Global Surface Water download and formats: files, access methods, and GIS workflows
Global Surface Water is a global Earth-observation dataset from the European Commission Joint Research Centre that describes the location, occurrence, seasonality, recurrence, transitions, change, and maximum extent of visible surface water. The current download site provides an extended release covering 1984–2024, alongside earlier 1984–2021 assets. Data are available as tiled GeoTIFF files, full-dataset downloads, Google Earth Engine assets, and web map tiles. The dataset is valuable for regional water analysis, environmental mapping, change detection, and spatial statistics, but it is not a live flood-monitoring feed, hydrologic model, legal wetland inventory, or complete representation of subsurface or obscured water.
Global Surface Water is designed for a specific geographic question: where has open surface water appeared, and how has its presence changed over time? Produced by the European Commission Joint Research Centre from Landsat observations, it provides global raster evidence at approximately 30-metre pixel size. The download and formats question matters because the dataset is distributed through several different pathways, and those pathways are not interchangeable. GeoTIFF downloads are appropriate for local GIS analysis, Earth Engine assets are useful for cloud-based raster computation, and web map tiles are intended primarily for visualization.
How EDK fits
For: GIS analysts, Earth-observation researchers, environmental planners, hydrologists, cartographers, conservation teams, and developers building location-based applications that need historical evidence about visible surface water.
Global Surface Water is a useful environmental layer for understanding the physical world as a computable geographic system. It turns long-term satellite observations into spatial attributes that can be joined to places, watersheds, roads, buildings, infrastructure, agricultural land, protected areas, and administrative regions. In an E2-style world model, the dataset can provide historical water context around physical features: whether a road corridor crosses frequently observed water, whether development is near a changing water boundary, or how water persistence varies across a region. The connection is analytical rather than product-specific: Global Surface Water supplies raster evidence about visible surface water, while a broader world model can relate that evidence to other geographic objects and compute comparisons across locations.
- •Combine Global Surface Water occurrence, recurrence, or maximum-extent rasters with searchable roads, buildings, bridges, utilities, parks, and other physical features to rank locations by proximity to historically observed surface water.
- •Aggregate water occurrence or change values by city, watershed, administrative area, infrastructure corridor, or development region, then compare geographic exposure and environmental change across places.
- •Use the historical raster layers as one evidence layer in location-based screening, while retaining the distinction between observed surface-water patterns and authoritative hydrology, flood-zone, wetland, or regulatory data.
Dataset profile
Formats
GeoTIFF raster tiles, generally distributed as 10° by 10° files.
Full-dataset download initiated through an official Python 3 script.
Google Earth Engine image and image-collection assets.
Rendered PNG web map and WMTS-style tiles for visualization.
Separate QGIS symbology files and ISO 19139 metadata files.
Access methods
Download individual tiles from the official Global Surface Water Data Access page.
Use the official Python download script for bulk or full-dataset acquisition.
Access documented historical and newer assets through Google Earth Engine.
Use the Global Surface Water Explorer for visual inspection and temporal profiles.
Add web map or WMTS layers to desktop GIS or web maps when display, not analysis, is the objective.
Quality and limits
The dataset represents visible open surface water detected from satellite observations; it does not directly represent groundwater, subsurface water, total soil moisture, or all inundation beneath vegetation or clouds.
Observation counts and effective precision vary by location because of cloud cover, acquisition plans, and available valid observations.
The transition from Landsat Collection 1 to Collection 2 can create residual co-registration differences, including offsets approaching or exceeding one 30-metre pixel in some areas.
The JRC states that some 2024 products require merging with earlier assets to create complete historical series.
The official site identifies a known monthly-recurrence issue in v1.4 and v1.5 that has not yet been corrected.
Web map services are rendered RGB imagery and are unsuitable for analytical recovery of original raster values.
The data are not a live flood-warning service, hydraulic model, legal wetland inventory, authoritative hydrography layer, or substitute for local regulatory datasets.
Analysis workflows
Select the product according to the question: occurrence or maximum extent for broad water presence, recurrence or seasonality for temporal behavior, and transitions or change for historical comparison.
Download only intersecting GeoTIFF tiles for a local study, or use the official Python workflow for bulk acquisition; preserve the original files and release documentation.
Build a mosaic or virtual raster, confirm the coordinate reference system and value interpretation, and avoid unnecessary resampling before calculating statistics.
Use zonal statistics to summarize water occurrence, extent, or change by watershed, municipality, protected area, infrastructure corridor, or other geographic unit.
Join raster-derived summaries to roads, buildings, land use, agriculture, environmental features, or public assets to study geographic relationships and prioritize further review.
For cloud-scale processing, use Google Earth Engine assets and explicitly document which historical and recent collections were merged.
Use the explorer or WMTS layers for cartographic context, but use GeoTIFF or analytical Earth Engine data for measurements and computation.
Validate high-consequence conclusions against local hydrologic, flood, wetland, survey, or regulatory sources.
Value inside EDK
What Global Surface Water contains
Global Surface Water maps the location and temporal distribution of visible surface water using Landsat observations. The data products describe several different aspects of water behavior rather than providing one single water mask. Occurrence summarizes how often water was observed at each pixel over the analysis period. Recurrence describes repeated water presence, seasonality characterizes the months in which water appears, and transitions classify changes between water and non-water states. Other products include change intensity and maximum water extent. (global-surface-water.appspot.com)
The dataset is intended to capture open water visible from space. The project’s FAQ describes coverage that includes freshwater and saltwater areas, rivers, lakes, coastal margins, wetlands, reservoirs, flooded extraction sites, irrigated fields, and other artificial water bodies. It does not mean that every pixel is a permanently water-covered feature, nor that a detected pixel identifies legal ownership, water rights, wetland status, channel boundaries, or flood hazard. (global-surface-water.appspot.com)
- •Core products: occurrence, change, seasonality, recurrence, transitions, and maximum water extent.
- •Primary spatial resolution: approximately 30 metres per pixel.
- •Observation basis: Landsat-derived classification of visible surface water.
- •Geographic scope: global land and coastal areas represented by the product.
- •The 2024 extension uses Landsat Collection 2 for recent years, while earlier periods retain Collection 1 processing in parts of the product lineage.
Geography, temporal coverage, and update cadence
Global Surface Water is global in geographic scope, but its temporal record must be described by product version. The established JRC/GSW1_4 Earth Engine asset covers observations from 1984 through 2021. The official download page now documents a newer 1984–2024 release and identifies it as the current extended dataset available for download. (developers.google.com)
There is no fixed annual or monthly update cadence stated on the official access page. The European Commission data catalogue describes the broader dataset’s update frequency as irregular, and the current download page records a specific update on August 26, 2026 while reporting observations through 2024. Therefore, analysts should treat the release version and processing date as explicit metadata rather than assuming that the newest calendar year is present. (data.jrc.ec.europa.eu)
- •Use the 1984–2021 JRC/GSW1_4 assets when reproducibility with earlier Earth Engine workflows is required.
- •Use the newer 1984–2024 download products when the extended record is appropriate, but record the specific product version and processing lineage.
- •Do not interpret every apparent change at the Collection 1/Collection 2 transition as physical change.
- •Check valid-observation information where available because cloud cover, acquisition patterns, and observation counts vary geographically.
Download options and file formats
The official data-access page provides several delivery mechanisms. For desktop GIS and local processing, the principal option is tiled download. Individual 10° by 10° files are provided as GeoTIFF files, with separate files for the available mapped products. GeoTIFF is a practical format for QGIS, ArcGIS Pro, GDAL, rasterio, and other raster-processing environments because it preserves a georeferenced grid and can be mosaicked or clipped for a study area. (global-surface-water.appspot.com)
The site also documents a Python 3 script for downloading the full global dataset. This is more appropriate than manually downloading thousands of tiles when an analysis requires global coverage, but it introduces operational considerations such as storage, network reliability, decompression time, and tile management. The page separately lists symbology files for QGIS and ISO 19139 metadata files; it warns that the downloadable raster files themselves do not contain all metadata information. (global-surface-water.appspot.com)
- •Local analysis: GeoTIFF tiles, mosaicking, clipping, reprojection, masking, and raster statistics.
- •Large-scale cloud analysis: Google Earth Engine assets and image collections.
- •Visualization: Global Surface Water Explorer and web map or WMTS tiles.
- •Cartography: downloadable symbology files can reproduce the explorer’s visual classes, but labels and metadata may require separate supporting files.
- •Automation: the official Python downloader is preferable for repeatable bulk acquisition, subject to the instructions supplied with the release.
Suitable GIS and spatial-analysis workflows
A defensible workflow begins by choosing the question and the product. For a broad screening map, occurrence or maximum extent may be appropriate. For seasonal water, use seasonality or recurrence. For historical change, use transitions or change products, while checking the observation record and release notes. Avoid treating a rendered web tile as an analytical source; download the raster or use an analytical Earth Engine asset instead. (global-surface-water.appspot.com)
For desktop GIS, download only the tiles intersecting the study area when possible, build a virtual raster or mosaic, inspect the coordinate reference system and value range, and clip the result to the area of interest. Preserve the original files and release metadata before resampling. If the analysis combines water with vector features such as roads, parcels, buildings, protected areas, or infrastructure, align the raster grid and document whether statistics are based on pixel centers, zonal overlap, or resampled values.
- •Record product version, observation period, processing collection, download date, and source attribution.
- •Use raster values for computation; use web map tiles only for display.
- •Inspect valid observations and cloud-related coverage differences before interpreting change.
- •Avoid unnecessary resampling, especially for pixel-level change detection.
- •Validate high-consequence results against local hydrology, flood, wetland, or survey data.
License, strengths, and tradeoffs
The official JRC access page states that the data are produced under the Copernicus Programme and provided free of charge without restriction of use, subject to the applicable Copernicus Regulation. Publications, models, and data products using the data must include the required acknowledgement and citation. Published maps should include “Source: EC JRC/Google.” (global-surface-water.appspot.com)
The principal strength is consistency at global scale: the same family of raster products supports comparison across countries and regions. The long historical record also makes the dataset useful for detecting broad patterns in water persistence, expansion, contraction, and seasonality. Its multiple access methods allow analysts to choose between local GeoTIFF processing, cloud computation, and lightweight visualization.
- •Strength: global geographic coverage at approximately 30-metre pixel size.
- •Strength: multi-decade temporal record with occurrence, seasonality, recurrence, transitions, change, and extent products.
- •Strength: open access with GeoTIFF, Earth Engine, explorer, and web-map delivery options.
- •Tradeoff: update timing is irregular and product vintages do not all share identical processing histories.
- •Tradeoff: Collection 1 to Collection 2 transitions can introduce residual co-registration differences.
- •Tradeoff: visible open water is not equivalent to groundwater, total inundation, legal wetlands, flood risk, or hydrologic connectivity.
Questions teams ask
What is the best format for using Global Surface Water in QGIS or ArcGIS?
Use the downloadable GeoTIFF files. The official site provides 10° by 10° tiles for products including occurrence, change, seasonality, recurrence, transitions, and maximum water extent. Mosaic or virtualize the relevant tiles, inspect the supplied symbology and metadata files, and clip them to the study area before analysis. Rendered web map tiles are better for visualization than for raster calculations. (global-surface-water.appspot.com)
Does Global Surface Water provide data through 2024?
The current official download page documents an extended 1984–2024 release. However, not every product is delivered in the same way across the full period: the page explains that some 2022–2024 history layers must be merged with earlier assets, and the established JRC/GSW1_4 Earth Engine collection remains a 1984–2021 asset. Always record the exact product and version used. (global-surface-water.appspot.com)
Is Global Surface Water a flood map or wetland inventory?
No. It maps visible open-surface-water observations and their temporal patterns. It can support floodplain, wetland, environmental, or water-resource screening, but it is not a live flood-monitoring service, a hydraulic model, a legal wetland inventory, or an authoritative regulatory boundary. (global-surface-water.appspot.com)
Can I use Global Surface Water commercially?
The official JRC access page states that the data are provided free of charge without restriction of use under the Copernicus Programme. Users must still provide the required acknowledgement and citation, including the “Source: EC JRC/Google” attribution for published map layers. Review the official terms and citation guidance for the release you use. (global-surface-water.appspot.com)
Keep exploring
global surface water analysis workflow
A complementary workflow page for turning the Global Surface Water layers into reproducible raster, zonal-statistics, and change-detection analyses.
copernicus dem download and formats
A related Earth-observation dataset page covering another global raster foundation commonly combined with surface-water data for terrain, watershed, and environmental analysis.
Sources
Global Surface Water – Data Access
Official JRC access page documenting the 1984–2024 release, downloadable products, GeoTIFF tiles, full-dataset Python download, Earth Engine assets, web map services, licensing, attribution, and processing notes.
Open sourceGlobal Surface Water Explorer
Official JRC explorer describing the dataset’s purpose, Landsat basis, global coverage, historical record, and environmental applications.
Open sourceJRC Global Surface Water Mapping Layers, v1.4
Google Earth Engine’s official catalog entry documenting the established JRC/GSW1_4 asset, its 1984–2021 coverage, producer, citation, and access pattern.
Open sourceGlobal Surface Water FAQ
Official FAQ explaining what counts as mapped surface water, the types of natural and artificial water bodies included, observation variability, and data-policy requirements.
Open sourceGlobal Surface Water Explorer dataset — European Commission JRC Data Catalogue
European Commission metadata record identifying the publisher, dataset identifier, access route, catalogue description, and irregular update frequency.
Open source