Unified Brazilian Rainfall Dataset (UNIPLU-BR): A Standardized National Database of Point Precipitation from Major Brazilian Monitoring Networks (1885 - 2025)
收藏资源简介：
Unified Brazilian Rainfall Dataset (UNIPLU-BR): A Standardized National Database of Point Precipitation from Major Brazilian Monitoring Networks (1885 – 2025) This dataset is the first unified and standardized national database of point precipitation (non-interpolated) in Brazil, consolidating raw data from five primary monitoring networks: CEMADEN: National Center for Monitoring and Early Warning of Natural Disasters. INMET: National Institute of Meteorology. ANA (Hidroweb): National Water and Sanitation Agency. Telemetria: Telemetry system for hydrological monitoring. ICEA: Institute of Air Space Control. The primary contribution of this work is overcoming the high fragmentation of rainfall data in Brazil through a rigorous curation process: Structural and Nominal Standardization: Harmonization of column names, attributes, and storage formats, addressing historical changes in data protocols and formats within the same agency. Time Zone Adjustment: Standardization of timestamps based on the station's geographical location (UTC offset). Temporal Resolution: Records ranging from 10-minute intervals to 1-day resolution. Data Quality Disclaimer: The processing of this dataset is strictly focused on structural standardization. No qualitative assessment, physical consistency checks, or outlier filtering were performed. The rainfall values remain as originally reported by the agencies, now organized into a unified, analysis-ready structure. Impressive numbers: The dataset covers the period from 1885 to 2025, highlighted by the presence of stations with historical series exceeding a century. With broad national coverage, the database consolidates approximately 2.2 billion precipitation records from over 21,000 stations. This information features varied temporal resolutions, ranging from 10-minute intervals to 24-hour totals. The distribution of these records among the main Brazilian monitoring networks is detailed below. Source Potential average number of years Quantity Initial and Final Year CEMADEN 7,99 5.061 2014 - 2025 Hidroweb 32,38 12.067 1885 - 2025 ICEA 33,19 183 1951 - 2025 INMET daily 45,03 627 1889 - 2025 INMET sub-daily 15,87 629 2000 - 2025 Telemetria 6,63 2.819 2014 - 2025 Total — 21.386 The following figures illustrate the spatial distribution of the stations and the annual data availability for both daily and sub-daily networks. Figure 1 Figure 2 Accessing the Data The data is stored in compressed ZIP files, which function as optimized containers. Each ZIP file internally contains two files in Parquet format: table_info.parquet (station metadata) and table_data.parquet (rainfall time series). The primary advantage of this structure is that, using Python or R, you can read the data directly from memory. This eliminates the need to manually decompress files to the disk, saving storage space and accelerating processing within automation workflows. Within these files, the gauge_code (station code) serves as the primary key that links the registration information to the measurement data. Metadata Rainfall gauge information (table_info) This dataframe functions as the 'identity document' for the rain gauge stations. It contains the static characteristics of each monitoring point: Column Description gauge_code Unique station identifier (ID). This serves as the link to table_data. city / state The administrative location of the station (e.g. João Pessoa, PB). lat / long Geographic coordinates in decimal degrees. elevation The station's altitude above mean sea level (meters). time_step Estimated temporal resolution of the data (1440 minutes = 24 hours/daily). network Data network source (e.g. Hidroweb). responsible The agency responsible for operations (ANA or SGB-CPRM). utc Local time zone relative to the Greenwich Meridian (-3 for Brasília Time). Time Series (table_data) Whilst table_info defines the location and identity of the station, table_data records the rainfall measurements. It contains the following columns: Column Description gauge_code Unique station identifier (ID). This serves as the link to table_info. datetime The date and time of the reading. rain_mm The volume of precipitation recorded during that interval, measured in millimeters (mm). Script Examples Examples of scripts for accessing and filtering data, as well as generating plots, can be found at the following link: GitHub - LARHENA/UNIPLU-BR: Unified Brazilian Rainfall Dataset (UNIPLU-BR): A Standardized National Database of Point Precipitation from Major Brazilian Monitoring Networks (1885 – 2025) · GitHub How to Cite Das Neves Almeida, C., Francis Bertrand, G., Carvalho Lemos, F., da Silva Freitas, E., Lins Silva, A., Vidal Barbosa, J. L., ... & Coelho, V. H. R. (2025). The design of the Brazilian Sub-Daily Rainfall dataset (BR-SDR): two decades of high-time-resolution data in Brazil. Hydrological Sciences Journal, 70(11), 1850-1862. https://doi.org/10.1080/02626667.2025.2506193 Main papers published by the group Das Neves Almeida, C., Francis Bertrand, G., Carvalho Lemos, F., da Silva Freitas, E., Lins Silva, A., Vidal Barbosa, J. L., ... & Coelho, V. H. R. (2025). The design of the Brazilian Sub-Daily Rainfall dataset (BR-SDR): two decades of high-time-resolution data in Brazil. Hydrological Sciences Journal, 70(11), 1850-1862. Vidal-Barbosa, J. L., Lemos, F. C., da Silva Freitas, E., Coelho, V. H. R., da Silva, G. N. S., Patriota, E. G., ... & das Neves Almeida, C. (2025). BRain-D: A quality-controlled methodology for constructing the BRazilian Daily rainfall gridded data. Atmospheric Research, 108552. Freitas, E. D. S., Coelho, V. H. R., Bertrand, G. F., Lemos, F. C., & Almeida, C. D. N. (2024). IMERG BraMaL: An improved gridded monthly rainfall product for Brazil based on satellite‐based IMERG estimates and machine learning techniques. International Journal of Climatology, 44(11), 3976-3997. Lemos, F. C., Coelho, V. H. R., Freitas, E. D. S., Tomasella, J., Bertrand, G. F., Meira, M. A., ... & Almeida, C. D. N. (2023). Spatiotemporal distribution of precipitation and its characteristics under tropical atmospheric systems of Brazil: Insights from a large sub‐hourly database. Hydrological Processes, 37(11), e15017. Ramos Filho, G. M., Coelho, V. H. R., da Silva Freitas, E., Xuan, Y., Brocca, L., & das Neves Almeida, C. (2022). Regional-scale evaluation of 14 satellite-based precipitation products in characterising extreme events and delineating rainfall thresholds for flood hazards. Atmospheric Research, 276, 106259. Meira, M. A., Freitas, E. S., Coelho, V. H. R., Tomasella, J., Fowler, H. J., Ramos Filho, G. M., ... & Almeida, C. D. N. (2022). Quality control procedures for sub-hourly rainfall data: An investigation in different spatio-temporal scales in Brazil. Journal of Hydrology, 613, 128358. Freitas, E. D. S., Coelho, V. H. R., Xuan, Y., de CD Melo, D., Gadelha, A. N., Santos, E. A., ... & Almeida, C. D. N. (2020). The performance of the IMERG satellite-based product in identifying sub-daily rainfall events and their properties. Journal of Hydrology, 589, 125128. File containing the stations that underwent spatial adjustments: Click here
统一化巴西降雨数据集（Unified Brazilian Rainfall Dataset, UNIPLU-BR）：巴西主要监测网络站点降水标准化国家级数据库（1885 – 2025） 本数据集为巴西首个统一化、标准化的站点降水（非插值）国家级数据库，整合了来自5个核心监测网络的原始数据： CEMADEN：国家自然灾害监测与预警中心（National Center for Monitoring and Early Warning of Natural Disasters） INMET：国家气象研究所（National Institute of Meteorology） ANA（Hidroweb）：国家水资源与卫生署（National Water and Sanitation Agency） Telemetria：水文监测遥测系统 ICEA：空域管控研究所（Institute of Air Space Control） 本研究的核心贡献在于通过严格的数据整理流程，解决了巴西降雨数据高度碎片化的难题： 1. 结构与名义标准化：统一列名、属性与存储格式，解决同一机构内数据协议与格式的历史变更问题； 2. 时区校准：基于监测站地理位置（UTC偏移量）标准化时间戳； 3. 时间分辨率：记录间隔覆盖10分钟至1日尺度。 数据质量声明：本数据集仅针对结构进行标准化处理，未开展任何定性评估、物理一致性校验或异常值过滤。降雨数值仍保留各机构原始上报结果，仅整合为统一的、可直接用于分析的结构化格式。 核心数据规模： 本数据集覆盖1885年至2025年的时间跨度，包含大量历史序列超过百年的监测站点。数据库实现了全国范围的广泛覆盖，整合了来自21000余个监测站点的约22亿条降水记录。这些数据的时间分辨率跨度较大，从10分钟间隔至24小时累积量不等。 各监测网络的记录分布详情如下： | 数据网络 | 平均潜在使用年限 | 站点数量 | 起始与终止年份 | |----------------|------------------|----------|----------------| | CEMADEN | 7.99 | 5061 | 2014 - 2025 | | Hidroweb | 32.38 | 12067 | 1885 - 2025 | | ICEA | 33.19 | 183 | 1951 - 2025 | | INMET日分辨率 | 45.03 | 627 | 1889 - 2025 | | INMET次日分辨率| 15.87 | 629 | 2000 - 2025 | | Telemetria | 6.63 | 2819 | 2014 - 2025 | | 总计 | — | 21386 | — | 以下图表展示了监测站点的空间分布，以及日分辨率与次日分辨率网络的年度数据可用性： 图1 图2 数据获取方式 数据以压缩ZIP文件存储，作为优化的容器格式。每个ZIP文件内部包含两个Parquet格式文件："table_info.parquet"（站点元数据表）与"table_data.parquet"（降雨时间序列数据表）。 该结构的核心优势在于，可通过Python或R直接从内存中读取数据，无需手动将文件解压至磁盘，既节省存储空间，又可加速自动化工作流中的数据处理流程。 在这些文件中，站点编码（gauge_code）作为主键，将注册信息与测量数据关联起来。 元数据 雨量站信息（table_info） 该数据框作为雨量监测站点的“身份凭证”，包含每个监测点的静态属性： | 列名 | 列说明 | |--------------|----------------------------------------------------------------------| | gauge_code | 唯一站点标识符（ID），即站点编码（gauge_code），用于关联table_data表。 | | city / state | 站点所属行政区域（例如：若昂佩索阿，帕拉伊巴州）。 | | lat / long | 以十进制度表示的地理坐标。 | | elevation | 站点海拔高度（单位：米，相对于平均海平面）。 | | time_step | 数据估算的时间分辨率（1440分钟=24小时/日尺度）。 | | network | 数据来源网络（例如：Hidroweb）。 | | responsible | 负责运维的机构（ANA或SGB-CPRM）。 | | utc | 相对于格林威治标准时间的本地时区（巴西利亚时间为UTC-3）。 | 时间序列数据（table_data） table_info定义了站点的位置与身份信息，而table_data则记录降雨测量数据，其包含以下列： | 列名 | 列说明 | |--------------|----------------------------------------------------------------------| | gauge_code | 唯一站点标识符（ID），即站点编码（gauge_code），用于关联table_info表。 | | datetime | 观测的日期与时间。 | | rain_mm | 该时段内记录的降水量，单位为毫米（mm）。 | 脚本示例 可通过以下链接获取用于数据读取、筛选及绘图的脚本示例：GitHub - LARHENA/UNIPLU-BR: Unified Brazilian Rainfall Dataset (UNIPLU-BR): A Standardized National Database of Point Precipitation from Major Brazilian Monitoring Networks (1885 – 2025) · GitHub 引用方式 Das Neves Almeida, C., Francis Bertrand, G., Carvalho Lemos, F., da Silva Freitas, E., Lins Silva, A., Vidal Barbosa, J. L., ... & Coelho, V. H. R. (2025). The design of the Brazilian Sub-Daily Rainfall dataset (BR-SDR): two decades of high-time-resolution data in Brazil. 《水文科学杂志》, 70(11), 1850-1862. https://doi.org/10.1080/02626667.2025.2506193 团队已发表的核心论文 1. Das Neves Almeida, C., Francis Bertrand, G., Carvalho Lemos, F., da Silva Freitas, E., Lins Silva, A., Vidal Barbosa, J. L., 等（2025）. 巴西次日分辨率降雨数据集（BR-SDR）的构建：巴西二十年高时间分辨率数据. 《水文科学杂志》, 70(11), 1850-1862. 2. Vidal-Barbosa, J. L., Lemos, F. C., da Silva Freitas, E., Coelho, V. H. R., da Silva, G. N. S., Patriota, E. G., 等（2025）. BRain-D：构建巴西逐日降雨格点数据的质量控制方法. 《大气研究》, 108552. 3. Freitas, E. D. S., Coelho, V. H. R., Bertrand, G. F., Lemos, F. C., & Almeida, C. D. N. (2024). IMERG BraMaL: An improved gridded monthly rainfall product for Brazil based on satellite‐based IMERG estimates and machine learning techniques. 《国际气候学杂志》, 44(11), 3976-3997. 4. Lemos, F. C., Coelho, V. H. R., Freitas, E. D. S., Tomasella, J., Bertrand, G. F., Meira, M. A., 等（2023）. 巴西热带大气系统下降水的时空分布及其特征：基于大型亚小时级数据库的洞察. 《水文过程》, 37(11), e15017. 5. Ramos Filho, G. M., Coelho, V. H. R., da Silva Freitas, E., Xuan, Y., Brocca, L., & das Neves Almeida, C. (2022). Regional-scale evaluation of 14 satellite-based precipitation products in characterising extreme events and delineating rainfall thresholds for flood hazards. 《大气研究》, 276, 106259. 6. Meira, M. A., Freitas, E. S., Coelho, V. H. R., Tomasella, J., Fowler, H. J., Ramos Filho, G. M., 等（2022）. 亚小时级降雨数据的质量控制流程：巴西不同时空尺度的研究. 《水文杂志》, 613, 128358. 7. Freitas, E. D. S., Coelho, V. H. R., Xuan, Y., de CD Melo, D., Gadelha, A. N., Santos, E. A., 等（2020）. IMERG卫星产品在识别次日降雨事件及其特征中的性能. 《水文杂志》, 589, 125128. 包含经过空间校正站点的文件：点击此处
