Publications

Abstract : The present study is conducted to assess the influence of climate model biases in the predictions of yield and water requirement of cassava in one of the major cassava growing regions in India. Simple linear bias correction methods are used for temperature, and non-linear corrections are used for other meteorological variables. The WOFOST and CROPWAT models are used to predict the crop yield and water requirement of cassava using the scenarios of 2030, 2050, and 2070 for the representative concentration pathway 4.5 derived from the Long Ashton Research Station Weather Generator (LARS-WG). The percentage change in crop yield predictions with and without bias corrections of meteorological variables ranges from 7.6 to 10.8%, 1.6 to 5.4%, and − 3.0 to 4.0% respectively for 2030, 2050, and 2070. The bias corrections made an increment in the gross irrigation requirements of cassava with 16.5, 17.8, and 16.0% in 2030, 2050, and 2070 respectively, compared to the values without bias corrections. The outcome of this study indicates that raw meteorological variables directly from the climate models result over-/underestimation of yield and irrigation requirements of cassava, and the bias corrections help to issue reliable crop yield predictions. Results show zero yield reductions of cassava until 2050, and beyond that, there can be reductions in the crop yield. The gross irrigation requirements of cassava increase in the future to achieve higher productivity. However, this study needs to extend to other major growing regions in India to derive a general conclusion.