This file contains the Supplement (both raw observed precipitation data and figures obtained as output of the analysis) accompanying the manuscript 'hess-2016-453' submitted to the HESS journal (http://www.hydrology-and-earth-system-sciences.net/).

Title: "Temporal and spatial evaluation of satellite-based rainfall estimates across the complex topographical and climatic gradients of Chile"

Abstract

Accurate representation of the real spatio-temporal variability of catchment rainfall inputs is currently severely limited. Moreover, spatially interpolated catchment precipitation is subject to large uncertainties, particularly in developing countries and regions which are difficult to access. Recently, satellite-based rainfall estimates (SRE) provide an unprecedented opportunity for a wide range of hydrological applications, from water resources modelling to monitoring of extreme events such as droughts and floods.

This study attempts to exhaustively evaluate -for the first time- the suitability of seven state-of-the-art SRE products (TMPA 3B42v7, CHIRPSv2, CMORPH, PERSIANN-CDR, PERSIAN-CCS-adj, MSWEPv1.1 and PGFv3) over the complex topography and diverse climatic gradients of Chile. Different temporal scales (daily, monthly, seasonal, annual) are used in a point-to-pixel comparison between precipitation time series measured at 366 stations (from sea level to 4600 m a.s.l. in the Andean Plateau) and the corresponding grid cell of each SRE (rescaled to a 0.25° grid if necessary). The modified Kling-Gupta efficiency was used to identify possible sources of systematic errors in each SRE. In addition, five categorical indices (PC, POD, FAR, ETS, fBIAS) were used to assess the ability of each SRE to correctly identify different precipitation intensities.

Results revealed that most SRE products performed better for the humid South (36.4-43.7°S) and Central Chile (32.18-36.4°S), in particular at low- and mid-elevation zones (0-1000 m a.s.l.) compared to the arid northern regions and the Far South. Seasonally, all products performed best during the wet seasons autumn and winter (MAM-JJA) compared to summer (DJF) and spring (SON). In addition, all SREs were able to correctly identify the occurrence of no rain events, but they presented a low skill in classifying precipitation intensities during rainy days. Overall, PGFv3 exhibited the best performance everywhere and for all time scales, which can be clearly attributed to its bias-correction procedure using 217 stations from Chile. Good results were also obtained by the research products CHIRPSv2, TMPA 3B42v7 and MSWEPv1.1,while CMORPH, PERSIANN-CDR and the real-time PERSIANN-CCS-adj were less skillful in representing observed rainfall. While PGFv3 (currently available up to 2010) might be used in Chile for historical analyses and calibration of hydrological models, the high spatial resolution, low latency and long data records of CHIRPS and TMPA 3B42v7 (in transition to IMERG) show promising potential to be used in meteorological studies and water resources assessments. We finally conclude that despite improvements of most SRE products, a site-specific assessment is still needed before any use in catchment-scale hydrological studies.

Information on the spatio-temporal distribution of rainfall is very critical for addressing water related disasters, especially in the arid to semi-arid regions of the Middle East and North Africa region. However, availability of reliable rainfall datasets for the region is limited. In this study we combined observation from satellite-based rainfall data, in situ rain gauge observation and rainfall climatology to create a reliable regional rainfall dataset for Jordan, West Bank and Lebanon. First, we validated three satellite-based rainfall products using rain gauge observations obtained from Jordan (205 stations), Palestine (44 stations) and Lebanon (8 stations). We used the daily 25-km Tropical Rainfall Measuring Mission over 2000 – 2016; daily 10-km Rainfall Estimate for Africa (RFE) rainfall over 2001 – 2016; daily 5-km Climate Hazards Group Infrared Precipitation with Station (CHIRPS) rainfall over 1981-2015; daily 25-km Multi-Source Weighted-Ensemble Precipitation (MSWEP) ov ...

Understanding the drought characteristics of mountainous areas in northwest China with sparse rainfall stations requires high precision, as well as high-resolution precipitation data. Considering the spatial relationship of precipitation and environmental factors, this study downscales Global Precipitation Measurement (GPM) and Multi-Source Weighted-Ensemble Precipitation (MSWEP) based on the geographically weighted regression (GWR) and multi-scale geographically weighted regression (MGWR) models integrated with interpolation. A high-resolution (1 km×1 km) precipitation dataset during 1979–2020 is reconstructed in the Tianshan Mountains, and the drought characteristics are analyzed by using the optimal dataset. The results show that: 1) Compared with GWR, MGWR model has higher downscaling accuracy; 2) The optimal MSWEP downscaling dataset (CC = 0.93, |BIAS| = 0.48%) compared to GPM (CC = 0.81, |BIAS| = 1.87%) is closer to the observed precipitation; 3) In the past 40 years, 71% and 9% of the Tianshan Mountains show significant wetting and drying trends respectively, and 16 drought events are identified. 4) The West subregion of the Tianshan Mountains is characterized by low frequency, long duration and high severity of drought events. The characteristics of the East are opposite to those of the West. Occasional extreme drought events occur in the North and South. This paper provides data support and method reference for the study of water-vapor balance and regional ecohydrological process in the arid area of Northwest China.

The presented database is a set of hydrological, meteorological, environmental and geometric values for Russia Federation for the period from 2008 to 2020.

Database consist of next items:

• Point geometry for hydrological observation stations from Roshydromet network across Russia
• Geometry of the catchment for correspond observation station point
• Daily hydrological values
  • Water level
    • In relative representation (sm)
    • In meters of Baltic system (m)
  • Water discharge
    • as an observed value (qms/s)
    • as a layer (mm/day)
• Daily meteorological values
  • Maximum and minimum daily temperatures (°C) from ERA5 and ERA5-Land
  • Total precipitation (mm/day) from ERA5, ERA5-Land, IMERG v06, GPCP v3.2 and MSWEP
  • Different kind of evaporation (mm/day) corresponded to each variable calculated in GLEAM model
• Set of hydro-environmental characteristics derived from HydroATLAS database

Each variable derived from the grid data was calculated for each watershed, taking into account the intersection weights of the watershed contour geometry and grid cells.

Coordinates of hydrological stations were obtained from resource of Federal Agency for Water Resources of Russia Federation—AIS GMVO

To calculate the contours of the catchment areas, a script was developed that builds the contours in accordance with the rasters of flow directions from MERIT Hydro. To assess the quality of the contour construction, the obtained value of the catchment area was compared with the archival value from the corresponded table from AIS GMVO. The average error in determining the area for 2080 catchments is approximately 2%

To derive values for different hydro-environmental values from HydroATLAS were developed approach which calculate aggregated values for catchment, leaning on type of variable: qualitative (Land cover classes, Lithological classes etc.) Or quantitive (Air temperature, Snow cover extent etc.). Every quantitive variable were calculated as mode value for intersected sub-basins and target catchment, e.g. most popular attribute from sub-basins will describe whole catchment which are they relating. Quantitative values were calculated as mean value of attribute from each sub-basin. More detail could be found in publication.

Files are distributed as follows:

Each file has some connection with the unique identifier of the hydrological observation post. Files in netcdf format (hydrological and meteorological series) are named in response to identifier.

Every file which describe geometry (point, polygon, static attributes) has and column named gauge_id with same correspondence.

• attributes/static_data.csv – results from HydroATLAS aggregation
• geometry/russia_gauges.gpkg – coordinates of hydrological observation stations

  • 	gauge_id	name_ru	name_en	geometry
    0	49001	р. Ковда – пос. Софпорог	r.Kovda - pos. Sofporog	POINT (31.41892 65.79876)
    1	49014	р. Корпи-Йоки – пос. Пяозерский	r.Korpi-Joki - pos. Pjaozerskij	POINT (31.05794 65.77917)
    2	49017	р. Тумча – пос. Алакуртти	r.Tumcha - pos. Alakurtti	POINT (30.33082 66.95957)

• geometry/russia_ws.gpkg – catchments polygon for each hydrological observation stations

  • 	gauge_id	name_ru	name_en	new_area	ais_dif	geometry
    0	9002	р. Енисей – г. Кызыл	r.Enisej - g.Kyzyl	115263.989	0.230	POLYGON ((96.87792 53.72792, 96.87792 53.72708...
    1	9022	р. Енисей – пос. Никитино	r.Enisej - pos. Nikitino	184499.118	1.373	POLYGON ((96.87792 53.72708, 96.88042 53.72708...
    2	9053	р. Енисей – пос. Базаиха	r.Enisej - pos.Bazaiha	302690.417	0.897	POLYGON ((92.38292 56.11042, 92.38292 56.10958...

  • Column ais_diff is corresponded to % error in area definition
• nc_all_q
  • netcdf files for hydrological observation stations which has no missing values on discharge for 2008-2020 period
• nc_all_h
  • netcdf files for hydrological observation stations which has no missing values on level for 2008-2020 period
• nc_all_q_h
  • netcdf files for hydrological observation stations which has no missing values on discharge and level for 2008-2020 period
• nc_concat
  • data for all available geometry provided in dataset

More details on processing scripts which were used for development of this database can be found in folder of GitHub repository where I store results for my PhD dissertation

05.04.2023 – Significant data changes. Removed catchments and related files that have more than ±15% absolute error in calculated area relative to AIS GMVO information. Now these are data for 1886 catchments across the Russia.

17.05.2023 – Significant data changes. Major review of parsing algorithm for AIS GMVO data. Fixed the way of how 0.0xx values were read. Use previous versions with caution.

11.10.2023 – Significant data changes. Added 278 catchments for CIS region from GRDC resource. Calculate meteorological and environmental attributes for each catchment. New folder /nc_all_q_h with no missing observations on discharge and level. Now these are data for 2164 catchments across CIS.

This is REDB-BR, the Rainfall Erosivity Database for Brazil from the MSWEP rainfall dataset.

It provides the R factor from the Universal Soil Loss Equation (USLE) in a 0.1º resolution grid, developed with 37 years of rainfall data from the MSWEP dataset.

The R factor was calculated trough 73 erosivity index regression equations, which mostly uses a relation between monthly precipitation and annual precipitation, the Modified Fournier Index (MFI), and represents a good approximation to locals with no sub-hourly data for long periods.

The main product of REDB-BR is the R factor map, available also as a .tif raster. The database also includes the equations shapefile, Thiessen Polygons shapefile and the equations table.

Datasets were generated by the University of East Anglia Climatic Research Unit (CRU), the University of Delaware Terrestrial Precipitation (UDEL), the Global Precipitation Climatology Centre (GPCC), the European Centre for Medium-Range Weather Forecasts Reanalysis v5 (ERA5), the Swiss Federal Institute for Forest, Snow and Landscape Research (CHELSA), the Center for Hydrometeorology and Remote Sensing at the University of California (PERSIANN), and the Department of Civil and Environmental Engineering at Princeton University (MSWEP).

Morphological properties, collocated synoptic conditions, and collocated rainfall for mesoscale convective systems in 1) the ISCCP Convective Tracking (CT) dataset with coincident data from the ERA-Interim (ERA-I) reanalysis and the Multi-Source Weighted-Ensemble Precipitation (MSWEP) product and 2) long-channel radiative-convective equilibrium (RCE) simulations in the System for Atmospheric Modeling (SAM). ISCCP_tracking_colloc.tar.gz - NetCDF files by year from 2000 to 2004 inclusive including ISCCP-CT morphological properties of MCSs, a series of collocated synoptic variables from ERA-5 (including specific humidity, temperature, vertical velocity, and cloud condensate profiles), and collocated precipitation intensity and accumulation from MSWEP. RCE_colloc_execution1.tar.gz - NetCDF files including RCE-SAM extent of MCSs (RCE_COL_cluster-sizes_*.nc), as well as collocated synoptic variables. Collocation of synoptic variables is performed for these files by extracting and averaging the variables over grid cells where the precipitation is greater than either its mean (RCE_COL_MEAN_*.nc) or its 99th percentile (RCE_COL_99_*.nc). RCE_colloc_execution2.tar.gz - NetCDF files including RCE-SAM extent of MCSs (RCE_COL_cluster-sizes_*.nc), as well as collocated synoptic variables. Collocation of synoptic variables is performed for these files by taking either the mean (RCE_COL_MEAN_*.nc) or the 99th percentile (RCE_COL_99_*.nc) value over all grid cells within the MCS.For the NetCDF files from RCE output, the numeric value in the file name is the corresponding sea surface temperature from 280 to 310 K.

Drought duration strongly depends on the definition thereof. In meteorology, dryness is habitually measured by means of fixed thresholds (e.g. 0.1 or 1 mm usually define dry spells) or climatic mean values (as is the case of the Standardised Precipitation Index), but this also depends on the aggregation time interval considered. However, robust measurements of drought duration are required for analysing the statistical significance of possible changes. Herein we have climatically classified the drought duration around the world according to their similarity to the voids of the Cantor set. Dryness time structure can be concisely measured by the n-index (from the regular/irregular alternation of dry/wet spells), which is closely related to the Gini index and to a Cantor-based exponent. This enables the world’s climates to be classified into six large types based upon a new measure of drought duration. We performed the dry-spell analysis using the full global gridded daily Multi-Source Weighted-Ensemble Precipitation (MSWEP) dataset. The MSWEP combines gauge-, satellite-, and reanalysis-based data to provide reliable precipitation estimates. The study period comprises the years 1979-2016 (total of 45165 days), and a spatial resolution of 0.5º, with a total of 259,197 grid points.

FILES

1. "drought_class" (geotiff)

2. "legend_drought_class" (csv): legend values for drought classification.

3. "rasterbrick_index_HurstCantorGini" (geotiff): raster with three layers (Hurst, Cantor and Gini Index applied to dry spells).

4. "rasterbrick_nindex_spells" (geotiff): raster with four layers (Dry Spell Spells n-index, maximum expected dry spellY1 , mean dry spell and mean wet spell).

Projection: "+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs" (EPSG.4326)

The climatic water content is calculated as previously described [1, 2] using the average number of consecutive dry days obtained from precipitation timeseries (MSWEP [3]), potential evapotranspiration [4] (based on WorldClim [5]) and soil information (SoilGrids [6]).
The maps are generated at 0.1° resolution with a global extent of -180 to 180 °E and -60 to 90 °N.

References
[1] Bickel, Samuel, Xi Chen, Andreas Papritz, and Dani Or. “A Hierarchy of Environmental Covariates Control the Global Biogeography of Soil Bacterial Richness.” Scientific Reports 9, no. 1 (August 20, 2019): 1–10. https://doi.org/10.1038/s41598-019-48571-w.
[2] Bickel, Samuel, and Dani Or. “Soil Bacterial Diversity Mediated by Microscale Aqueous-Phase Processes across Biomes.” Nature Communications 11, no. 1 (January 8, 2020): 1–9. https://doi.org/10.1038/s41467-019-13966-w.
[3] Beck, Hylke E., Eric F. Wood, Ming Pan, Colby K. Fisher, Diego G. Miralles, Albert I. J. M. van Dijk, Tim R. McVicar, and Robert F. Adler. “MSWEP V2 Global 3-Hourly 0.1° Precipitation: Methodology and Quantitative Assessment.” Bulletin of the American Meteorological Society 100, no. 3 (March 2019): 473–500. https://doi.org/10.1175/BAMS-D-17-0138.1.
[4] Jensen, M. E., and H. R. Haise. “Estimating Evapotranspiration from Solar Radiation.” Proceedings of the American Society of Civil Engineers, Journal of the Irrigation and Drainage Division 89, no. 0023 (1963): 15–41.
[5] Fick, Stephen E., and Robert J. Hijmans. “WorldClim 2: New 1-Km Spatial Resolution Climate Surfaces for Global Land Areas: New Climate Surfaces for Global Land Areas.” International Journal of Climatology 37, no. 12 (October 2017): 4302–15. https://doi.org/10.1002/joc.5086.
[6] Hengl, Tomislav, Jorge Mendes de Jesus, Gerard B. M. Heuvelink, Maria Ruiperez Gonzalez, Milan Kilibarda, Aleksandar Blagotić, Wei Shangguan, et al. “SoilGrids250m: Global Gridded Soil Information Based on Machine Learning.” PLOS ONE 12, no. 2 (February 16, 2017): e0169748. https://doi.org/10.1371/journal.pone.0169748.

This NetCDF4 dataset contains gridded rainfall estimates created from a blend of Global Satellite Mapping of Precipitation (GSMaP) satellite rainfall and Australian Gridded Climate Dataset (AGCD) rain gauge analysis data. The blending process consisted of a two-step method. The first step involved correcting the data through the use of multiplicative ratio grids. For each month, the ratio of the satellite data to the rain gauge data was found at each station. These ratios were then converted into a grid using Ordinary Kriging. The ratio grid was then applied onto the original GSMaP data to form the corrected GSMaP data. The second step involved blending the corrected GSMaP data and AGCD data. The blend is formed from the weighted average of the two datasets using weights derived from their error variances. The weights were inversely proportional to the error variances of the respective datasets. The error variances were calculated on a seasonal basis using the Multi-Source Weighted-Ensemble Precipitation (MSWEP) dataset as truth. The weighted average is the final blended product. The temporal coverage of the data spans a total of 20 years from January 2001 to December 2020, on a monthly basis. The spatial domain of the data is a rectangular domain centred over Australia. The latitude ranges from 108 to 156 degrees east while the longitude ranges from -45 to -9 degrees north. The resolution is 0.1 degrees. The data was created in an attempt to provide better representation of rainfall away from rain gauges whilst retaining strong correlations to rain gauges where they exist. The algorithm described earlier was performed using Python 3. This is version 1 of the data. Refinements are planned in the future.

The Lagrangian Tropical Cyclone Precipitation Estimates and Moisture Sources (LagTCPMoS) dataset compiles Lagrangian precipitation estimates and moisture sources for all tropical cyclones (TCs) that occurred in the North Atlantic basin from 1980 to 2023 (Landsea and Franklin, 2013). This dataset was created using the Lagrangian moisture and heat tracking (LATTIN) tool (Pérez-Alarcón et al., 2024), which backtracked up to 10 days the air parcels residing over the TC area (delimited by a fixed and symmetrical radius of 500 km) at each 6-hourly time interval. The air parcel trajectories were filtered from the global outputs (Vázquez et al., 2024) of the Lagrangian particle dispersion model FLEXPART (Pisso et al., 2019).

To identify moisture sources, we applied the source attribution method based on the one proposed by Sodemann et al. (2008). We also followed the specific threshold values for TCs found by Pérez-Alarcón et al. (2025). Additionally, we utilized the high-resolution Multi-Source Weighted-Ensemble Precipitation (MSWEP) dataset (Beck et al., 2019) to bias-correct the moisture sources and Lagrangian precipitation estimates, as detailed in Pérez-Alarcón et al. (2025).

For each TC, LagTCPMoS has a NetCDF file containing the 6-hourly Lagrangian precipitation estimates and moisture uptake. The file has been named according to the TC code provided by the HURDAT2 dataset (Landsea and Franklin, 2013).

This database is fully described in Pérez-Alarcón, A.; Trigo, R.M.; Nieto, R.; and Gimeno, L. (2025). Lagrangian Tropical Cyclone Precipitation Estimates and Moisture Sources (LagTCPMoS) Dataset. Scientific Data. under review

Morphological properties, collocated synoptic conditions, and collocated rainfall for mesoscale convective systems in 1) the ISCCP Convective Tracking (CT) dataset with coincident data from the ERA-Interim (ERA-I) reanalysis and the Multi-Source Weighted-Ensemble Precipitation (MSWEP) product and 2) long-channel radiative-convective equilibrium (RCE) simulations in the System for Atmospheric Modeling (SAM).

ISCCP_tracking_colloc.tar.gz - NetCDF files by year from 1983 to 2008 including ISCCP-CT morphological properties of MCSs, a series of collocated synoptic variables from ERAI (including sea surface temperature, winds, specific humidity profiles, cloud condensate profiles, etc.), and collocated precipitation intensity and accumulation from MSWEP.

RCE_colloc_execution1.tar.gz - NetCDF files including RCE-SAM extent of MCSs (RCE_COL_cluster-sizes_*.nc), as well as collocated synoptic variables. Collocation of synoptic variables is performed for these files by extracting and averaging the variables over grid cells where the precipitation is greater than either its mean (RCE_COL_MEAN_*.nc) or its 99th percentile (RCE_COL_99_*.nc).

RCE_colloc_execution2.tar.gz - NetCDF files including RCE-SAM extent of MCSs (RCE_COL_cluster-sizes_*.nc), as well as collocated synoptic variables. Collocation of synoptic variables is performed for these files by taking either the mean (RCE_COL_MEAN_*.nc) or the 99th percentile (RCE_COL_99_*.nc) value over all grid cells within the MCS.

For the NetCDF files from RCE output, the numeric value in the file name is the corresponding sea surface temperature from 280 to 310 K.

The Multimetric Drought Dataset for the Greater Antilles (Cuba, Jamaica, La Española, and Puerto Rico) provides monthly drought related products derived from the Standardized Precipitation Index (SPI; McKee et al., 1993), calculated using MSWEP precipitation data (Beck et al., 2019) at 0.1° spatial resolution for the period January 1980 to December 2023. The SPI was computed for accumulation periods ranging from 1 to 24 months, producing both spatial and temporal outputs. For accumulation periods longer than one month, data availability begins after the corresponding delay (e.g., SPI-12 starts in January 1981).

To define different drought severity levels moderate (SPI < –0.84), severe (SPI < –1.28), and extreme (SPI < –1.65) the classification proposed by Agnew (2000) was applied, and the corresponding affected areas were calculated accordingly.

The repository is organized into directories by variable type and temporal scale for each island, as follows:

SPI Spatial drought (SPI1 to SPI24): provided in NetCDF (.nc) format.

SPI Temporal drought (SPI1 to SPI24): provided in CSV format.

Monthly percentage of area affected by drought, categorized as Moderate Drought (MD), Severe Drought (SD), and Extreme Drought (ED), in CSV format. (Available for SPI1, SPI3, SPI6, SPI12, SPI18, and SPI24)

Drought event catalogues, detailing start, end, severity, and duration of each event, in CSV format. (Available for SPI1, SPI3, SPI6, SPI12, SPI18, and SPI24)

Percentage of area affected during drought events, categorized by severity level (MD, SD, and ED), in CSV format. (Available for SPI1, SPI3, SPI6, SPI12, SPI18, and SPI24)

Derived indicators directory for SPI1, SPI3, SPI6, SPI12, SPI18, and SPI24, providing insights into drought temporal dynamics, spatial synchrony, and abrupt hydroclimatic regime shifts across the Greater Antilles. This includes:

DPI (Drought Persistence Index) – measures the average duration of drought events.

OSI (Onset Severity Index) – quantifies how quickly droughts intensify after onset.

DER (Drought Event Ratio) – captures the asymmetry between drought development and recovery phases.

Co-occurrence matrices of drought years across the islands – highlighting temporal synchronicity.

Whiplash transition records (MD, SD, and ED) – identifying and characterizing abrupt dry-to-wet and wet-to-dry shifts.

Moisture Analysis: This directory includes time series (in CSV format) of moisture contributions to precipitation (E − P < 0), derived from a forward analysis of air parcels originating from the Caribbean and North Atlantic (NATL) moisture sources, as well as from each island individually. It also includes the corresponding monthly anomalies.

This database is fully described in:

Stojanovic, M.; Sorí, R.; Pérez-Alarcón, A.; Nieto, R.; and Gimeno, L. (2025). Multimetric drought dataset for the Greater Antilles: a resource for environmental and adaptation studies. Scientific Data. Submitted.

The dataset ‘Rainfall Associated with Flood Fatalities in the Euro-Mediterranean Region (1980–2020)’ links rainfall extracted from the MSWEP database to flood fatalities recorded in FFEM-DB. The FFEM-DB provides detailed information on each fatality, including geographical locations. The dataset includes rolling maximum precipitation values for multiple durations (3h, 6h, 12h, 24h, and 48h) and daily rainfall (5-day window) at each FF location, allowing for a detailed assessment of rainfall intensity leading to fatal flood events. The data support the findings of the peer-reviewed article: Papagiannaki, K., Kotroni, V., Lagouvardos, K., Diakakis, M. & Kyriakou, P. Rainfall patterns and their association with flood fatalities across diverse Euro-Mediterranean regions over 41 years. npj Nat. Hazards 2, 39 (2025).

https://doi.org/10.1038/s44304-025-00095-2

Notes for the 13.10.2019 update- The months of January 2019 to July 2019 were added to the ERA5 reconstructions- The robustness of the ERA5 reconstruction was improved for a few Greenland and Antarctica mascons by better handling a special case occuring when air temperature is always lower than 0°C during the calibration period.- The updated ERA5 time series might differ from the previous version (especially individual ensemble members). With the exception of the special case mentioned above, these differences are not significant.List of all filesReadme file 00_readme.txtMonthly grids - ensemble means 01_monthly_grids_ensemble_means_allmodels.zipMonthly grids - ensembles, model 1 to 6 02_monthly_grids_ensemble_JPL_MSWEP_1979_2016.zip 02_monthly_grids_ensemble_JPL_GSWP3_1901_2014.zip 02_monthly_grids_ensemble_JPL_ERA5_1979_201907.zip 02_monthly_grids_ensemble_GSFC_MSWEP_1979_2016.zip 02_monthly_grids_ensemble_GSFC_GSWP3_1901_2014.zip 02_monthly_grids_ensemble_GSFC_ERA5_1979_201907.zipDaily grids - ensemble means, model 1 to 6 03_daily_grids_ensemble_means_JPL_MSWEP_1979_2016.zip 03_daily_grids_ensemble_means_JPL_GSWP3_1901_2014.zip 03_daily_grids_ensemble_means_JPL_ERA5_1979_201907.zip 03_daily_grids_ensemble_means_GSFC_MSWEP_1979_2016.zip 03_daily_grids_ensemble_means_GSFC_GSWP3_1901_2014.zip 03_daily_grids_ensemble_means_GSFC_ERA5_1979_201907.zipGlobal averages - daily and monthly time series 04_global_averages_allmodels.zipContent of readmeGRACE TWS Reconstruction (GRACE_REC_v03)The dataset contains reconstructed time series of daily and monthly anomalies of terrestrial water storage (TWS) based on two different GRACE solutions and three different meteorological forcing datasets. There is a total of 6 different models:JPL_MSWEP - trained with GRACE JPL mascons, forced with MSWEP forcing (1979-2016)JPL_GSWP3 - trained with GRACE JPL mascons, forced with GSWP3 forcing (1901-2014)JPL_ERA5 - trained with GRACE JPL mascons, forced with ERA5 forcing (1979-present)GSFC_MSWEP - trained with GRACE GSFC mascons, forced with MSWEP forcing (1979-2016)GSFC_GSWP3 - trained with GRACE GSFC mascons, forced with GSWP3 forcing (1901-2014)GSFC_ERA5 - trained with GRACE GSFC mascons, forced with ERA5 forcing (1979-present)The reconstruction aims at reproducing the sub-decadal climate-driven variability observed in the GRACE data. Seasonal cycle and human impacts on TWS are not reconstructed. A GRACE-based seasonal cycle is provided for convenience. Long-term signals (trends over a period >15 years) are removed during the model calibration procedure but are still present in the final dataset and mainly represent precipitation-driven trends. The interpretation of the reconstructed long-term trends should be done with the awareness that there can be some uncertainty in the reconstructed trends.For most applications, uncertainty ranges can be derived from the 100 ensemble members available for each model.The grids are stored in NetCDFv4 files in units of mm (kg m^-2). Although the data is provided on a 0.5 degrees grid, the effective spatial resolution should be considered to be 3 degrees, similar to the original resolution of the GRACE datasets. This might need to be taken into account when comparing this dataset against other sources.The global means are stored as csv files in units of Gt of water. To convert back to mm of water, use the land area values given in the reference paper below.When using this dataset, please cite:Humphrey, V., & Gudmundsson, L. (2019). GRACE-REC: a reconstruction of climate-driven water storage changes over the last century. Earth System Science Data, 11(3), 1153-1170.Vincent Humphrey, October 2019California Institute of TechnologyYour feedback is always welcome:vincent.humphrey[-a-t-]caltech.edu (vincent.humphrey[-a-t-]bluewin.ch) Abstract

The amount of water stored on continents is an important constraint for water mass and energy exchanges in the Earth system and exhibits large inter-annual variability at both local and continental scales. From 2002 to 2017, the satellites of the Gravity Recovery and Climate Experiment mission (GRACE) have observed changes in terrestrial water storage (TWS) with an unprecedented level of accuracy. In this paper, we use a statistical model trained with GRACE observations to reconstruct past climate-driven changes in TWS from historical and near real time meteorological datasets at daily and monthly scales. Unlike most hydrological models which represent water reservoirs individually (e.g. snow, soil moisture, etc.) and usually provide a single model run, the presented approach directly reconstructs total TWS changes and includes hundreds of ensemble members which can be used to quantify predictive uncertainty. We compare these data-driven TWS estimates with other independent evaluation datasets such as the sea level budget, large-scale water balance from atmospheric reanalysis and in-situ streamflow measurements. We find that the presented approach performs overall as well or better than a set of state-of-the-art global hydrological models (Water Resources Reanalysis version 2). We provide reconstructed TWS anomalies at a spatial resolution of 0.5°, at both daily and monthly scales over the period 1901 to present, based on two different GRACE products and three different meteorological forcing datasets, resulting in 6 reconstructed TWS datasets of 100 ensemble members each. Possible user groups and applications include hydrological modelling and model benchmarking, sea level budget studies, assessments of long-term changes in the frequency of droughts, the analysis of climate signals in geodetic time series and the interpretation of the data gap between the GRACE and the GRACE Follow-On mission.Check reference for additional details and caveats.ReferenceHumphrey, V., & Gudmundsson, L. (2019). GRACE-REC: a reconstruction of climate-driven water storage changes over the last century. Earth System Science Data, 11(3), 1153-1170.

The rainfall data set of the southern Qinghai Tibet Plateau is fused by the satellite and the ground station. The data is in ASCII format, with a temporal resolution of 1 day and a horizontal spatial resolution of 0.1 °， The time coverage is from June 10 to October 31 in 2014-2019, which can provide driving data for rainfall verification and hydrological simulation in the southern Tibetan Plateau. The data set is based on the rainfall data of China Meteorological Administration and Hydrological Bureau of the Ministry of water resources after strict quality control , which is the highest density ground station network in the region so far. Dynamic Bayesian Model Average method is used to merge satellite precipitation products, i.e., GPM-IMERG, GSMaP, and CMORPH, based on the likelihood measurements of a high-density rainfall gauge network. The statistical accuracy evaluation and hydrological simulation verification of the merged data preforms better than the source satellite data, and also better than the popular reanalysis data CHIRPS and MSWEP.

Hydrological variables (evapotranspiration, precipitation, runoff, and terrestrial water storage) are distributed through several datasets that do not have the same characteristics (spatial and temporal coverage, spatial and temporal resolution). This dataset aims at providing a coherent gathering of all datasets available as of January 2021, covering at least the period from 2003 to 2014, and from 50°S to 50°N.

This file contains:

14 datasets for evapotranspiration (ERA5-Land, FLUXCOM, GLDAS2.2 CLSM2.5, GLDAS2.1 CLSM2.5, GLDAS2.1 NOAH3.6, GLDAS VIC4.1.2, GLDAS2.0 CLSM2.5, GLDAS2.0 NOAH3.6, GLDAS2.0 VIC4.1.2, GLEAM, JRA55, MERRA2, MOD16, and SEBBop)

11 datasets for precipitation (CPC, CRU, ERA5-Land, PGF, GPCC, GPCP, GPM, JRA55, MERRA2, MSWEP, and TRMM)

11 datasets for runoff (ERA5-Land, GLDAS2.2 CLSM2.5, GLDAS2.1 CLSM2.5, GLDAS2.1 NOAH3.6, GLDAS VIC4.1.2, GLDAS2.0 CLSM2.5, GLDAS2.0 NOAH3.6, GLDAS2.0 VIC4.1.2, GRUN, JRA55, and MERRA2)

2 datasets for terrestrial water storage (GRACE CSR mascons, and GRACE JPL mascons)

Each dataset is given as the original version (on a regular grid) and also as a post-treated file averaged for 189 river basins (whose borders are also provided).

Current global multi-source merged precipitation datasets can facilitate better utilization of the complementary nature of gauge-, satellite-, and reanalysis-based precipitation estimates, particularly for capturing precipitation variability. However, merging these datasets at high resolutions of 1-hourly and 0.1° on a full global scale remains a substantial challenge for the scientific community owing to high spatiotemporal heterogeneities. This study proposed a merging-and-calibration framework to optimally integrate the advantages of gauge-, satellite-, and model-based precipitation estimates, focusing on precipitation occurrences and providing a new fully Global multi-source Merging-and-Calibration Precipitation dataset (GMCP: 1-hourly, 0.1°, global, 2000–Present). The main conclusions included: (1) GMCP generally outperformed the input datasets, ERA5-Land, GSMaP-MVK, and IMERG-Late, across various spatiotemporal scales, both in regional statistics and extreme precipitation systems; (2) GMCP significantly outperformed IMERG-Final, calibrated by gauge analysis at the monthly scale, with the improvements in correlation coefficient (CC), root mean square error (RMSE), and Heidke skill score (HSS) by approximately 66.67%, 39.25%, and 26.83%, respectively, from 2016 to 2020 over the Continental United States (CONUS); (3) compared to the state-of-the-art multi-source merged product with a daily gauge correction scheme, MSWEP V2 (3-hourly and 0.1°), GMCP demonstrated the notable improvements with an approximately 20% enhancement in accurately capturing the precipitation occurrences against approximately 67, 000 rain gauges over Mainland China in 2016; (4) in comparison to another well-known multi-source merged quasi-global daily and 0.05° precipitation product, CHIPRS integrating the gauge-, satellite-, and reanalysis-based precipitation estimates, GMCP also demonstrated the notable improvements at the daily scale, achieving the increases in CC, RMSE, and HSS by around 57.45%, 38.18%, and 75.76%, respectively, against approximately 67, 000 rain gauges over Mainland China in 2016; and (5) this framework was suitable for generating the fully global precipitation datasets at 1-hourly and 0.1° scales, significantly mitigating the inherent shortcomings of each input dataset, with GMCP demonstrating the great potential as a valuable resource for worldwide scientific research and societal applications.

Data used for analysis in "Explainable Clustering Applied to the Definition of Terrestrial Biome" - using Decision Tree and Clustering techniques to identify biomes.

Land surface properties:

• TreeCover - Vegetation Continuous Fields (VCF) collection 6 fractional tree cover from ¹, regridded as per ².
• urban cover from the History Database of the Global Environment, Version 3.1 (HYDE) ^3,4
• crop cover (from HYDE)

• pas - Pasture Cover (from HYDE)PopDen (population density from HYDE)
• BurntArea_xxxxx - Burnt area with xxxx denoting different products, provided by fireMIP ^5–7:
  • GFED_four: Global Fire Emissions Database, Version 4 (GFED4) ⁸
  • GFED_four_s: Global Fire Emissions Database, Version 4.1, including small fires (GFEDv4.1) ⁹
  • MCD_forty_five: MCD45 ¹⁰
  • Meris: Fire_CCI4.0 ¹¹
  • MODIS: Fire_CCI5.1 ¹²

Climate:

• MAP_xxx - Mean annual precipitation where xxx denotes data source:
  • CMORPH ^13,14
  • CRU from version 4.03 of the Climatic Research Unit Time Series high-resolution gridded dataset (CRU TS v4.01) ¹⁵
  • GPCC: ¹⁶
  • MSWEP: ¹⁷
• MAT - Mean annual temperature from CRU)
• MConc_xxx – Mean annual concentration of rainfall as defined by ¹⁸, where xxx denotes precip data source (see “MAP_xxx”)
• MADD_xxx- Mean annual fractional dry days from CRU - i.e. seasonality of rainfall), where xxx denotes precip data source (see “MAP_xxx”)
• MDDM_xxx – Mean fractional dry days of the driest month.
• MADM_xxx – Mean annual precipitation of the driest month.
• MTWM - Mean Maximum Temperature of the warmest month from CRU
• MTCM - Mean minimum temperature of the coldest month from CRU
• SW1 - direct downwards SW simulated using the SLASH model using CRU cloud cover
• SW2 - diffuse downwards SW simulated using the SLASH model using CRU cloud cover
• MaxWind (Mean Max Windspeed from CRU-(National Centers for Environmental Prediction ¹⁵

‘output_summary’ contains framework output. There are several directories for different experiments, each containing a netcdf file. Along with standard latitude and longitude,each file contains ‘model_level_number’ dimension, with each layer representing the 1, 5, 10, 25, 50, 75, 90, 95 and 99% quantiles of the model posterior. The folder represents the experiment:

• Control – standard full model reconstruction
• noHumans – without human influence (from crop, pasture, population density or urban influence)
• noMortality – without disturbance stress (burnt area, wind, heat stress, rainfall seasonality
• noMAP – without mean annual precip influence.
• noNoneMAT – without mean annual temperature influence.
• noFire – tree cover without the influence of fire
• noDrought – without the influence of rainfall distribution
• noTasMort – without mortality from heat stress
• noWind – without influence from max. windspeed
• noPas – without exclusion from pasture
• noCrop – without exclusion from crop
• noPop – without reduction from population density
• noUrban – without exclusion from urban
• firePlus1pc – tree cover with burnt area was 1% higher.

References

1. Dimiceli, C. & Others. MOD44B MODIS/Terra Vegetation Continuous Fields Yearly L3 Global 250m SIN Grid V006 (NASA EOSDIS Land Processes DAAC, 2015). Preprint at (2015).

2. Kelley, D. I. et al. How contemporary bioclimatic and human controls change global fire regimes. Nat. Clim. Chang. 9, 690–696 (2019).

3. Klein Goldewijk, K., Goldewijk, K. K., Beusen, A., Van Drecht, G. & De Vos, M. The HYDE 3.1 spatially explicit database of human-induced global land-use change over the past 12,000 years. Glob. Ecol. Biogeogr. 20, 73–86 (2010).

4. Hurtt, G. C. et al. Harmonization of land-use scenarios for the period 1500–2100: 600 years of global gridded annual land-use transitions, wood harvest, and resulting secondary lands. Climatic Change vol. 109 117–161 Preprint at https://doi.org/10.1007/s10584-011-0153-2 (2011).

5. Hantson, S., Arneth, A., Harrison, S. P. & Kelley, D. I. The status and challenge of global fire modelling. (2016).

6. Hantson, S. et al. Quantitative assessment of fire and vegetation properties in simulations with fire-enabled vegetation models from the Fire Model Intercomparison Project. Geoscientific Model Development vol. 13 3299–3318 Preprint at https://doi.org/10.5194/gmd-13-3299-2020 (2020).

7. Rabin, S. S., Melton, J. R. & Lasslop, G. The Fire Modeling Intercomparison Project (FireMIP), phase 1: experimental and analytical protocols with detailed model descriptions. Geoscientific Model (2017).

8. Giglio, L., Randerson, J. T. & van der Werf, G. R. Analysis of daily, monthly, and annual burned area using the fourth-generation global fire emissions database (GFED4). J. Geophys. Res. Biogeosci. 118, 317–328 (2013).

9. van der Werf, G. R. et al. Global fire emissions estimates during 1997–2016. Earth Syst. Sci. Data 9, 697–720 (2017).

10. Roy, D. P., Boschetti, L., Justice, C. O. & Ju, J. The collection 5 MODIS burned area product — Global evaluation by comparison with the MODIS active fire product. Remote Sensing of Environment vol. 112 3690–3707 Preprint at https://doi.org/10.1016/j.rse.2008.05.013 (2008).

11. Alonso-Canas, I. & Chuvieco, E. Global burned area mapping from ENVISAT-MERIS and MODIS active fire data. Remote Sens. Environ. 163, 140–152 (2015).

12. Chuvieco, E. et al. Generation and analysis of a new global burned area product based on MODIS 250 m reflectance bands and thermal anomalies. Earth System Science Data vol. 10 2015–2031 Preprint at https://doi.org/10.5194/essd-10-2015-2018 (2018).

13. Joyce, R. J., Janowiak, J. E., Arkin, P. A. & Xie, P. CMORPH: A Method that Produces Global Precipitation Estimates from Passive Microwave and Infrared Data at High Spatial and Temporal Resolution. J. Hydrometeorol. 5, 487–503 (2004).

14. Marthews, T. R., Blyth, E. M., Martínez-de la Torre, A. & Veldkamp, T. I. E. A global-scale evaluation of extreme event uncertainty in the eartH2Observe project. Hydrol. Earth Syst. Sci. 24, 75–92 (2020).

15. Harris, I. C. & Jones, P. D. CRU TS4.03: Climatic Research Unit (CRU) Time-Series (TS) version 4.03 of high-resolution gridded data of month-by-month variation in climate (Jan. 1901- Dec. 2018). (2019) doi:10.5285/10D3E3640F004C578403419AAC167D82.

16. Schneider, U., Becker, A., Finger, P., Meyer-Christoffer, A. & Ziese, M. GPCC Full Data Monthly Product Version 2018 at 0.5◦: Monthly Land-Surface Precipitation from Rain-Gauges Built on GTS-Based and Historical Data. Deutscher Wetterdienst: Offenbach am Main, Germany (2018).

17. Beck, H. E., Van Dijk, A. & Levizzani, V. MSWEP: 3-hourly 0.25 global gridded precipitation (1979-2015) by merging gauge, satellite, and reanalysis data. Hydrol. Earth Syst. Sci. (2017).

18. Kelley, D. I., Harrison, S. P., Wang, H. & Simard, M. A comprehensive benchmarking system for evaluating global vegetation models. (2013).

Tropical river basins – crucial components of global water and carbon cycles – are threatened by logging, mining, agricultural conversion, and climate change. Thus, decision-makers require hydrological impact assessments to sustainably manage threatened basins, such as the ∼68,000 km2 Essequibo River basin in Guyana. Emerging global data products offer the potential to better understand sparsely-gauged basins. We combined new global meteorological and soils data with established in situ observations to build the first physically-based spatially-distributed hydrological model of the Essequibo. We developed new, open source, methods to translate global data (ERA5-Land, WFDE5, MSWEP, and IMERG) into a grid-based SHETRAN model. Comparing the performance of several global and local precipitation and evaporation datasets showed that WFDE5 precipitation, combined with ERA5-Land evaporation, yielded the best daily discharge simulations from 2000 to 2009, with close water balances (PBIAS = −3%) and good discharge peaks (NSE = 0.65). Finally, we tested model sensitivity to key parameters to show the importance of actual to potential evapotranspiration ratios, Strickler runoff coefficients, and subsurface saturated hydraulic conductivities. Our data translation methods can now be used to drive hydrological models nearly anywhere in the world, fostering the sustainable management of the Earth’s sparsely-gauged river basins.