Title: Inter-comparison and Validation of Remote Sensing Satellite based Soil Moisture.

Speaker: Stephanie Marquez, NOAA EPP/MSI CESSRST-II Fellow at UTEP

Date: Ocotber 24, 20223

Time: 1:30 PM ET

Venue: Virtual

Meeting Link : meet.google.com/spo-vxqc-nmb

Phone Number
(‪US‬)‪+1 219-321-0478‬
PIN: ‪651 945 406#‬

Learn more about the speaker

Abstract:

Inter-comparison and Validation of Remote Sensing Satellite based Soil Moisture.: This study explores the potential of using a random forest, gradient boosting and support vector machine model to predict soil moisture levels by incorporating both meteorological and biogeophysical data. The study utilized meteorological data such as temperature, albedo, and precipitation, and biogeophysical data including soil type, land cover, slope, elevation, and normalized differenced vegetation index (NDVI) from remote sensing sources. Additionally, in-situ soil moisture measurements from a Campbell Scientific Hydrosense II-12 sensor were incorporated as input parameters to train and test the model. The dataset was collected from a soil moisture sensor network installed in Jornada Experimental Range, New Mexico during the summer of 2022. The results indicate that the gradient boosting model accurately predicts soil moisture levels, with an NSE score of 0.62, demonstrating the importance of combining remote sensing data with ground-based observations. This study showcases the potential of machine learning algorithms, particularly random forest models, to accurately predict future soil moisture levels, which could enhance the accuracy of satellite data and weather predictions.