Hybrid Bias Correction of IMERG Late Run V07 over Indonesia (2001-2025): Input Data, Land-Sea Masks, and Corrected Products

This deposit contains the full input, mask, and output data bundle for the hybrid statistical-deep learning bias correction of daily IMERG Late Run V07 precipitation over Indonesia (95-141 deg E, -11 to 6 deg N) for the period 2001-2025. The bundle accompanies the companion manuscript (Istanto, in preparation) and the open-source code at https://github.com/bennyistanto/hybrid-bias-correction. The framework combines three correction stages applied per 10-day window (dekad). Linear Scaling matches the long-term mean. Empirical Quantile Mapping with a Generalized Pareto Distribution tail (80th-percentile threshold) reshapes the full daily distribution. A lightweight two-layer convolutional network refines the spatial residual on extreme days. A station-density confidence mask, built from the BMKG gauge network, modulates the CNN influence so that the correction degrades gracefully to LSEQM in gauge-sparse areas. The input folder contains IMERG Late Run V07 (idn_imergl.nc4, 2001-01-01 to 2025-12-31) IMERG Final Run V07 baseline (idn_imergf.nc4, 2001-01-01 to 2025-09-30) CPC-UNI regridded to 0.1 deg (idn_cpcuni.nc4) and at its native 0.5 deg resolution (idn_cpcuni_native05.nc4) BMKG station location and daily observation CSVs. The mask folder contains the Indonesia land-sea mask at 0.1 deg together with the area-of-interest subsets used for clipping (`idn_subset.nc`). The output folder contains Per-method corrected products (corrected_ls, corrected_lseqm, corrected_lseqmdl) Trained CNN models per dekad (trained_models) Per-method performance metrics (metrics_ls, metrics_lseqm, metrics_lseqmdl) Continuous Quality Index outputs (quality_ls, quality_lseqm, quality_lseqmdl) Station-density confidence mask Independent BMKG station validation CSVs Figures All gridded files are CF-1.8 compliant NetCDF (float32, zlib compression) on a common 0.1 deg lat-lon grid. The CPC-UNI 0.5 deg file is kept on the original gauge-analysis grid and is used by the parameter-fitting path that avoids the 5x5 nearest-neighbour block boundary. Global attributes record the framework version (v2026.01), the source IMERG and CPC files, and the sensitivity parameters used in the run (blend_alpha = 0.70, gpd_threshold_percentile = 80, station_density saturation_count = 3). These three parameters were not formally optimised and were chosen by inspection. Users adapting the framework to another region should treat them as sensitivity parameters rather than fixed defaults. To reproduce the outputs from the inputs, clone the code repository, point config.yml at the directories in this bundle, and run notebooks 02 through 06 in sequence. A smaller Bali example (about 11 MB) ships with the code repository so users can validate the toolchain on a free Google Colab session before downloading the full Indonesia bundle. Full documentation is available at https://bennyistanto.github.io/hybrid-bias-correction. If you use these data, please cite both this Zenodo deposit and the companion manuscript (DOI to be added on publication). NASA GPM IMERG V07 data are courtesy of the NASA Goddard Earth Sciences Data and Information Services Center. CPC Unified gauge-based daily precipitation data are provided by the NOAA Physical Sciences Laboratory. BMKG station observations are provided by the Indonesian Agency for Meteorology, Climatology and Geophysics.本数据集包含2001-2025年期间，针对印度尼西亚区域（东经95°-141°，北纬-11°至6°）的每日IMERG延迟运行版V07（IMERG Late Run V07）降水产品进行混合统计-深度学习偏差校正所需的完整输入、掩码与输出数据包。本数据包配套待发表的合著论文（Istanto，待刊）以及开源代码库：https://github.com/bennyistanto/hybrid-bias-correction。 该校正框架针对每10天窗口（旬）设置三级校正流程：首先通过线性缩放匹配长期均值；其次采用带有广义帕累托分布（Generalized Pareto Distribution, GPD）尾部（80百分位阈值）的经验分位数映射重塑全日降水分布；最后借助轻量级双层卷积神经网络（Convolutional Neural Network, CNN）修正极端日期下的空间残差。此外，基于印度尼西亚气象气候与地球物理局（Badan Meteorologi, Klimatologi dan Geofisika, BMKG）雨量站网络构建的站网密度置信掩码，可调节卷积神经网络的校正权重，在站网稀疏区域自动降级为LSEQM（线性缩放-经验分位数映射）校正方法，确保校正效果平稳过渡。 输入文件夹包含以下数据： 1. IMERG延迟运行版V07产品（idn_imergl.nc4，时间范围2001-01-01至2025-12-31） 2. IMERG最终运行版V07基准产品（idn_imergf.nc4，时间范围2001-01-01至2025-09-30） 3. 重采样至0.1°分辨率的CPC-UNI降水数据（idn_cpcuni.nc4），以及保留原始0.5°分辨率的CPC-UNI数据（idn_cpcuni_native05.nc4） 4. BMKG站点位置与每日观测CSV文件。 掩码文件夹包含0.1°分辨率的印度尼西亚海陆掩码，以及用于区域裁剪的研究区域子集文件（idn_subset.nc）。 输出文件夹包含以下内容： 1. 各方法校正后的降水产品（corrected_ls、corrected_lseqm、corrected_lseqmdl） 2. 各旬训练得到的卷积神经网络模型（trained_models） 3. 各方法的性能评估指标（metrics_ls、metrics_lseqm、metrics_lseqmdl） 4. 连续质量指数（Continuous Quality Index）输出结果（quality_ls、quality_lseqm、quality_lseqmdl） 5. 站网密度置信掩码 6. 独立BMKG站点验证CSV文件 7. 可视化图表 所有网格化文件均采用符合CF-1.8规范的NetCDF格式（float32数据类型，zlib压缩），统一使用0.1°经纬度网格。其中0.5°分辨率的原始CPC-UNI数据保留其原生雨量分析网格，用于规避5×5最近邻块边界效应的参数拟合流程。全局属性记录了校正框架版本（v2026.01）、原始IMERG与CPC数据源，以及本次运行使用的敏感性参数：blend_alpha=0.70、GPD阈值百分位=80、站网密度饱和计数=3。上述三项参数未经过正式优化，仅通过人工检验确定，用户若将本框架适配至其他区域，应将其视为敏感性参数而非固定默认值。 若需基于输入数据复现输出结果，请克隆代码仓库，将config.yml文件指向本数据包中的目录，并依次运行Notebook 02至06。代码仓库附带一个小型巴厘岛示例数据集（约11 MB），用户可在免费Google Colab环境中先验证工具链完整性，再下载完整印度尼西亚数据包。完整文档可访问：https://bennyistanto.github.io/hybrid-bias-correction。若使用本数据集，请同时引用本Zenodo存档与合著论文（出版后将补充DOI）。 NASA GPM IMERG V07数据由NASA戈达德地球科学数据与信息服务中心提供。CPC统一雨量站逐日降水数据由NOAA物理科学实验室提供。BMKG站点观测数据由印度尼西亚气象、气候与地球物理局提供。