Comparing Methods of Identifying Irrigation Using High Resolution SMAP Soil Moisture

With the wide variety of methods using satellite soil moisture (SM) observations to identify irrigation, our study aims to recreate several of the methods with a newly developed soil moisture product that uses a downscaling algorithm to produce a 400-m resolution soil moisture product from the native 36-km Soil Moisture Active Passive (SMAP) soil moisture data (Fang et al. 2025). From this product, we can estimate deeper soil moisture (20-cm and 50-cm from the original 5-cm depth). The objectives of this project are to see which of the methods perform best in the state of Colorado, and if performance differs with crop type, irrigation type, precipitation levels, and soil moisture depth as compared with the irrigation/crop type spatiotemporal field data available through Colorado’s Decision Support System (CDSS).

Our methods for identifying irrigation include:

• Summation of SM over the growth period: Because we are studying a relatively small area, we assume that all soil is receiving an approximately equal amount of moisture from precipitation. We deduce that regions with higher SM than those around them receiving additional moisture from irrigation.
• High SM despite low precipitation: If SM is detected despite there being a lack of precipitation for 4 days or more, one can assume that detected moisture came from irrigation (Lawston et al., 2017; Shellito et al., 2016).
• Changes in mean absolute deviation (MAD): MAD is used to understand the variability of SM within each pixel- because irrigation causes frequent and significant changes in SM, higher variability is a sign of irrigation (Jalilvand et al., 2021).
• Isolating irrigation signals: To isolate irrigation signals, this method incorporates a soil water balance model that accounts for vertical fluxes such as evapotranspiration and drainage, which influence soil moisture changes independently of irrigation (Zappa et al., 2021).
• Increases in NDVI after irrigation events: The normalized difference vegetation index (NDVI) measures plant greeness and vigor and is elevated for healthy plants. Previous studies have found that irrigated crops have an NDVI value greater than 0.8, while non irrigated vegetation is below 0.75 (Brown & Pervez, 2014; Ibrahim et al., 2023).