Publications

Kumar, U; Rashmi; Srivastava, A; Kumari, N; Chatterjee, C; Raghuwanshi, NS (2023). Evaluation of Standardized MODIS-Terra Satellite-Derived Evapotranspiration Using Genetic Algorithm for Better Field Applicability in a Tropical River Basin. JOURNAL OF THE INDIAN SOCIETY OF REMOTE SENSING, 51(5), 1001-1012.

Abstract
Evapotranspiration (ET) estimation at different spatial and temporal scales with a paucity of climatic parameters in a river basin is becoming a challenging task. Accurate estimation of ET is necessary for efficient water resource management and improving water efficiency at the field scale. Therefore, this study attempts to indirectly estimate actual ET from version 006 of MODIS-Terra product (MOD16A2.006), Sentinel-2A and Variable infiltration capacity (VIC-3L) model using survey information collected from a traditional paddy field in Kangsabati river basin. Further, this study is undertaken to standardize raw MODIS-Terra ET product (MOD16A2.06) using a genetic-based algorithm for better field applicability at local condition. The MODIS-standardized ET and ET estimated using different methods along with raw MODIS-Terra ET product were evaluated against observed ET estimated using globally recommended FAO-56 Penman-Monteith (PM) equation coupled with a crop coefficient. MODIS-Terra ET estimates were standardized using a genetic-based algorithm to enhance the efficacy of MODIS-Terra ET (MODIS-raw ET) for better field applicability. The result revealed that the genetic-based algorithm (MODIS-standardized ET) improved significantly with the NSE and RMSE from approximately - 0.03 to 0.86 and 13.89 to 2.56 (mm/8 day). Of various ET models Sentinel-2A ET performed best followed by MODIS-standardized ET, VIC-3L ET and MODIS-raw ET with R-2 = 0.92, NSE = 0.89, RMSE = 1.89 (mm/8 day), R-2 = 0.88, NSE = 0.86, RMSE = 2.47 (mm/8 day), R-2 = 0.77, NSE = 0.76, RMSE = 3.02 (mm/8 day) and R-2 = 0.41, NSE = - 0.03, RMSE = 7.31 (mm/8 day), respectively. The result showed that Sentinel 2A and MODIS-standardized-based ET can be used under data scarce conditions for better field applicability and water management practices.

DOI:
10.1007/s12524-023-01675-3

ISSN:
0974-3006