analyses demonstrate response ranges of modeled variables including maximum and minimum temperatures, precipitation and overwinter snow water equivalent for a variety of terrain derived modeling units in the watershed.
C21C-0571
Effects of Climate Change on White-Water Recreation on the Salmon River, Idaho
White-water recreation on the Salmon River generates tens of millions of dollars each summer for central Idaho's economy. This tourism revenue is highly dependent on a healthy snowpack melting throughout the summer to meet minimum streamflow requirements for the rafting industry. A number of previous studies have shown that in a warming climate this vital snowpack will diminish and so will summer streamflows. In areas such as the Middle Fork of the Salmon River this will result in less streamflow in July and August, which are the critical months for the rafting industry. Current estimates approximate that eight percent of scheduled trips are canceled due to low summer streamflows. In this study we project future impacts to white-water recreation in the Salmon River basin, associated with an ensemble of climate change scenarios. The University of Washington's Climate Impacts Group has statistically downscaled 20 GCMs A1B and B1 climate change scenarios from the IPCC 2007 Report. We use these forcings to run the Variable Infiltration Capacity (VIC) land-surface model to determine future streamflows for the Pacific Northwest. To verify the likelihood of non-boatable days in the future due to low summer streamflows, we compare this suite of projected results for the Salmon River streamflow to historical streamflows for the Middle Fork and Main Fork Salmon. Preliminary analysis shows a two degree Celsius increase could result in a twenty-five percent cancellation of future Middle Fork trips as a result of low summer streamflows. On the Middle Fork section alone this translates into a two million dollar loss in annual revenue generation for the rafting industry, with impacts stretching deeper into the economy. We also discuss additional costs to the users, the tourist economy and potential analysis for other river systems.
C21C-0572
Snowmelt infiltration and evapotranspiration in Red Fir forest ecosystems of the Sierra Nevada
Measurements from two forested catchments in the mixed conifer, Red Fir zone of the southern Sierra Nevada (2,200-2,600 m elevation) demonstrate the controls that topography and canopy cover exert over snow cover. Snow-depth, soil-moisture, stream-stage and sap-flow measurements from the Wolverton basin in Sequoia National Park and Teakettle Experimental Area in the Sierra National Forest exhibit distinct and rapid responses to spring snowmelt. Spatial heterogeneity in snow water equivalent is influenced by tree clusters and individual canopies. Snowmelt and soil moisture timing are controlled by proximity to stem of tree and canopy clustering. Aspect and slope position affect soil moisture, with drier conditions predominating on the steeper slopes. Synchronous fluctuations in soil moisture, stream flow and sap flow were observed. Continuous instrumental and synoptic survey data show snow water equivalent is approximately 15% less 1 m from the tree stem than in open areas at peak accumulation, with snowmelt occurring in shaded open areas 1 to 4 weeks later than under canopy. Soil moisture tracks the snowmelt pattern closely, with diel fluctuations in soil moisture under saturated conditions followed by an exponential dry-down to field capacity after snowmelt. This is followed by a prolonged summer drought punctuated by rain events of <5% of total precipitation. Synoptic surveys of steeper slopes and multiple aspects show consistent patterns of drier conditions on steeper terrain.
C21C-0573 INVITED
Preferential Deposition of Precipitation in Alpine Terrain and some Implications
The inhomogeneous snow distribution typical for alpine terrain is the result of wind and precipitation interacting with the (snow) surface over topography. We discuss preferential deposition of precipitation as the deposition process without erosion of previously deposited snow and thus in absence of saltation. A numerical model, Alpine3D, describing all relevant surface and transport processes uses high-resolution wind fields calculated with a meteorological model, ARPS. The model is used to simulate deposition patterns over a steep alpine ridge and small glaciers in Switzerland and Austria. Preferential deposition is shown to be a dominant process at the ridge scale. Snow deposition correlates well with the horizontal and vertical component of the mean wind vector as influenced by the topography and predicted by ARPS. Based on simulated and observed patterns of snow deposition it is concluded that the survival of small Alpine glaciers critically depends on preferential deposition of snow precipitation.
C21C-0574
Sub-canopy radiant energy during snowmelt in non-uniform forests spanning a latitudinal transect
In mountainous, forested environments, snowcover dynamics exert a strong control on hydrologic and atmospheric processes. Snowcover ablation patterns in forests are controlled by a complex combination of depositional patterns coupled with radiative and turbulent heat flux patterns related to topographic and canopy cover variations. Quantification of small-scale variations of radiant energy in forested environments is necessary to understand how canopy structure affects snowcover energetics to improve the representation of snowmelt processes in spatially-explicit physically-based snowmelt models. Incoming solar and thermal radiation were measured during the melt season within continuous and discontinuous forest stands, and at the interface between forest patches and small clearings along a transect spanning the North American Cordillera. Results indicate that reductions in solar radiation at the snow surface are partially balanced by increased thermal radiation from the forest canopy, relative to open locations. The differences between the transfer processes for solar and thermal radiation can produce two net incoming and net snowcover radiation paradoxes in heterogeneous environments. In discontinuous canopies, net radiation in forested areas may exceed radiation in open sites, whereas in other situations, net radiation may be less than net radiation in closed canopy forests. The empirical results coupled with theoretical modeling indicates that the effects of forest canopies on the radiative regimes at the snow surface are controlled by complex interactions of slope, aspect, gap sizes, canopy height, canopy density, canopy temperature, snow surface temperature and snowcover albedo. In higher latitude, closed canopy forests, radiative regimes may be characterized by relatively simple geometric optical radiation transfer methods, whereas at lower latitude and more non- uniform forests, other processes such as canopy and stem heating must be considered. These net radiation differences coupled with decreased turbulent fluxes due to lower wind velocities and reduced snow water equivalent values due to canopy interception losses help to explain small-scale patterns of snowmelt in non- uniform forested areas. Future investigations will use physically based models coupled with LiDAR derived topographic and vegetation data to assess how these small-scale processes integrate in both space and time to control the timing and rates of snowcover ablation in complex vegetated terrain.
C21C-0575
Using wind fields from a high resolution atmospheric model for simulating snow dynamics in mountainous terrain
Wind-induced snow transport processes lead to a significant variability of the snow cover. Knowledge about this variability is important for e.g. determining the temporal dynamics of the snowmelt runoff. For predicting the correct amount of transported snow knowledge of the local wind-field is an essential. In high-alpine rugged relief wind fields can hardly be provided by a simple interpolation of station recordings. In this work we use a modified version of the PSU/NCAR Mesoscale Model MM5 to derive wind fields for a 450 km² area at a target resolution of 200 m, accounting for topography and related dynamic effects. We have modelled 220 wind fields representing the most characteristic wind situations within the test-area. The criteria for the extraction of the wind field for the current snowmodel (SNOWTRAND-3D) time step are mean wind speeds and directions in the 700 hPa level derived from DWD (German Weather Service) Local Model reanalysis data with a temporal resolution of one hour. These data are then compared with the corresponding mean wind speeds and directions from the appropriate MM5 nesting area indicating which one of the library files represents the best fit. Verification is conducted by comparison of historical station measurements with corresponding downscaled simulation results. For this downscaling a semi-empirical approach is utilized which accounts for topographic effects. Results for the winter seasons 2003/04 and 2004/05 showing that the presented scheme is able to improve the quality of SNOWTRAN-3D runs with respect to the snow height.
C21C-0576
Assimilation of AATSR, MERIS and MODIS Data in the Snowmelt Runoff Model (SRM) on the Upper Rio Grande (USA)
Current efforts for simulating or forecasting snowmelt are time-consuming and laborious; the AWARE project (A tool for monitoring and forecasting Available WAter REsource in mountain environments) has been motivated by the urgent need to facilitate the prediction of medium-term flows from snowmelt for an effective and sustainable water resources management. Its main goal is to provide innovative tools for monitoring and predicting water availability and distribution in drainage basins where snowmelt is a major component of the annual water balance. The particular objective of the effort reported here is to compare results obtained from the MODIS sensor on NASA Terra and Aqua satellite and next generation sensors AATSR and MERIS on board ESA Envisat satellite. The vehicle for this comparison is the AWARE Geoportal (http://www.aware- eu.info/eng/home.htm) which is a WWW implementation of the Snowmelt Runoff Model (SRM). The river basin chosen for analysis is the Upper Rio Grande of North America. The time period for analysis encompasses the Water Years 2005, 2006, and 2007 (October 2004 – September 2007). The reason for this is to ensure that data from all three sensors are available for use and to investigate variable climate conditions. A successful comparison between the various sensors will help demonstrate that the AWARE approach will facilitate future processing of several years' worth of snow cover data from a variety of sensors that covers large extremes in climate variability. This will allow greater success in developing forecasts and understanding of longer term climate change impacts.
C21C-0577
Snow Avalanche Disturbance Ecology: Examples From the San Juan Mountains, Colorado.
We evaluated landscape ecology approaches to characterize snow avalanche paths based on patterns of plant species composition and evidence of disturbance. Historical records of avalanche incidents, patterns in the annual growth layers of woody plants, and distributions of plant species can be used to quantify and map the frequency and magnitude of snow slide events. Near Silverton, Colorado, a series of snow storms in January of 2005 resulted in many avalanche paths running full track at 30 and 100 year return frequency. Many avalanches cut fresh trimlines, widening their tracks by uprooting, stripping, and breaking mature trees. Powerful avalanches deposited massive piles of snow, rocks, and woody debris in their runout zones. We used cross-section discs and cores of representative downed trees to detect dendro-ecological signals of past snow avalanche disturbance. Avalanche signals included impact scars from the moving snow and associated wind blast, relative width of annual growth rings, and development of reaction wood in response to tilting. Initial measurements of plant diversity and disturbance along the elevation gradient of an avalanche path near Silverton indicate that avalanche activity influences patterns of forest cover, contributes to the high local plant species diversity, and provides opportunities for new seedling establishment.
C21C-0578
Observed SWE trends and climate analysis for Northwest Pacific North America: validation for future projection of SWE using the CRCM and VIC
Observational databases of snow water equivalent (SWE) have been collected from Alaska, western US
states and the Canadian provinces of British Columbia, Alberta, Saskatchewan, and territories of NWT, and
the Yukon. These databases were initially validated to remove inconsistencies and errors in the station
records, dates or the geographic co-ordinates of the station. The cleaned data was then analysed for
historical (1950 to 2006) trend using emerging techniques for trend detection based on (first of the month)
estimates for January to June. Analysis of SWE showed spatial variability in the count of records across the
six month time period, and this study illustrated differences between Canadian and US (or the north and
south) collection. Two different data sets (one gridded and one station) were then used to analyse April 1st
records, for which there was the greatest spatial spread of station records for analysis with climate
information. Initial results show spatial variability (in both magnitude and direction of trend) for trend results,
and climate correlations and principal components indicate different drivers of change in SWE across the
western US, Canada and north to Alaska. These results will be used to validate future predictions of SWE
that are being undertaken using the Canadian Regional Climate Model (CRCM) and the Variable Infiltration
Capacity (VIC) hydrologic model for Western Northern America (CRCM) and British Columbia (VIC).
http://www.pacificclimate.org/resources/publications/bennett.agu2008/
C21C-0579
Important factors influencing microstructural estimates derived from the SnowMicroPen
Snow scientists utilize the SnowMicroPen (SMP) to record high resolution hardness profiles of seasonal snowpacks. Microstructural and statistical estimates have been derived from the signal to characterize snow types and changes in snow strength and stability. Two important factors affecting these estimates are signal quality and signal processing. In this study we utilize a spatial dataset of approximately 800 SMP profiles to demonstrate how these factors influence end results. If gone unchecked, poor signal quality significantly influences results and their implications. However, given the high percentage of good quality profiles in this dataset, faulty profiles could be excluded without hindering the analysis. Slight differences in applications of micromechanical models and sampling criteria also influence results significantly. These findings demonstrate that quality control is paramount at the outset of an analysis and that careful consideration and documentation of micro-mechanical applications and sampling criteria are necessary for results to be applicable to other studies.