PatagoniaMet: A multi-source hydrometeorological dataset for Western Patagonia

PatagoniaMet v1.0 (PMET from here on) is a new dataset for Western Patagonia that consists of two datasets: i) PMET-obs, a compilation of quality-controlled ground-based hydrometeorological data, and ii) PMET-sim, a daily gridded product of precipitation, and maximum and minimum temperature. PMET-obs was developed using a 4-step quality control process applied to 523 hydro-meteorological time series (precipitation, air temperature, potential evaporation, streamflow and lake level stations) obtained from eight institutions in Chile and Argentina.  Based on this dataset and currently available uncorrected gridded products (in this case ERA5), PMET-sim was developed using statistical bias correction procedures (i.e. quantile mapping), spatial regression models (random forest) and hydrological methods (Budyko framework). Details are given below.

- PMET-obs is a compilation of five hydrometeorological variables obtained from eight institutions in Chile and Argentina. The daily quality controlled data of each variable are stored in separate .csv files with the following naming convention: variable_PMETobs_timeperiod_version/timestep.csv. Each column represents a different gauge with its "gauge_id". Each variable has an additional .csv file containing the metadata for each station (variable_PMETobs_version_metadata.csv). In order to make transparent the possible erroneous data that were discarded from the quality-controlled version, a .zip file with the raw data of all variables is attached. The metadata file (final and raw versions) contains the station name (gauge_name), the institution, the station location (gauge_lat and gauge_lon), the NASADEM elevation (gauge_alt) and the total number of daily records (length). In addition, the precipitation and temperature metadata include the number of monthly outliers (step Nº3 in the methods) and the number of changepoints (step Nº4 in the methods).

The streamflow metadata file (Q_PMETobs_version_metadata.csv) contains more than just the location data. Following current guidelines for hydrological datasets, the upstream area corresponding to each stream gauge was delimited (.shp file in Basins_PMETobs_version.zip), and several climatic and geographic attributes were derived. The details of the attributes can be found in the README file. For the basins that were part of the hydrological modelling (and that achieved a Kling-Gupta efficiency greater than 0.5), the file Q_PMETobs_version_water_balance.csv is attached, which contains the water balance for each basin estimated for the period 1985-2019. 

- PMET-sim is a daily gridded product with a spatial resolution of 0.05° covering the period 1980-2020. The data for each variable (precipitation and maximum and minimum temperature) are stored in separate netcdf files with the following naming convention: variable_PMETsim_1980_2020_v10d.nc.

Citation:   Aguayo, R., León-Muñoz, J., Aguayo, M., Baez-Villanueva, O., Fernandez, A. Zambrano-Bigiarini, M., and Jacques-Coper, M. (2023) PatagoniaMet: A multi-source hydrometeorological dataset for Western Patagonia. Sci Data 11, 6 (2024). https://doi.org/10.1038/s41597-023-02828-2

Code repository: https://github.com/rodaguayo/PatagoniaMet

PatagoniaMet v1.0（下文简称PMET）是面向西巴塔哥尼亚的全新水文气象数据集，包含两个子数据集：一是PMET-obs，即经过质量控制的地面水文气象数据汇编；二是PMET-sim，即降水、最高与最低气温的逐日格点数据集。PMET-obs基于智利与阿根廷8家机构提供的523条水文气象时间序列（涵盖降水、气温、潜在蒸散发、径流与湖泊水位站点），通过四步质量控制流程构建。本数据集结合当前公开的未校正格点产品（此处为ERA5），采用统计偏差校正方法（分位数映射法（quantile mapping））、空间回归模型（随机森林模型（random forest））以及水文方法（Budyko框架（Budyko framework））生成PMET-sim。详细说明如下。

- PMET-obs包含从智利与阿根廷8家机构获取的5类水文气象变量。各变量的逐日质量控制数据以单独的.csv文件存储，命名规则为：variable_PMETobs_timeperiod_version/timestep.csv。每一列代表一个不同的测站，带有对应的测站ID（gauge_id）。每个变量还附带一份单独的.csv元数据（metadata）文件，命名格式为variable_PMETobs_version_metadata.csv，内含各测站的相关信息。为透明展示质控过程中剔除的可疑数据，我们附带了包含所有变量原始数据的.zip压缩包。
元数据（metadata）文件（最终版与原始版）包含测站名称（gauge_name）、所属机构、测站位置（测站纬度（gauge_lat）与测站经度（gauge_lon））、通过NASADEM获取的测站海拔（gauge_alt）以及单日记录总数（length）。此外，降水与气温元数据还包含月度异常值数量（方法第3步）与突变点数量（方法第4步）。

径流元数据（metadata）文件（Q_PMETobs_version_metadata.csv）包含的信息不止位置数据。遵循当前水文数据集的规范，我们对每个径流测站对应的上游汇水区进行了划定（相关.shp文件存储于Basins_PMETobs_version.zip中），并推导了多项气候与地理属性。属性的详细说明可参阅README文件。对于参与水文建模且Kling-Gupta效率系数（Kling-Gupta efficiency）大于0.5的汇水区，我们附带了文件Q_PMETobs_version_water_balance.csv，内含1985-2019年各汇水区的水量平衡估算结果。

- PMET-sim是空间分辨率为0.05°的逐日格点产品，时间覆盖范围为1980-2020年。各变量（降水、最高与最低气温）的数据以单独的netcdf文件存储，命名规则为：variable_PMETsim_1980_2020_v10d.nc。

引用信息：Aguayo, R., León-Muñoz, J., Aguayo, M., Baez-Villanueva, O., Fernandez, A., Zambrano-Bigiarini, M. 及 Jacques-Coper, M. (2023) PatagoniaMet: A multi-source hydrometeorological dataset for Western Patagonia. Sci Data 11, 6 (2024). https://doi.org/10.1038/s41597-023-02828-2
代码仓库：https://github.com/rodaguayo/PatagoniaMet