GPM_3IMERGM全球0.1x0.1°月平均降水数据集

"该数据集获取自美国国家航空航天局戈达德宇宙飞行中心。GPM的综合多星反演（IMERG）是美国的统一算法，为美国GPM团队提供多星降水产品。使用2017版Goddard剖面算法（GPROF2017）计算包含GPM星座的各种与降水相关的卫星被动微波（PMW）传感器的降水量估计值，然后进行网格化，再与GPM组合Ku雷达辐射计算法（CORRA）产品进行相互校准，并合并为半小时0.1°x0.1°（约10x10 km）场。请注意，CORRA调整为高纬度海洋和热带陆地上的每月全球降水气候学项目（GPCP）卫星测量仪（SG）产品，以纠正已知偏差。然后，将半小时相互校准的合并PMW估计输入气候预测中心（CPC）变形卡尔曼滤波器（CMORPH-KF）拉格朗日时间插值方案和使用人工神经网络云分类系统（PERSINN-CCS）重新校准方案的遥感信息降水量估计。并行地，CPC组装天顶角校正、相互校准的合并地理红外场，并将其转发给PPS，以输入到PERSIANN-CCS算法（由异步重新校准周期支持），然后输入到CMORPH-KF变形（准拉格朗日时间内插）方案。CMORPH-KF变形（由异步KF权重更新周期支持）使用PMW和IR估计值创建半小时估计值。变形的运动矢量是通过最大化垂直积分蒸汽（TQV）连续小时的模式相关性来计算的，TQV是由现代研究和应用回顾性分析提供的，版本2（MERRA-2）和戈达德地球观测系统模型版本5（GEOS-5）前向处理（FP），分别用于后实时（最终）运行和近实时（早期和晚期）运行。KF使用变形数据作为“预测”，IR估计值作为“观测值”，权重取决于距离微波天桥时间的时间间隔。距离天桥时间约±90分钟后，IR变得非常重要。
IMERG系统近实时运行两次，观测时间后4小时内仅使用前向变形的“早期”多星产品，观测时间后14小时内使用前向和后向变形的“晚期”多星产品，并在收到月度计量分析后进行一次；“最终”，观测月后3.5个月的卫星测量产品，使用正向和反向变形，包括月度测量分析。目前，近实时早半小时和晚半小时估计值没有最终校准，而在实时后的最终运行中，多卫星半小时估计值会进行调整，以使其总和达到最终运行的每月卫星仪表组合。在所有情况下，输出都包含多个字段，这些字段提供有关输入数据、选定中间字段和估计质量的信息。一般来说，对于大多数用户来说，完全校准的降水量（DepositionCal）是首选的数据字段。
简要描述最终运行，将从各种卫星无源微波传感器计算的输入降水量估计值与CORRA产品进行相互校准（因为在调整到每月GPCP SG后，它被认为是最佳的快照TRMM/GPM估计值），然后“前向/后向变形”，并与微波降水校准的geo IR场相结合，并根据季节性GPCP SG地面降水数据进行调整，以提供全球0.1°x0.1°（约10x10 km）网格上每半小时和每月的降水量估计。利用地表温度、湿度和压力分析计算降水阶段。"
将下载的HDF文件，转换成TIF格式，便于查看。"
GlobeLand30采用WGS-84坐标系，共包括10个一级类型，分别是：耕地、林地、草地、灌木地、湿地、水体、苔原、人造地表、裸地、冰川和永久积雪。
GlobeLand30分幅数据由地表覆盖数据文件、坐标信息文件、分类影像接图表文件、元数据文件4部分组成。其中：
1) 地表覆盖数据文件是指存储分幅地表覆盖分类信息的文件；
2) 坐标信息文件是指记录分幅数据坐标信息的文件；
3) 分类影像接图表文件是指记录分类所用的主要影像范围及获取时间的矢量文件；
4) 元数据文件指记录分幅数据元数据信息的文件。"

This dataset is sourced from the NASA Goddard Space Flight Center. The Integrated Multi-satellitE Retrievals for GPM (IMERG) is the unified U.S. algorithm that provides multi-satellite precipitation products for the U.S. GPM team. Precipitation estimates from various precipitation-related satellite passive microwave (PMW) sensors in the GPM constellation are computed using the 2017 Goddard Profiling Algorithm (GPROF2017), then gridded, cross-calibrated against the GPM Combined Ku Radar-Radiometer Algorithm (CORRA) product, and merged into half-hourly 0.1°×0.1° (≈10×10 km) grids.

Note that CORRA is adjusted to the monthly Global Precipitation Climatology Project (GPCP) Satellite Gauge (SG) product over high-latitude oceans and tropical land areas to correct known biases.

Subsequently, the half-hourly cross-calibrated merged PMW estimates are fed into the Climate Prediction Center (CPC) Morphing Kalman Filter (CMORPH-KF) Lagrangian temporal interpolation scheme and remote sensing precipitation estimates utilizing the Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks-Cloud Classification System (PERSINN-CCS) recalibration framework.

Concurrently, CPC assembles zenith-angle-corrected, cross-calibrated merged geostationary infrared (IR) fields and forwards them to the PPS, which inputs them into the PERSIANN-CCS algorithm (supported by asynchronous recalibration cycles) before passing them to the CMORPH-KF morphing (quasi-Lagrangian temporal interpolation) scheme.

The CMORPH-KF morphing scheme (supported by asynchronous Kalman filter weight update cycles) generates half-hourly precipitation estimates using both PMW and IR retrievals. Morphed motion vectors are calculated by maximizing the temporal correlation of vertically integrated water vapor (TQV) over consecutive hours, with TQV provided by the Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2) and the Goddard Earth Observing System Model, Version 5 (GEOS-5) forward processing (FP) for post-real-time (final) and near-real-time (early and late) runs, respectively.

The Kalman filter uses the morphed data as the "forecast" and IR retrievals as the "observations", with weights dependent on the temporal interval from the microwave overpass time. IR retrievals become highly influential approximately ±90 minutes away from the overpass time.

The IMERG system runs twice in near-real-time: the "early" multi-satellite product uses only forward morphing within 4 hours after observation time, while the "late" multi-satellite product uses both forward and backward morphing within 14 hours after observation time; additionally, a "final" run is conducted 3.5 months after the observation month, utilizing both forward and backward morphing and incorporating monthly gauge analysis.

Currently, the near-real-time early and late half-hourly estimates lack final calibration, whereas in the post-real-time final run, the multi-satellite half-hourly estimates are adjusted to match the monthly satellite-gauge composite totals from the final run.

In all cases, the output includes multiple fields that provide information on input data, selected intermediate fields, and estimation quality. Generally, the fully calibrated precipitation (DepositionCal) is the preferred data field for most users.

A brief description of the final run: Input precipitation estimates computed from various satellite passive microwave sensors are cross-calibrated against the CORRA product (which is regarded as the optimal snapshot TRMM/GPM precipitation estimate after adjustment to the monthly GPCP SG product), then subjected to "forward/backward morphing", combined with microwave-calibrated geostationary IR fields, and adjusted using seasonal GPCP SG surface precipitation data to produce half-hourly and monthly precipitation estimates on a global 0.1°×0.1° (≈10×10 km) grid. Precipitation phase is calculated using surface temperature, humidity, and pressure analyses.

Downloaded HDF files are converted to TIF format for easy visualization.

GlobeLand30 adopts the WGS-84 coordinate system and includes 10 primary land cover types, namely: cropland, forestland, grassland, shrubland, wetland, water bodies, tundra, artificial surfaces, bare land, glaciers and permanent snow.

The tiled GlobeLand30 dataset consists of four components: land cover data file, coordinate information file, classified image index map file, and metadata file. Specifically:
1) Land cover data file: A file that stores the land cover classification information for each tile;
2) Coordinate information file: A file that records the coordinate information of the tiled dataset;
3) Classified image index map file: A vector file that records the main image ranges and acquisition times used for classification;
4) Metadata file: A file that records the metadata information of the tiled dataset.

搜集汇总

背景与挑战

背景概述

GPM_3IMERGM全球0.1x0.1°月平均降水数据集是一个高分辨率的全球降水数据集，覆盖2010年至2021年，由NASA戈达德太空飞行中心通过多卫星数据融合和校准生成，适用于大气科学和遥感应用研究。