CAMELS-BR: Hydrometeorological time series and landscape attributes for 897 catchments in Brazil - link to files.

Version 1.2 (March 2025): Now with longer time series, expanded stream gauge coverage, meteorological data from additional sources, soil moisture time series, and observed rainfall time series from 11,853 rain gauges.

This is the CAMELS-BR dataset (Catchment Attributes and MEteorology for Large-sample Studies – Brazil) accompanying the paper: Chagas, V. B. P., Chaffe, P. L. B., Addor, N., Fan, F. M., Fleischmann, A. S., Paiva, R. C. D., and Siqueira, V. A.: CAMELS-BR: hydrometeorological time series and landscape attributes for 897 catchments in Brazil, Earth Syst. Sci. Data, 12, 2075–2096, https://doi.org/10.5194/essd-12-2075-2020, 2020.

CAMELS-BR provides daily observed streamflow time series for 4,025 stream gauges, daily observed rainfall for 11,853 rain gauges, daily meteorological time series and 65 attributes for 897 catchments in Brazil.

The daily hydrometeorological time series include (i) observed streamflow accompanied by quality control information, (ii) precipitation extracted from five products, (iii) actual evapotranspiration extracted from three products, (iv) potential evapotranspiration extracted from two products, (v) reference evapotranspiration extracted from one product, (vi) minimum, mean, and maximum temperature extracted from three products, and (vii) soil moisture extracted from two products.

The 65 catchment attributes cover properties such as (i) topography, (ii) climate, (iii) hydrology, (iv) land cover, (v) geology, (vi) soil, and (vii) human intervention.

The data follow the same standards as other CAMELS datasets such as for the United States (https://doi.org/10.5194/hess-21-5293-2017), Chile (https://doi.org/10.5194/hess-22-5817-2018), and Great Britain (https://doi.org/10.5194/essd-2020-49).

How to cite: Chagas, V. B. P., Chaffe, P. L. B., Addor, N., Fan, F. M., Fleischmann, A. S., Paiva, R. C. D., and Siqueira, V. A.: CAMELS-BR: hydrometeorological time series and landscape attributes for 897 catchments in Brazil, Earth Syst. Sci. Data, 12, 2075–2096, https://doi.org/10.5194/essd-12-2075-2020, 2020.

Changes in CAMELS-BR version 1.2:

Major changes

• Updated streamflow time series up to February 2025 (where available), as obtained from ANA's website on 27 February 2025 (ANA – Brazilian National Water and Sanitation Agency – http://www.snirh.gov.br/hidroweb/). Some historical records have changed slightly due to ANA's quality control procedures. For eight gauges (see the readme.txt file), data are merged from 2025 and 2019 records (i.e. from CAMELS-BR version 1.1).
• Increased the stream gauge coverage to 4025 stream gauges (including both quality-controlled and non-quality-controlled series), up from 3679 in version 1.1.
• Added daily observed rainfall time series for 11853 rain gauges (not catchment averages), as obtained from ANA's website on 27 February 2025 (ANA – Brazilian National Water and Sanitation Agency – http://www.snirh.gov.br/hidroweb/). Data include quality flags but are mostly not quality-controlled.
• Added a GeoPackage file with coordinates for 11853 rain gauges.
• Updated precipitation time series (catchment averages) up to October 2024 (where available). Now derived from: CHIRPS v2.0; CPC; ERA5-Land; MSWEP v2.8; and BR-DWGD v3.2.3 (when at least 95% of the catchment area lies within Brazil – 864 catchments).
• Updated actual evapotranspiration time series (catchment averages) up to October 2024 (where available). Now derived from: GLEAM v4.2a; ERA5-Land; and MGB-SA.
• Updated potential evapotranspiration time series (catchment averages) up to October 2024 (where available). Now derived from GLEAM v4.2a and ERA5-Land.
• Added reference evapotranspiration time series (catchment averages). Derived from BR-DWGD v3.2.3 (when at least 95% of the catchment area lies within Brazil).
• Updated daily maximum, mean, and minimum temperature time series (catchment averages) up to October 2024 (where available). Now derived from: CPC; ERA5-Land; and BR-DWGD v3.2.3 (when at least 95% of the catchment area lies within Brazil).
• Added daily soil moisture time series (catchment averages) up to December 2024 (where available). Computed from GLEAM v4.2a and ERA5-Land.
• Improved meteorological data processing. Catchment averages now account for pixel fraction coverage.
• Reformatted meteorological time series files. Files now includes data from different products, with columns renamed for clarity.
• Hydrological and climatic indices were not updated, despite the new streamflow and meteorological data.

Minor changes

• Updated stream gauge coordinates based on ANA's website on 27 February 2025. Coordinates were updated for 73 gauges in the 897 selected catchments and for 298 gauges across all catchments.
• Streamflow time series now include quality flag values from 0 to 7 (see the readme.txt file), previously from 0 to 4 in CAMELS-BR version 1.1. Flags from 5 to 7 may be present only in the last few years of data.
• Streamflow time series files for the 897 selected gauges now include values in both millimeters per day and cubic meters per second.
• Removed streamflow time series with fewer than 180 days of measurement.
• Converted gauge and catchment spatial data from Shapefile (.shp) to GeoPackage (.gpkg).
• Catchment areas computed by GSIM (in files "camels_br_location.txt" and "location_gauges_streamflow.gpkg") flagged as "caution" for quality were set to "nan" due to low reliability.
• Updated catchment areas computed by ANA (in files "camels_br_location.txt" and "location_gauges_streamflow.gpkg") to reflect the newest ANA's data from 27 February 2025. Streamflow values in millimeters per day remain unchanged because unit conversions rely on GSIM areas.
• Set catchment areas with zero squared kilometers, as computed by ANA, to "nan".
• Removed CPC daily mean temperature time series (catchment averages) because they were a simple average of minimum and maximum temperatures. For daily mean temperatures, refer to ERA5-Land data (now included) as they are computed from hourly data.