Hardware Acceleration of Simulations of Extreme Weather Events

这是对日本京都大学防灾研究所进行为期两个月的研究访问的提议，与大气和水圈灾害司的Tetsuya Takemi教授就极端天气事件模拟的硬件加速进行合作。硬件加速意味着使用特殊的硬件——在我们的例子中是通用图形处理单元（gpgpu）和现场可编程门阵列（fpga）来加速计算任务。特别是，我们的目标是减少使用天气研究与预报（WRF）模型的模型的模拟时间，WRF是下一代天气和气候模拟的领先模型。这将使模型能够以更高的精度运行，这对于预测极端天气事件至关重要。此次研究访问的主要目的是为包括日本和英国的研究小组在内的长期合作奠定基础，目的是加速完整的天气研究和预报（WRF）模型，并增加对FPGA加速的支持。

This is a proposal for a two-month research visit to the Disaster Prevention Research Institute of the University of Kyoto in Japan, to work with Prof. Tetsuya Takemi of the Atmospheric and Hydrospheric Disasters Division on hardware acceleration of simulations of extreme weather events. Hardware acceleration means the use of special hardware -- in our case General-Purpose Graphics Processing Units (GPGPUs) and Field-Programmable Gate Arrays (FPGAs) to speed up a computational task.In particular, our aim is to reduce the simulation times for models using the Weather Research and Forecasting (WRF) model, the leading next-generation model for weather and climate simulations. This will allow the models to be run at higher precision, which is essential in the prediction of extreme weather events.The main purpose of this research visit is to lay the groundwork for a long-term collaboration including research groups in Japan and the UK, with the aim of accelerating the full Weather Research and Forecasting (WRF) model and adding support for FPGA acceleration.