Satellite remote sensing of crop water use across the Missouri River Basin for 1986–2018 period

Abstract

Understanding historical crop water use (CWU) dynamics is important to improve land and water management. In this study, well-validated (coefficient of determination = 0.91, percent bias = 4%, and percent root mean square error = 11.8%) Landsat-based actual evapotranspiration (ETa) time-series estimations were used to (1) assess summer season CWU (CWU-Su) dynamics, (2) investigate CWU-Su trends over the study period (1986–2018; 33 years) at the regional- and pixel-scales, and (3) attribute CWU-Su driving factors across Missouri River Basin. Spatial variability of the ETa estimations along with the observed bimodal probability density distribution of ETa highlighted a strong relation between land cover and water uses across the basin. The bimodal distribution of ETa also indicated the presence of two major landcovers in the basin. The drier foothill regions in northwestern Missouri River Basin, dominated by grassland/shrubland, showed lower ETa (< 500 mm/year), whereas cropland dominated regions in lower semi-humid basin and forested subbasins exhibited higher ETa (> 600 mm/year). The CWU-Su anomalies revealed the vulnerability of the basin to year-to-year weather conditions. The CWU-Su trend analysis revealed a significant positive trend (p < 0.1) at the regional-scale affecting 30% of basin’s cropland pixels. The cropland pixels under positive CWU-Su trend were found to be clustered in the eastern and central Missouri River Basin as a result of the combined effect of increased crop production area, increased crop yields, crop practice shifts to higher biomass crops, and increased irrigated land. The effect of improved irrigation and water management practices on reducing CWU-Su was observed in western Missouri River Basin, which had a stable major crop throughout the study period. Overall, the study highlights the usefulness of Landsat imagery and remote sensing-based ETa modeling approaches in generating historical time-series ETa maps over a wide range of elevation, vegetation, and climate.