H038-0009
Comparing SoilMERGE Root Zone Soil Moisture and IMERG Precipitation as Predictors of Vegetation Greenness in the Colorado River Basin, 2001-2019

Tuesday, 8 December 2020
Poster
Kevin McKee1,2, Linnea Saby3, Venkataraman (Venkat) Lakshmi4, Jonathan L Goodall3 and Lawrence E Band5, (1)Virginia Polytechnic Institute and State University, Statistics, Charlottesville, Virginia, United States, (2)Virginia Polytechnic Institute and State University, Statistics, Blacksburg, VA, United States, (3)University of Virginia, Charlottesville, VA, United States, (4)University of Virginia, Engineering Systems and Environment, Charlottesville, VA, United States, (5)University of Virginia, Environmental Sciences, Charlottesville, VA, United States
Abstract:
Lack of root zone soil moisture (RZSM) data has historically limited the accuracy of spatially-distributed vegetation forecasting. Newly available 40-cm depth RZSM estimations from SoilMERGE, derived from satellite observations and National Land Data Assimilation System model outputs, are promising to help overcome this limitation. We compare SoilMERGE RZSM estimates with precipitation as predictors of vegetation greenness using a distributed lagged correlation analysis in the semi-arid Colorado River Basin (CRB) over 18 years (2001-2019). MODerate Resolution Imaging Spectroradiometer (MODIS) Normalized Difference Vegetation Indices (NDVI) was used as a measure of vegetation greenness, and precipitation data was from the Integrated Multi-satellitE Retrievals for GPM (IMERG). Our results show that SoilMERGE was a better predictor of NDVI than precipitation, with higher lagged correlations and shorter lag times over most of the study area. Our results help reveal timing and sensitivity of NDVI to critical hydrologic variables and can serve as a reference for the improvement of vegetation models and forecasts of water supply.
<< Previous Abstract | Next Abstract >>