PARENT SESSION
Poster Session 14: GIS and Remote Sensing
Tuesday, August 9, 5:00 PM - 6:30 PM, Exhibit Hall 220 A-E, Level 2, Palais des congrès de Montréal

Estimation of actual evapotranspiration by using satellite-driven water stress index and potential evapotranspiration models in a rice paddy.

Ryu, Youngryel^*,1, Kang, Sinkyu¹, Kim , Seung¹, ¹ Seoul National University, Seoul, Korea (ROK)

ABSTRACT- Evapotranspiration (ET) is an essential component for understanding the terrestrial ecosystem productivity and hydrological cycle. Eddy covariance method directly measure ET; however, it can only provide point-wise data sparsely distributed over the world. Fusion with remote sensing and field data remains a challenging task to estimate ET at various spatial and temporal scales. In this study, estimation of actual evapotranspiration (AET) was calculated by multiplying potential evapotranspiration (PET) by soil moisture function induced by water stress index derived from MODIS near- and shortwave infrared data. Study site is the Haenam site, a typical farmland venue- low-lying crops mixed with scattered rice paddies. A few small forests, small hills and residential areas complement the site. To construct a soil moisture function, the water stress index was introduced from MODIS near- and shortwave infrared data. Soil water content derived from MODIS was well fitted with in situ data (r²=0.81). Two PET models including Penman equation and Priestly-Taylor equation were converted into AET by multiplying soil water function, and then each AET was validated with flux tower data. AET derived from Priestly-Taylor model predicted ET best with flux tower data (r²=0.83).

Key words: evapotranspiration, water stress index, potential evapotranspiration, MODIS