The economy and livelihood of the State of West Bengal in India are mainly dependent on agriculture. Rice is one of the major crops of this state, and it contributes a significant proportion to the rice production of India. The present study deals with the prediction of Kharif rice over Gangetic West Bengal using forecast products from IITM-IMD extended range prediction (ERP) system. It is a multi-model ensemble prediction system that comprises 16 different ensemble members obtained from the Climate Forecast System (CFSv2) and the stand-alone atmospheric component of CFSv2 (i.e., GFSv2). The performance of ERP rainfall forecast has been assessed over a relatively larger domain covering Indian landmass before its utilization in crop yield prediction. Satisfactory skills, e.g., higher correlation and lower normalized root mean squared error (nRMSE), have been found in ERP rainfall forecast during the first 3 weeks for most of the initial conditions (ICs). Next, bias-corrected ERP weekly forecast data of incoming solar radiation, rainfall, and maximum and minimum temperatures have been incorporated into a process-based crop model (CERES-rice) available in the Decision Support System for Agro-technology Transfer (DSSAT). ERP-driven crop model has performed better to reproduce inter-annual variability of observed rice yield compared to yield simulated using crop model driven by climatology alone. Also, the ERP-driven model has been able to capture the below- and above-normal yield categories relatively better than the climatology-driven model. Hence, the incorporation of ERP in crop models may provide value-added prediction, which will be helpful for the stakeholders and decision-makers.

印度西孟加拉邦的经济和生计主要依赖农业。水稻是该邦的主要农作物之一，对印度的水稻产量贡献很大。本研究使用来自 IITM-IMD 扩展范围预测 (ERP) 系统的预测产品来预测恒河西孟加拉邦的哈里夫水稻。它是一个多模式集合预测系统，包括从气候预报系统 (CFSv2) 和 CFSv2 的独立大气分量（即 GFSv2）获得的 16 个不同集合成员。在将 ERP 降雨预测用于作物产量预测之前，已经在覆盖印度陆地的相对较大的领域内对其性能进行了评估。令人满意的技能，例如较高的相关性和较低的归一化均方根误差 (nRMSE)，在大多数初始条件 (IC) 的前 3 周内，在 ERP 降雨预报中已经发现。接下来，已将入射太阳辐射、降雨量以及最高和最低温度的偏差校正 ERP 每周预测数据合并到农业技术转让决策支持系统 (DSSAT) 中提供的基于过程的作物模型 (CERES-水稻) 中. 与使用单独由气候学驱动的作物模型模拟的产量相比，ERP 驱动的作物模型在重现观察到的水稻产量的年际变化方面表现更好。此外，ERP 驱动的模型能够比气候学驱动的模型更好地捕捉低于和高于正常水平的产量类别。因此，将 ERP 纳入作物模型可能会提供增值预测，这将有助于利益相关者和决策者。