Use of Remote Sensing Data for Estimation of Aman Rice Yield

Abstract

Weather related crop losses have always been a concern for farmers, governments, traders and policy makers for the purpose of balanced food supplies, demands, trade, and distribution of aid to nations in need. This paper discusses the utility of Advanced Very High Resolution Radiometer (AVHRR)-based vegetation health (VH) indices as proxies for modelling inter annual variation in Aman rice (AR) yield in Bangladesh and for early estimation. We compare annual local and hybrid AR yield with VH Indices computed for each week during 1991-2005. A strong correlation was found between AR yield and VH during the period of AR development that occurs during one/two months in advance of harvest (early October to early November). Stepwise principal components regression (PCR) was used to construct a model to estimate yield as a function of critical-period VH indices. The model reduced the yield prediction error variance by 97% and 92% compared with a prediction of average local Aman rice (LAR) and hybrid Aman rice (HAR) yield for each year respec- tively.