Improving gridded snow water equivalent products in British
Columbia, Canada: multi-source data fusion by neural network
models

Understanding and characterizing the spatial distribution of snow are critical to represent the energy balance and runoff production in mountain environments. In this study, we investigate the interannual and seasonal variability in snow depth scaling behavior at the Izas experimental catchment of the Spanish Pyrenees (2,000 to 2,300 m above sea level). We conduct variogram analyses of 24 snow depth maps derived from terrestrial light detection and ranging scans, acquired during six consecutive snow seasons (2011–2017) that span a range of hydroclimatic conditions. We complement our analyses with bare ground topography data and wind speed and direction measurements. Our results show temporal consistency in the spatial variability of snow depth, with short‐range fractal behavior and scale break lengths that are similar to the optimal search distance (25 m) previously reported for the topographic position index, a terrain‐based predictor of snow depth. Beyond the 25‐m scale break, there is little to no fractal structure. We report a long‐range scale break of the order of 185–300 m for most dates—aligned with the dominant wind direction—and patterns between anisotropies in scale break lengths of shallow snow cover and directional terrain scaling behavior. The temporal consistency of snow depth scaling patterns suggests that, in addition to guiding the spatial configuration of physically based models, fractal analysis could be used to inform the design of independent variables for statistical models used to predict snow depth and its variability.

Section: Discussionmentioning

confidence: 99%

Interannual and Seasonal Variability of Snow Depth Scaling Behavior in a Subalpine Catchment

Mendoza

Musselman

Revuelto³

et al. 2020

“…Improving simulated and observation‐based gridded estimates of SWE is a major field of research (see, e.g., Snauffer et al, ; Painter et al, ), and, compared to the VIC model, more complex snow models exist. Feng et al () compared several snow models by validating them against observations and showed that VIC performs better than the complex Community Land Model, version 3 (Dai et al, ), and agrees well with the more complicated Snow Thermal Model (Jordan, ).…”

Section: Methodsmentioning

confidence: 99%

Snow Drought Risk and Susceptibility in the Western United States and Southwestern Canada

Dierauer

Allen

Whitfield

2019

In western North America (WNA), mountain snowpack supplies much of the water used for irrigation, municipal, and industrial uses. Thus, snow droughts (a lack of snow accumulation in winter) can have drastic ecological and socioeconomic impacts. In this study, the historical (1951–2013) frequency, severity, and risk (frequency × severity) of dry, warm, and warm and dry snow droughts are quantified at the grid‐cell and ecoregion scale for snow‐dominated regions in the western United States and southwestern Canada (sWNA). Based on multiple linear regression analysis, relationships between mean winter temperature, snow drought risk, and snow water equivalent sensitivity are explored. Piecewise linear regression is used to identify temperature thresholds for mapping temperature‐related snow drought susceptibility. Results highlight spatial differences in snow drought regimes across sWNA and reveal that temperature thresholds exist at −3.1 °C (±0.3 °C) and 1.4 °C (±0.3 °C), above which the warm snow drought risk increases more rapidly. Approximately 3% of the nonglaciated snow storage in this region has high susceptibility to temperature‐related snow drought, representing 11 km3 of water, or approximately one third the capacity of Lake Mead. Under a +2 °C climate scenario, an additional 8% (28 km3) of this snow storage volume will transition to high susceptibility.

“…These statistical models, which include a variety of approaches including regression models (such as multiple linear regression—MLR), binary regression trees, and lookup tables, have been applied using observations that span both larger (e.g., continental) scales (Bormann et al, ; Sturm et al, , ) and smaller (e.g., watershed) scales (Jonas et al, ; Wetlaufer et al, ). Other machine learning approaches such as Random Forests and Artificial Neural Networks have also become popular for estimating snow quantities (particularly SWE and snow cover) using a variety of input data including data from satellite sensors (e.g., Bair et al, ; Dobreva & Klein, ; Tedesco et al, ), land surface models (e.g., Snauffer et al, ), and ground observations (e.g., Tabari et al, ; Buckingham et al, ; Gharaei‐Manesh et al, ). These approaches have been shown to be highly adaptable to capture nonlinear relationships involved in snow measurement (Czyzowska‐Wisniewski et al, ), allowing them to outperform linear approaches such as MLR.…”

Section: Introductionmentioning

confidence: 99%

Improving Snow Water Equivalent Maps With Machine Learning of Snow Survey and Lidar Measurements

Broxton

Leeuwen

Biederman

2019

In the semiarid interior western USA, where a majority of surface water supply comes from mountain forests, high‐resolution aerial lidar‐based surveys are commonly used to study snow. These surveys provide rich information about snow depth, but they are usually not accompanied with spatially explicit measurements of snow density, which leads to uncertainty in the estimation of snow water equivalent (SWE). In this study, we use a novel approach to distribute ~300 field measurements of snow density with artificial neural networks. We combine the resulting density maps with aerial lidar snow depth measurements, bias corrected with a very large and precisely geolocated array of field‐measured snow depths (~4,000 observations), to create and validate maps of snow depth, snow density, and SWE over two sites along Arizona's Mogollon Rim in February and March 2017. These maps show differences between midwinter and late‐winter snow conditions. In particular, compared to that of snow depth, the spatial variability of snow density is smaller for the later snow survey than the earlier snow survey. These gridded data also show that the representativeness of Snow Telemetry and other point measurements is different for the midwinter and late‐winter snow surveys. Overall, the lidar artificial neural network SWE estimates can be as much as 30% different than if Snow Telemetry density were used with lidar snow depths to estimate SWE.