Multi-satellite-based water budget components in South Korea

Abstract

Interpreting and predicting variations of the water cycle are a significant concern given the emerging threat of climate change. Generically, hydrological components of the water cycle are routinely observed with ground-based measurements, yet it is difficult to measure their spatiotemporal variability. Remote sensing approach is recognized as one of the most promising tools to obtain continuous data over large areas, thereby offering the unique possibility to assess the complicated and non-local features of hydrological phenomena. To estimate water budget components using remote sensing, this research considers precipitation (P), evapotranspiration (ET), and the change in water storage (∆S) calculated from satellites (i.e., Communication, Ocean and Meteorological Satellite; COMS, and Gravity Recovery and Climate Experiment; GRACE) and the Global Land Data Assimilation System (GLDAS) model-based datasets in South Korea from April to December 2011. The P estimates from the COMS rainfall intensity (COMS RI), COMS CM (which employs conditional merging [CM] to improve the accuracy of COMS RI), and GLDAS were compared with the measured P values from the two flux towers on a monthly scale. These results showed that COMS CM and GLDAS are in reasonable agreement, and additionally, their correlation, bias, and root-mean-square errors are favorable compared to the original COMS RI. The ET estimation of GLDAS and COMS applied from the revised RS-PM method were compared which indicated reasonable agreement with the two flux tower measurements. The derived runoff from COMS RI, COMS CM, and GLDAS was evaluated with that of the flux towers. The statistical results indicated that COMS CM and GLDAS were slightly better than that of COMS RI. The spatial distribution of P from COMS CM and GLDAS indicated similar pattern with that of ground-based measurement with the exception of COMS RI. ET from COMS and GLDAS showed slightly analogous pattern. The spatial distribution of runoff from both COMS and GLDAS showed evidence of a seasonality, which mainly resulted from the seasonally varying effects of ET and P. This research shows that it is possible to conduct the analysis of COMS products for efficient water resource planning, monitoring, and water budget modeling.