Machine Learning Modeling of SWESARR and Lidar Data to Understand How Snow Water Equivalent Changes Spatially and Temporally

The Snow Water Equivalent Synthetic Aperture Radar and Radiometer (SWESARR) is a dual microwave instrument meant to fill in information gaps in the remote sensing data of Snow Water Equivalent (SWE). This work aims to improve and validate SWESARR measurements of SWE for areas with tree canopy using machine learning. This information is critical to NASA’s SnowEx mission in understanding the spatial and temporal variability of snow. SWE is an integral part of the climate system that affects many other climate-related processes, so an accurate understanding of how SWE is changing spatially and temporally with climate change is crucial for future water resource management. We have used unsupervised machine learning algorithms to help identify features most important in analyzing SWE spatially in areas with missing satellite data due to vegetation.

Furthermore, a supervised machine-learning algorithm, specifically a long short-term memory neural network, was used to establish a time series forecasting algorithm to predict changes in maximum temperature and SWE. Our aim is to investigate SWE spatially, including quantifying how SWE measurements with vegetation differ from SWE measurements without vegetation, as the results will inform predictions of how SWE will change in time using time-series forecasting. Other relevant datasets central to identifying parameters for predicting SWE and data validation have been from the ASO (Airborne Snow Observatory), NSIDC (National Snow and Ice Data Center), and ground snow pit observations taken by the SnowEx fieldwork team. This work is a result of NASA’s SnowEx team based out of Goddard Space Flight Center within NASA GISS’s Climate Change Research Initiative.