ERA5_precipitation_sum_1950_2023

ERA5-Land is a reanalysis dataset that provides a consistent view of the evolution of terrestrial variables over decades at a higher resolution than ERA5, which was generated by replaying the terrestrial component of the ERA5 climate reanalysis of ECMWF. The reanalysis uses the laws of physics to combine model data with observations from around the world to produce a globally complete and consistent dataset. The reanalysis produces data that goes back decades and provides an accurate picture of past climate. The dataset includes all 50 variables available on the CDS. The asset is a monthly summary of the ECMWF ERA5 Land hourly asset, including both mobile and non-mobile bands. The mobile bands are created by collecting data from the first hour of the second day of each day of the month and adding them together, while the non-mobile bands are created by averaging all hourly data for the month. The flow bands are labelled with the ‘_sum’ identifier, which is different from the monthly data generated by the Copernicus Climate Data Store, which also averages the flow bands. When sub-monthly fields are not required, monthly totals have been pre-calculated for many applications that require easy and quick access to the data. ERA5-Land monthly summary data are available in real time from 1950, three months before the present. More information can be found in the Copernicus Climate Data Repository. Precipitation and other flow (accumulation) bands may occasionally have negative values that are physically unreasonable. Sometimes their values may be too high. This problem is caused by the way the GRIB format saves data: it simplifies or ‘packages’ the data into smaller, less precise numbers, which can lead to errors. These errors are made worse when the data changes a lot. As a result, when we look at a whole day's worth of data to calculate daily totals, sometimes the highest rainfall recorded at one time appears to be greater than the total rainfall measured throughout the day.ERA5-Land是一款再分析数据集，其通过复现欧洲中期天气预报中心（ECMWF）的ERA5气候再分析的陆面分量模块，能够以高于ERA5的分辨率，提供数十年间陆面变量演变的一致性观测视图。该再分析依托物理定律，将数值模式数据与全球各地的观测数据进行融合，从而生成一套全球完整且自洽的数据集。其回溯数据可覆盖数十年跨度，能够精准刻画过往气候的真实状况。本数据集包含气候数据存储（Climate Data Store, CDS）平台上可用的全部50种变量。本次提供的资产为ECMWF ERA5 Land逐小时资产的月度汇总产物，涵盖移动波段（mobile bands）与非移动波段（non-mobile bands）两类。其中移动波段的生成规则为：采集当月每日次日第一小时的观测数据并进行求和；非移动波段则通过对当月所有逐小时数据取算术平均得到。需注意，带有‘_sum’标识的流量波段（flow bands），其生成方式与哥白尼气候数据服务（Copernicus Climate Data Store）的常规月度数据不同——后者会对流量波段执行平均处理。对于无需次月度数据的应用场景，我们已预先计算好月度总量，以满足各类对数据获取便捷性和时效性有要求的需求。ERA5-Land月度汇总数据自1950年起即可实时获取，数据更新较当前时间滞后三个月。更多详细信息可查阅哥白尼气候数据仓库（Copernicus Climate Data Repository）。降水及其他流量（累积）波段偶尔会出现不符合物理规律的负值，部分情况下数值也可能偏高。该问题源于GRIB格式的数据存储机制：GRIB会将数据简化或“打包”为更小、精度更低的数值，进而引发误差。当数据波动幅度较大时，这类误差会被进一步放大。最终表现为，在基于单日逐小时数据计算日降雨总量时，有时会出现单次记录的最高降雨量大于当日总降雨量的反常情况。