Application of SEBAL for rice water consumption and productivity estimation through integrating remote sensing and census data in the Sanjiang Plain, China.

摘要

Crop consumptive water use and productivity are key elements to understanding water management performance. This study mapped rice water consumption, yield, and water productivity (WP) in 23 counties from May to October 2006 in the Sanjiang Plain by combining remotely sensed images, agricultural census and meteorological data. The actual evapotranspiration (ET_a) was calculated using the Surface Energy Balance Algorithm for Land (SEBAL) model using Moderate Resolution Imaging Spectroradiometer (MODIS) products in 12 snapshots in the growing season. The accuracy evaluation of daily ET_a for the SEBAL indicates a relative error ranging from -11.3% to 27.6% compared with the Eddy Covariance system (10.5% on average). The seasonal SEBAL estimated ET_a was comparable to that from the ground observation with a relative error of 8.9%. The results indicated that the ET_a retrieval method based on remote sensing techniques could satisfy the requirements of regional ET_a estimation. The calculated average and maximum water consumption for rice were 446 and 494 mm, respectively. The WP ranged from 1.21 kg.m-3 to 1.94 kg.m-3 in the Sanjiang Plain. A close linear relationship between WP and rice yield was observed (R2=0.84), which indicated that the spatial pattern of WP was similar to that of yield, namely high WP associated with the higher yield in the Sanjiang Plain. A comparative analysis of ET_a, rice yield and WP maps indicated greater space for improvement of water use efficiency in the Sanjiang Plain. By improving the irrigation schedule, the WP would be increased, which would in turn significantly reduce the irrigation water use and alleviate the water shortage in the Sanjiang Plain.