Integrated Forecasting System of Flood Inundation and Agricultural Damage in the Lower Mekong Basin

洪水预报和相关的实时损失评估具有挑战性，目前在湄公河下游地区无法提供。本文主要研究了湄公河下游洪水淹没实时系统的开发。特别是，这项研究的重点是开发一个系统，以同时评估洪水灾害和LMB的损失。开发的系统还可用于气候变化影响下的极端洪水事件未来预测的农业损失风险评估。首先，我们收集和评估了观测的水文数据。利用大型集合气候资料(D4PDF)、MRI-AGCM3.2和CMIP6 GCM数据集，在LMB进行了气候变化影响下的洪水淹没模拟。对洪水灾害及其相关损失进行了评估。此外，通过NICAM-LETKF资料同化，利用来自GSMaPxNEXRA的数值天气预报数据集对柬埔寨普雷克桑特河流域的洪水预报进行了初步评估。最后，结合机器学习技术，对洪水预报进行了评估和改进，以提高预报的性能和精度。水文模型的混合方法。

Flood forecasting and the related damage assessment in real-time are challenging and currently not available in the Lower Mekong region. This study focuses on the development of a real-time system of flood inundation in the Lower Mekong Basin (LMB). In particular, this study focuses on the developing a system to assess flood hazards and damages in the LMB simultaneously. The developed system can also be used for risk assessment of agricultural damages for future projections of extreme flood events under climate change effects.Firstly, we have collected and evaluated the observed hydrological data. Flood inundation simulation was conducted in the LMB under the effect of climate change using a large ensemble climate data (d4PDF), MRI-AGCM3.2, and CMIP6 GCM datasets. Flood hazards and its related damages were assessed. Moreover, flood forecasting was primarily evaluated using numerical weather prediction dataset from GSMaPxNEXRA through NICAM-LETKF data assimilation in the Prek Thnot River Basin, Cambodia. Finally, the evaluation and improvement of flood forecasting has been done to improve its performance and accuracy by combining with machine learning technique. Hybrid approach of hydrological model.