Advancing Terrestrial Snow Depth Monitoring with Machine Learning and L-band InSAR Data - A Case Study Using NASA’s SnowEx 2017 Data

Aug 5, 2026

Accurate snow depth observations are essential for hydrologic science, water management, and climate modeling, yet large-scale monitoring remains constrained by spatial coverage, revisit frequency, and cost. This study evaluates whether L-band InSAR can support snow depth estimation through machine learning. Using 3 m resolution L-band InSAR over Grand Mesa, Colorado, it compares XGBoost, ExtraTrees, and neural networks on open, vegetated, and combined datasets with RMSE, MBE, and R² metrics. XGBoost performs best and validates against in situ measurements with RMSE around 16 cm, indicating cloud-penetrating, wide-area potential for future NISAR-enabled mapping.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages19

Reading Time48 min

AccessFree to download

Introduction

Challenges in Large-Scale Snow Depth Monitoring

Limitations of In Situ and Airborne Observations

Motivation for L-band InSAR and Machine Learning

TYPE Original Research PUBLISHED 22 January 2025 DOI 10.3389/frsen.2024.1481848 OPEN ACCESS EDITED BY Kun-Shan Chen, Nanjing University, China REVIEWED BY Mukesh Gupta, Université du Québec à Rimouski, Canada Divyesh Varade, Indian Institute of Technology Jammu, India Surendar Manickam, VIT University, India *CORRESPONDENCE Ibrahim Olalekan Alabi, [ibrahimolalekana@u.boisestate.edu](ibrahimolalekana@u.boisestate.edu) RECEIVED 16 August 2024 ACCEPTED 24 December 2024 PUBLISHED 22 January 2025 CITATION Alabi IO, Marshall H-P, Mead J and Trujillo E (2025) Advancing terrestrial snow depth monitoring with machine learning and L-band InSAR data: a case study using NASA’s SnowEx 2017 data. Front. Remote Sens. 5:1481848 . doi: 10.3389/frsen.2024.1481848 COPYRIGHT © 2025 Alabi, Marshall, Mead and Trujillo. This isan open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Advancing terrestrial snow depth monitoring with machine learning and L-band InSAR data: a case study using NASA’s SnowEx 2017 data Ibrahim Olalekan Alabi 1,2*, Hans-Peter Marshall 2, Jodi Mead 3 and Ernesto Trujillo 2 1Computing PhD Program, Boise State University, Boise, ID, United States, 2Department of Geoscience, Boise State University, Boise, ID, United States, 3Department of Mathematics, Boise State University, Boise, ID, United States Current terrestrial snow depth mapping from space faces challenges in spatial coverage, revisit frequency, and cost. Airborne lidar, although precise, incurs high costs and has limited geographical coverage, thereby necessitating the exploration of alternative, cost-effective methodologies for snow depth estimation. The forthcoming NASA-ISRO Synthetic Aperture Radar (NISAR) mission, with its 12-day global revisit cycle and 1.25 GHz L-band frequency, introduces a promising avenue for cost-effective, large-scale snow depth and snow water equivalent (SWE) estimation using L-band Interferometric SAR (InSAR) capabilities. This study demonstrates InSAR’s potential for snow depth estimation via machine learning. Using 3 m resolution L-band InSAR products over Grand Mesa, Colorado, we compared the performance of three machine learning approaches (XGBoost, ExtraTrees, and Neural Networks) across open, vegetated, and the combined (open + vegetated) datasets using Root Mean Square Error (RMSE), Mean Bias Error (MBE), and R2 metrics. XGBoost emerged asthe superior model, with RMSE values of 9.85 cm, 10.46 cm, and 9.88 cm for open, vegetated, and combined regions, respectively. Validation against in situ snow depth measurements resulted in an RMSE of approximately 16 cm, similar to in situ validation of the airborne lidar. Our ﬁndings indicate that L-band InSAR, with its ability to penetrate clouds and cover extensive areas, coupled with machine learning, holds promise for enhancing snow depth estimation. This approach, especially with the upcoming NISAR launch, may enable highresolution (~10 m) snow depth mapping over extensive areas, provided suitable training data are available, offering a cost-effective approach for snow monitoring. The code and data used in this work are available at [https://github](https://github). com/cryogars/uavsar-lidar-ml-project. KEYWORDS snow depth, InSAR, machine learning, NISAR, remote sensing Frontiers in Remote Sensing 01 [frontiersin.org](frontiersin.org) 1 Introduction Accurately measuring snow depth is critical for applications in hydrologic science, water resource management, and climate modeling (Lievens et al., 2022) . Seasonal snowpacks act as natural reservoirs, s

FAQs

Why is accurate snow depth monitoring important?

It underpins hydrologic science, water resource management, and climate modeling by tracking snow accumulation and melt and supporting understanding of snowpack storage and release.

What problem does this study address?

Current space-based snow depth mapping is limited by coverage, revisit frequency, and cost, and airborne lidar—while precise—has high costs and limited geographical coverage.

How is L-band InSAR used in the study?

3 m resolution L-band InSAR products over Grand Mesa are used with machine learning models (XGBoost, ExtraTrees, and neural networks), and performance is evaluated against in situ snow depth measurements using RMSE, MBE, and R².

Aug 5, 2026

Upload by Theodore