B43D-0595
Validation the Product of Evapotranspiration (MOD16) With Flux Tower Data to Rice Paddy in Southern Brazil

Thursday, 17 December 2015
Poster Hall (Moscone South)
Vanessa De Arruda Souza1, Debora Regina Roberti2, Rita Marques Alves1, Marcelo Diaz2 and Gisele Cristina Rubert2, (1)UFRGS, CEPSRM, Porto Alegre, Brazil, (2)UFSM Federal University of Santa Maria, Santa Maria, Brazil
Abstract:
The process of Evapotranspiration (ET) is second the major component of the hydrological cycle, connected directly with the amount of water vapor in the atmosphere. Rice is one of the most produced grains in the world, mainly flooded-irrigated. However, determine the ET for large areas of irrigated rice is a limiting factor due to the high cost of equipment and monitoring of the site. The utilization of data from remote sensing is an important tool for monitoring large areas and are of fundamental importance for the description of the behavior the ET estimate. The aim of this study is to evaluate the ET (MOD16) using Eddy Covariance flux tower-derived ET for a rice paddy in the Paraíso do Sul - Brazil. The data used in this study cover the period from July 2003 to July 2004, with the cultivated rice during the months from November to April. The comparison between ET-MODIS the central pixel with ET observed showed statistical values of RMSE = 8.83, PBIAS -25.8 and R2 = 0.5. For the pixels with same coverage vegetation were also extracted the average values of ET-MODIS and compared with data flux tower, resulting in RSME = 7.68, = -31.3 PBIAS and R2 = 0.63. Although the average of pixels with the same vegetation present greater underestimation of ET compared the observed, showed lower RMSE and higher R2. Much of this difference is associated to the period from July 2003 to January 2004, wherein the months January to April the ET-MODIS central pixel overestimates the observed data. However, in general mode we observed that MOD16 product shows similar behavior ET observed.