Robert E. Davis

Retrieval of subpixel snow-covered area and grain size from imaging spectrometer data

We describe and validate an automated model that retrieves subpixel snow-covered area and effective grain size from Airborne Visible/ Infrared Imaging Spectrometer (AVIRIS) data. The model analyzes multiple endmember spectral mixtures with a spectral library of snow, vegetation, rock, and soil. We derive snow spectral endmembers of varying grain size from a radiative transfer model; spectra for vegetation, rock, and soil were collected in the field and laboratory. For three AVIRIS images of Mammoth Mountain, California that span common snow conditions for winter through spring, we validate the estimates of snow-covered area with fine-resolution aerial photographs and validate the estimates of grain size with stereological analysis of snow samples collected within 2 h of the AVIRIS overpasses. The RMS error for snowcovered area retrieved from AVIRIS for the combined set of three images was 4%. The RMS error for snow grain size retrieved from a 3 � 3 window of AVIRIS data for the combined set of three images is 48 Am, and the RMS error for reflectance integrated over the solar spectrum and over all hemispherical reflectance angles is 0.018. D 2003 Elsevier Science Inc. All rights reserved.

Spatial Snow Modeling of Wind-Redistributed Snow Using Terrain-Based Parameters

Abstract Wind is widely recognized as one of the dominant controls of snow accumulation and distribution in exposed alpine regions. Complex and highly variable wind fields in rugged terrain lead to similarly complex snow distribution fields with areas of no snow adjacent to areas of deep accumulation. Unfortunately, these complexities have limited inclusion of wind redistribution effects in spatial snow distribution models. In this study the difficulties associated with physically exhaustive wind field modeling are avoided and terrain-based parameters are developed to characterize wind effects. One parameter, , was based on maximum upwind slopes relative to seasonally averaged winds to characterize the wind scalar at each pixel location in an alpine basin. A second parameter, , measured upwind breaks in slope from a given location and was combined with an upwind application of to create a drift delineator parameter, D0, which was used to delineate sites of intense redeposition on lee slopes. Based on 504 ...

Estimating the spatial distribution of snow water equivalence in a montane watershed

An approach to model distributed snow water equivalence (SWE) that merges field measurements of depth and density with remotely sensed snow-covered area (SCA) is described. In 1993, two teams conducted an intensive snow survey in the 92. 8k m 2 Blackcap Basin of the Kings River. Snow depth was measured at 709 points and density in five snow pits and along five transects using a Federal Sampler. Sample locations were chosen to be representative of the range of elevation, slope and aspect of the basin. Regression tree models showed that net radiation, elevation and slope angle account for 60‐70% of the variance in the depth measurements. Density was distributed over the basin on a 30 m grid with a multiple linear regression model that explained 70% of the observed variance as a function of the same three variables. The gridded depth estimates, combined with modelled density, produced spatially distributed estimates of SWE. An unsupervised spectral unmixing algorithm estimated snow cover fractions from Landsat-5 Thematic Mapper data acquired at the time as the snow survey. This method provides a snow cover fraction estimate for every pixel. This subpixel map was used as the best estimate for SCA and, combining it with the SWE map, allowed computation of the SWE volume. The estimated volume using the subpixel SCA map was compared with several SCA maps produced with simulations of binary SCA mapping techniques. Thresholds of 40, 50 and 60% fractional cover were used to map binary cases of full snow cover or no snow cover. The diAerence in basin SWE volume was up to 13% depending on the threshold used to classify snow-covered versus snow-free areas. The percentage diAerences in volumes show a significant correlation to the percentage diAerences in SCA between the methods. #1998 John Wiley & Sons, Ltd.

Snow ablation modeling at the stand scale in a boreal jack pine forest

The purpose of this study is to predict spatial distributions of snow properties important to the hydrology and the remote sensing signatures of the boreal ecosystem. This study is part of the Boreal Ecosystems Atmosphere Study (BOREAS) of central Saskatchewan and northern Manitoba. Forested environments provide unique problems for snow cover process modeling due to the complex interactions among snow, energy transfer, and trees. These problems are approached by coupling a modified snow process model with a model of radiative interactions with forest canopies. Additionally, a tree well model describes the influence of individual trees on snow distribution on the ground. The snow process and energy budget model calculates energy exchange at the snow surface, in-pack snow processes, melting and liquid water flow, heat conduction, and vapor diffusion. The surface radiation model provides input on the radiation receipt at the snow surface for model runs in the jack pine forest. Field data consisted of measured meteorological parameters above and within the canopy, spatial variability of snow properties, and variations of incoming solar irradiance beneath the forest canopy. Results show that the area beneath tree canopies accumulated 60% of the snow accumulated in forest openings. Peak solar irradiance on the snow cover was less than one half that measured above the canopy. Model runs are compared between the open and the forested sites and show the open area ablating four days before areas beneath the canopy and eight days before forest openings and compare favorably with measured data. Physically based modeling of snow ablation was successful at the forested site and nearby open area.

Transmission of solar radiation in boreal conifer forests: Measurements and models

A combined geometric-optical and radiative transfer (GORT) model allows incorporation of multiple scales of clustering in conifer canopies on the estimation of radiation transmission. Consideration of clustering of branches into whorls is the latest addition to this model. Modification of the GORT model to include whorl orientation improves the ability to model the observed patterns of solar radiation transmission as a function of solar zenith angle and height in the canopy. Whorl orientation distributions are derived from multidirectional measurements using a geometric optical mutual shadowing model. For BOREAS test stands, model estimates and vertical measurements of photosynthetically active radiation transmittance within the canopy show (1) general decreases in transmission as solar zenith angles increase in the range of solar zenith angles dominated by beam irradiance, (2) increases in PAR transmission at very high solar zenith angles where diffuse skylight is dominant, (3) maximum scattering and absorption occur in the middle of the canopy. Model estimates match measurements from the forest floor, indicating the value of the model for providing radiation inputs to snowmelt models in forested landscapes.

Variation of snow cover ablation in the boreal forest: A sensitivity study on the effects of conifer canopy

The duration and meteorological history of winter and thaw periods in the boreal forest affect carbon exchange during the growing season. Characteristics of conifer canopies exert important control on the energy exchange at the forest floor, which in turn controls snow cover processes such as melting. This analysis investigated the role of the conifer tree characteristics, including height and canopy density. Canopy and snow models estimated radiation incoming to the snow surface, the net energy budget of the snow, and melting rates of snow cover under conifer forests with different canopy density and tree height. This analysis assumed that canopy effects dominated snow surface energy exchange under conifers in the boreal forest. We used data layers of forest characteristics from the Boreal Ecosystem-Atmosphere Study (BOREAS) modeling subareas in Saskatchewan and Manitoba to guide the choice of modeled tree height and canopy density. Modeled stand characteristics assumed random location of trees and used a uniform tree height within a stand and regular crown geometry scaled to tree height. Measurements during winter and thaw in 1994 of incoming solar and longwave radiation, humidity, and wind speed above the forest canopy provided input to the models, along with air temperature measured in the canopy. Results showed the importance of canopy density and tree height as the first-order controls on cumulative incoming solar radiation at the forest floor for the range of these variables in the BOREAS test area. The combined canopy and snow models showed a large range of snow ablation within conifers, which showed the trade-offs between canopy density and tree height. Solar fluxes dominated the net transfer of energy to the snow in the north, while sensible heat exchange, net solar, and net longwave radiation played important roles in the south.

Climate and energy exchange at the snow surface in the Alpine Region of the Sierra Nevada: 1. Meteorological measurements and monitoring

A detailed evaluation of climate conditions in a small alpine watershed, typical of much of the southern Sierra Nevada, is presented for the 1986 water year. Measurements of snowfall, meteorological and snow cover conditions, and snow cover ablation are used to characterize the climate at four locations in the watershed during that snow season. Data from these locations are then combined into two representative sites for the watershed. Measurement approaches and methodologies and the effectiveness of instrumentation used in the study are discussed, and an estimate of the uncertainty of the monitored meteorological parameters is made. The data are integrated into a continuous hourly time series of solar and thermal radiation, air, snow and soil temperature, humidity, and wind at the two representative sites in this remote alpine watershed for an entire snow season. Snow deposition and snow cover depth and density are measured manually at regular intervals throughout the snow season. While problems were encountered monitoring air and snow surface temperature, humidity, and wind, because of the extreme conditions which are likely to occur in an alpine environment, radiation is easily monitored, and the estimated uncertainty of all measured parameters was acceptably low. This effort was required to develop a high quality time series of integrated climate data to evaluate the components of the energy balance of the snow cover during both deposition and ablation conditions.

Snow ablation modelling in a mature aspen stand of the boreal forest

Snow ablation modelling at the stand scale must account for the variability in snow cover and the large variations of components of energy transfer at the forest floor. Our previous work successfully predicted snow ablation in a mature jack pine stand by using a one-dimensional snow process model and models predicting radiation below forest canopies. This work represents a second test of our basic modelling scenario by predicting snow ablation in a leafless, deciduous aspen stand and verifying the results with field data. New modifications to the snow model accounted for decreased albedo owing to radiation penetration through optically thin snowpacks. A provisional equation estimates litter fall on the snowpack, thereby reducing the areal averaged albedo. We showed that subcanopy radiation measurements can be used with a canopy model to estimate a branch area index for defoliated aspen as an analogue to the foliage area index used for conifers. Modelled incoming solar and long-wave radiation showed a strong correlation with measurements, with r 2 a 0.96 and 0.91 for solar and long-wave radiation, respectively. Model results demonstrate that net radiation overwhelms turbulent exchanges as the most significant driving force for snowmelt in aspen forests. Predicted snow ablation in the aspen stand compared very favourably with available data on snow depth. #1998 John Wiley & Sons, Ltd.

Sublimation of Intercepted Snow within a Subalpine Forest Canopy at Two Elevations

Abstract To determine how elevation affects the sublimation rate from intercepted snow within a subalpine forest canopy, a cut subalpine fir and an artificial conifer were weighed at each of two elevations (3230 and 2920 m) at a U.S. continental site (39°53′N, 105°54′W) from 1 January to 1 May 2001. Measured stand characteristics included canopy density (67% and 75%) and basal area (43.4 and 24.1 m2 ha−1) for the higher and lower elevations, respectively. Temperature, relative humidity, net radiation, wind speed, and mass of snow on suspended trees provided data to determine whether sublimation rates of intercepted snow are more rapid at higher elevations associated with increased wind speed. Measurements showed the unexpected result that wind speed during sublimation periods was lower at higher elevations, probably because of terrain sheltering. The analysis examined 21 storm-free periods ranging in duration from 9 to 53 h. Sublimation rates per unit mass of intercepted snow were significantly larger at ...

Relating storm and weather factors to dry slab avalanche activity at Alta, Utah, and Mammoth Mountain, California, using classification and regression trees

Using classification and regression tree models, we evaluated 31 factors in terms of their importance to explaining avalanche activity indices at two ski areas: Alta, UT and Mammoth Mountain, CA. This study derived new empirical factors that combined wind velocity with new snow amount, air temperatures with time, and total snow depth with time. The analyses created over-fit tree models in exploring structures inherent in the data to obtain the relative ranking and scores of various combinations of the 31 factors. Avalanche activity indices included maximum size, number of releases and sum of sizes of released avalanches. Results showed that time lagged conventional factors describing snowfall and derived wind-drift parameters ranked highest in all tests. Snow drift factors segregated into prominent wind directions showed only moderate importance. Among the non-storm factors, the starting snow depth of a particular year ranked highest showing the importance of interannual variability. This was followed by the accumulated vapor pressure difference, which we formulated to better describe the conditioning of old snow with age. The average snow depth increase and other factors followed in importance.