Authors: Zack Leasor*, The Ohio State University, Chen Zhao, Synoptic Data, Luyu Liu, The Ohio State University, Steven Quiring, The Ohio State University
Topics: Climatology and Meteorology, Water Resources and Hydrology
Keywords: Soil moisture, drought, Remote sensing, downscale
Session Type: Virtual Paper
Start / End Time: 9:35 AM / 10:50 AM
Room: Virtual 40
Presentation File: No File Uploaded
Improved methods for visualizing drought at a fine resolution can help to bridge the gap between drought severity classifications and local impacts. Soil moisture data are critical for drought monitoring because they can provide reference measurements of meteorological and agricultural drought conditions. Recent advances in remote sensing technology have provided additional tools for monitoring near-surface soil moisture across the contiguous U.S. (CONUS). This research leverages remote sensing products to accurately downscale soil moisture and produce national maps of soil moisture at a fine resolution. Soil moisture data are standardized and disseminated according to the severity thresholds of the U.S. National Drought Monitor (USDM) to better communicate drought status. Remote sensing soil moisture data obtained from the NASA Soil Moisture Active Passive (SMAP) satellite mission are utilized in this study. To downscale these soil moisture data, ancillary variables such as precipitation, soil texture, vegetation, and other physiographic information are considered to downscale soil moisture to a 1-kilometer resolution. Random forest, a non-parametric machine learning modeling approach, is used to find the best relationship that can determine the optimal soil moisture value at any location. Results compare the downscaled products to direct in situ measurements and crop yield metrics to demonstrate the value of high-resolution drought products based on soil moisture.