气象卫星专题气候产品数据集-海表温度（SST）（2000-2019）

利用风云卫星再定标处理后的FY-1 VIRR L1级数据，采用基于深度学习的海温反演方法生成SST产品。风云一号卫星VIRR再处理日平均SST数据集包括FY1C/D星全周期全球日平均SST数据，空间分辨率为0.05°×0.05°。
FY-3A/B/C VIRR重处理SST日产品数据集V1.0，利用FY-3A/B/C VIRR L1级数据，采用一致性的统计回归非线性海温NLSST算法进行海温反演重处理，并对每个像元进行质量标识（质量等级QualityLevel，QL=0-5），生成5分钟段原始轨道投影（星下点1.1km）的海温产品。在每天约288个5分钟段海温的基础上，采用质量优先原则，经投影转换，进行5km全球等经纬度日合成，白天夜间单独存放，每颗星每天两次全球观测海温。时间范围为2009-2019，空间分辨率为0.05°×0.05°,全球5km等经纬度投影，HDF5格式存储。相比于业务数据，该数据集在精度和稳定性两方面均明显提升。气候用户建议仅使用夜间质量为优（QL=5）的海温，海洋监测用户可使用白天和夜间质量为优良（QL=5和4）的海温。数据集将在国内和国际免费共享。

SST products were generated using re-calibrated FY-1 VIRR L1 data from Fengyun satellites via a deep learning-based sea surface temperature (SST) retrieval method. The reprocessed daily-averaged SST dataset from FY-1 satellites' VIRR instruments covers full-cycle global daily-averaged SST data from FY-1C and FY-1D satellites, with a spatial resolution of 0.05° × 0.05°.

The FY-3A/B/C VIRR reprocessed daily SST product dataset V1.0 was developed using FY-3A/B/C VIRR L1 data. A consistent statistical regression-based nonlinear sea surface temperature (NLSST) algorithm was adopted for reprocessed SST retrieval, and quality flagging was performed for each pixel with quality levels (QualityLevel, QL=0-5). 5-minute interval original orbital-projected SST products with a sub-satellite point resolution of 1.1 km were generated. Based on approximately 288 sets of 5-minute interval SST data per day and following the quality-priority principle, global daily 5 km equal-latitude-longitude composite SST data was produced via projection transformation. Daytime and nighttime SST data are stored separately, with each satellite conducting two global SST observations per day. The dataset has a temporal coverage of 2009–2019, a spatial resolution of 0.05° × 0.05°, uses a global 5 km equal-latitude-longitude projection, and is stored in HDF5 format. Compared with operational datasets, this dataset exhibits significant improvements in both accuracy and stability. It is recommended that climate users only use nighttime SST data with excellent quality (QL=5), while ocean monitoring users can utilize both daytime and nighttime SST data with good or excellent quality (QL=5 and QL=4). This dataset will be freely shared both domestically and internationally.

搜集汇总

背景与挑战

背景概述

该数据集提供了2000-2019年全球海表温度（SST）的再处理数据，来源于FY-1和FY-3卫星的VIRR仪器，空间分辨率为5km，数据质量分级明确，适用于气候研究和海洋监测。