Data from: Development of a gridded meteorological data set over the Java island, Indonesia 1985-2014

We developed a gridded daily meteorology data set consisting of precipitation, minimum temperature and maximum temperature over the Java Island, Indonesia at 0.125 x 0.125 (∼14 km) resolution spanning 30 years from 1985 – 2014. Importantly, this data set represents a marked improvement from existing gridded data sets over Java as it has much higher spatial resolution, and is derived exclusively from ground-based observations unlike other satellite or reanalysis-based products. Gridding was conducted via Inverse Distance Weighting, IDW (radius, R, of 25 km and power of influence, , of 3 as optimal parameters) restricted to only those stations including at least 3650 days (~10 years) of valid data. A gap-infilling procedure was performed prior to the gridding process. We performed cross-validation on both the gap-infilling process and gridding process. It showed that spatial percent bias for gap-infilling and gridding are ~7% and ~1% respectively, while the temporal bias are ~25% and ~16% for gap-infilling and gridding respectively. Visual inspection reveals an increasing performance of gridded precipitation from grid, basin to island scale. The data set, stored in a network common data form (NetCDF), is intended to support basin-scale and island-scale studies of short-term and long-term climate, hydrology and ecology.本研究构建了一套覆盖印度尼西亚爪哇岛的网格化逐日气象数据集，包含降水、最低气温与最高气温要素，空间分辨率为0.125°×0.125°（约14公里），时间跨度为1985年至2014年，共计30年。 值得注意的是，相较于爪哇岛区域现有的网格化数据集，本数据集空间分辨率显著更高，且完全依托地面观测数据构建，区别于其他依赖卫星反演或再分析的产品。 网格化插值采用反距离加权法（Inverse Distance Weighting, IDW），以25公里的搜索半径与3的影响幂值作为最优参数，且仅纳入具备至少3650天（约10年）有效观测数据的气象站点参与插值。 在网格化插值流程启动前，已先行完成缺测值插补步骤。 我们分别针对缺测值插补流程与网格化插值流程开展了交叉验证。结果显示，缺测值插补与网格化插值的空间百分比偏差分别约为7%与1%，而二者对应的时间偏差则分别约为25%与16%。 目视评估结果表明，网格化降水产品的性能随分析尺度从格点、流域拓展至岛域而逐步提升。 本数据集以网络通用数据格式（Network Common Data Form, NetCDF）存储，旨在支撑流域尺度与岛域尺度下的短期、长期气候、水文及生态学相关研究。