Using a Random Forest Model to Combine Airborne Lidar and Snotel Data for Daily Estimates of Snow Depth Across Mountain Drainage Basins of Colorado
Year: 2025
Type: article
Abstract: Machine learning (ML) has emerged as an effective tool for estimating snow depth and snow water equivalent at unsampled times and locations. Airborne lidar surveys are particularly useful for ML applications: the high‐resolution, high‐precision snow depth data allow for algorithm training and testing to an extent and spatial resolution not previously possible. Here, we train a random forest model to estimate snow depth relative to a nearby Snotel site using static physiographic data and dynamic (i.e., time‐dependent) snowpack data as predictor variables and lidar for the target variable. The model output is daily, 50 m resolution snow depth for basins that have both lidar and Snotel data in Colorado. We evaluated multiple approaches for random forest training: using historic lidar data in a basin (temporal transfer), using lidar data from other basins in a region (spatial transfer), and both together. The three approaches yield RMSE values ranging from 0.37 to 0.44 m at 50 m resolution, achieving lower errors compared to ML studies at higher resolutions. Model error decreases when outputs are upscaled, with RMSE values of 0.17 and 0.10 m for the 4 km and basin scales, respectively. The model scenario which includes both temporally and spatially transferred lidar data is the most robust to the number and timing of lidar surveys used in model training. This framework extends the spatial footprint of Snotel and the temporal coverage of lidar by leveraging the strengths of the two data sets, with applications for water resource management and validation of gridded snow products.
Source: Water Resources Research
DOI: 10.1029/2024wr039775
Authors: J. N. Herbert, M. S. Raleigh, E. E. Small
Institutions: University of Colorado Boulder, Oregon State University
Language: English
Subfield: Atmospheric Science
Field: Earth and Planetary Sciences
Domain: Physical Sciences
SDG: Climate action
Open Access status: gold
Estimated APC paid: $3,000
Awards: G1060899602, G1451190392, G3424600768, G3897489853, WE PROPOSE THAT THE DESIGN OF AN IDEAL FIELD CAMPAIGN FOR MEASURING SNOW DENSITY AND SWE IS ONE THAT INTEGRATES REMOTE SENSING MULTIPLE MODELS AND FIELD DATA. AN IDEAL FIELD SAMPLING PROTOCOL SHOULD TARGET SITES SPANNING (1) A RANGE OF UNCERTAINTY IN MODELED DENSITY AND (2) A GRADIENT OF LANDSCAPE SETTINGS LIKE FOREST DENSITY (FIGURE 2). THE FIRST STEP IS TO QUANTIFY UNCERTAINTY IN SNOW MODELS THAT ARISE FROM VARIOUS SOURCES INCLUDING METEOROLOGICAL FORCING [RALEIGH ET AL. 2015 2016] PARAMETERS [REBA ET AL. 2014] AND PROCESS FORMULATIONS [ESSERY ET AL. 2013]. THIS LEADS TO THE GENERATION OF A SINGLE MAP OF DENSITY UNCERTAINTY (FIGURE 2A) THAT CAN GUIDE THE DESIGN OF A MORE TARGETED AND EFFICIENT SAMPLING STRATEGY. WITH THE GOAL OF ADVANCING SNOW MODELS WE WOULD SAMPLE AT LOCATIONS ACROSS THE RANGE OF UNCERTAINTY: FROM NEAR COMPLETE MODEL AGREEMENT (I.E. CV CLOSER TO 0) TO MODERATE AND HIGH UNCERTAINTY ACROSS MODELS. THIS SAMPLING CAN THEN BE STRATIFIED ACROSS A GRADIENT OF LANDSCAPE FEATURES (E.G. FORESTED AREA FIGURE 2C) TO ENHANCE UNDERSTANDING OF PROCESS CONTROLS. THUS A COMBINED CLASSIFICATION ENABLES A SIMULTANEOUS EXAMINATION OF BOTH MODEL UNCERTAINTY AND LANDSCAPE CONTROLS (FIGURE 2B). THIS TYPE OF CLASSIFICATION MAP CAN HELP EVALUATE THE ADEQUACY OF SAMPLING STRATEGIES IN PAST CAMPAIGNS AND CAN INFORM SITE SELECTION FOR FUTURE FIELD CAMPAIGNS (FIGURE 2B INSET). THIS GENERAL APPROACH CAN BE USED IN SNOWEX BOTH WITH AND WITHOUT ASSIMILATION OF REMOTELY SENSED SNOW DEPTH OBSERVATIONS AND OTHER RELEVANT VARIABLES., G5132276322, G6425854980
OpenAlex ID: W4413265872