Relations between Soil Moisture and Satellite Vegetation Indices in the U.S. Corn Belt

Abstract

Satellite-derived vegetation indices extracted over locations representative of midwestern U.S. cropland and forest for the period 1990–94 are analyzed to determine the sensitivity of the indices to neutron probe soil moisture measurements of the Illinois Climate Network (ICN). The deseasoned (i.e., departures from multiyear mean annual cycle) soil moisture measurements are shown to be weakly correlated with the deseasoned full resolution (1 km × 1 km) normalized difference vegetation index (NDVI) and fractional vegetation cover (FVC) data over both land cover types. The association, measured by the Pearson-moment-correlation coefficient, is stronger over forest than over cropland during the growing season (April–September). The correlations improve successively when the NDVI and FVC pixel data are aggregated to 3 km × 3 km, 5 km × 5 km, and 7 km × 7 km areas. The improved correlations are partly explained by the reduction in satellite navigation errors as spatial aggregation occurs, as well as the apparent scale dependence of the NDVI–soil moisture association. Similarly, stronger relations are obtained with soil moisture data that are lagged by up to 8 weeks with respect to the vegetation indices, implying that soil moisture may be a useful predictor of warm season satellite-derived vegetation conditions. This study suggests that a “long-term” memory of several weeks is present in the near-surface hydrological characteristics, especially soil water content, of the Midwest Corn Belt. The memory is integrated into the satellite vegetation indices and may be useful for predicting crop yield estimates and surface temperature anomalies.