Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #331869
Title: Investigating water use over the Choptank River Watershed using a multi-satellite data fusion approach

Author
item SUN, L. - Collaborator
item Anderson, Martha
item Gao, Feng
item HAIN, C. - University Of Maryland
item Alfieri, Joseph
item SHARIFI, AMIR - University Of Maryland
item Yang, Yun
item YANG, YANG - Collaborator
item Dulaney, Wayne

Submitted to: American Meteorological Society of the Conference on Hydrology Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: 9/5/2016
Publication Date: 1/22/2017
Citation: Sun, L., Anderson, M.C., Gao, F.N., Hain, C., Alfieri, J.G., Sharifi, A., Yang, Y., Yang, Y., Dulaney, W.P. 2017. Investigating water use over the Choptank River Watershed using a multi-satellite data fusion approach. American Meteorological Society of the Conference on Hydrology Proceedings. 2017 CD-ROM.

Interpretive Summary:

Technical Abstract: Satellite remote sensing technologies have been widely used to map spatiotemporal variability in consumptive water use (or evapotranspiration; ET) for agricultural water management applications. However, current satellite-based sensors with the high spatial resolution required to map ET at sub-field scales (<100 m) typically provide infrequent temporal sampling (bi-weekly), while satellites with hourly or daily revisit have too coarse a resolution (>1 km) to see individual fields. To overcome these limitations, our group has developed a multi-sensor satellite data fusion methodology (STARFM: Spatial and Temporal Adaptive Reflective Fusion Model), combined with a multi-scale ET retrieval algorithm (DisALEXI: Disaggregated Atmosphere-Land Exchange Inverse model). This system combines ET maps generated with the Geostationary Environmental Operational Satellites (GOES; 4-km spatial resolution, hourly temporal sampling), the Moderate Resolution Imaging Spectroradiometer (MODIS) data (1-km resolution, daily acquisition) and the Landsat satellite (sharpened to 30-m resolution, 16-day acquisition) to create geospatial water use datasets with both high spatial (30-m) and temporal (daily) detail. In this study, the ET fusion system was applied to a 30 by 30 km region including the Choptank River watershed located on the Eastern Shore of Maryland, USA - a focus watershed within the Lower Chesapeake Bay Long-Term Agricultural Research (LCB LTAR) site. Evaluations using in-situ flux tower measurements indicate that ET estimates directly retrieved on Landsat overpass dates have high accuracy with bias of 0.17 mm, Root Mean Square Error (RMSE) of 1.18 mm, and relative error (RE) of 9%. The fused daily ET, using MODIS to inform interpolation between Landsat dates, has reasonable errors of 18% - an improvement from 27% errors using standard interpolation techniques. Maps of cumulative annual ET for 2013 and 2014 showed similar spatial patterns, with high evaporative water use rates in the west close to the river, with large contributions from riparian and wetland systems, and relative low ET in the drier eastern portion of the domain. An accounting of annual water consumption by different landcover types was performed, showing reasonable distributions of water use. Developed areas showed lowest evaporative water loss, with 765 mm total annual ET on average. Wetlands have the highest consumption rates, with 1316 mm annually. Forest and crops annually consume 1137 and 1054 mm, respectively. Crop transpiration and soil evaporation in terms of crop types were extracted, and agree well with crop phenology at spatiotemporal scale. Additionally, irrigation event right happened at the shortage of rainfall period could also be captured by our fusion program. Efforts are underway to integrate these details water use datasets into hydrologic modeling to improve assessments of water quality and best management practices within the Chesapeake Bay.