SYSTEM AND METHOD FOR FORECASTING PRECIPITAION BASED ON ATMOSPHERIC NUMERICAL MODELS USING GENETIC ALGORITHM

摘要

PURPOSE: A genetic algorithm interface system applicable for a numerical model is provided to improve the accuracy of rainfall forecast by being applied for a weather forecasting model. CONSTITUTION: A system for forecasting rainfall includes a data inputting part(110), a model inputting part(120), a first parameter generating part(130), an ETS storing part(140), a second parameter generating part(150), and a convergence analyzing part(160). Rainfall data is input into the data inputting part. A forecasting model based on convective parameterization is inputted into the model inputting part. The first parameter generating part generates arbitrary parameters based on the input rainfall data. The ETS storing part stores rainfall forecast by implementing the forecasting model based on the parameters. The second parameter generating part evaluates the stored ETS based on a genetic algorithm and generates enhanced second parameters. The convergence analyzing part implements the forecasting model based on the second parameters and analyzes the convergence of parameter values.