Data underlying the research on: Downscaling precipitation

This document is a description of all files and scripts used for the MSc thesis research of Avelon Gerritsma:Super-resolution neural networks: a strategy to downscale model output dataThis document indicates the order in which the files were created/used, files are indicated with /. In the files are desciptions to explain what is done.The precipitation maps are downloaded from https://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-20th-century-using-surface-observations-onlyIntermediate files (TFRecord files that are rescaled, after step 3, can be found on 4TU name: Data underlying the research on: Downscaling precipitation)Scripts can be found on github<br><br>

本文件为阿维隆·格里茨玛（Avelon Gerritsma）硕士论文研究所用全部文件与脚本的说明文档，其研究主题为「超分辨率神经网络：一种模拟输出数据降尺度的策略」。本文件列明了各文件的创建与使用顺序，文件以「/」作为标识。各文件内部均附带有操作说明，用以阐释其具体功能与执行流程。本次研究所用的降水场数据，下载自欧洲中期天气预报中心（European Centre for Medium-Range Weather Forecasts, ECMWF）的「仅使用地面观测资料的20世纪再分析数据集」，获取链接为：https://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-20th-century-using-surface-observations-only。步骤3后生成的经过重缩放的TFRecord格式中间文件，可于4TU平台获取，对应数据集名称为「降尺度降水研究相关数据」（Data underlying the research on: Downscaling precipitation）。相关研究脚本可在GitHub平台获取。