Want to see VISTA flood frequency directly in QGIS?
Choose a period and paste its raster URL into QGIS. Zoom to your region without first downloading the global file.
Open a layer in QGIS
VISTA data access FAO STAC
Open a seasonal flood-frequency layer in QGIS, or read just your study area with Python. FAO’s STAC catalogue connects each layer’s metadata to a public GeoTIFF, so software can find and open the data without an account.
Flood-any frequency only (freq_flood_any): how often a pixel was flagged as flooded, at any severity, among its valid observations across 2012–2026. Values run from 0 to 1; 0.25 means 25%.
These are multi-year seasonal references. For dated flood observations, counts, dominant class or other frequency bands, use the full VISTA download bucket.
Choose a period and paste its raster URL into QGIS. Zoom to your region without first downloading the global file.
Open a layer in QGISFind a layer through STAC, then read a small area with rasterio. Start from the ready-to-run example below.
Use the Python exampleChoose a product and period, then copy its URL into QGIS. You can also download the entire global file if you need a local copy.
The picker needs JavaScript. A URL example is available under “Access without the picker” below.
03P1 spans 21 February–20 March; 01P1 spans 21 December–20 January. Available source dekads determine coverage at archive edges.N_valid in the matching count product. The sampled FAO files have no declared NoData value; a numeric value alone cannot establish valid coverage. See the interpretation guidance.Select the item by ID, then read a small region. This example uses a box in Sofala, Mozambique; it reads numeric values and does not establish observation support.
Install: pip install pystac-client rasterio. For monthly data, replace DEKAD with MONTHLY in both IDs, and 03D2 with 03P2.
from pystac_client import Client
import rasterio
from rasterio.windows import from_bounds
catalogue = Client.open("https://data.apps.fao.org/geospatial/search/stac/")
collection_id = "fao-gismgr:VISTA:raster:mapsets:N-FLOOD-ANY-FREQ-DEKAD"
collection = catalogue.get_collection(collection_id)
# Use the full ID, not a datetime filter: platform dates are placeholders.
item_id = collection_id + ":items:VISTA-N-FLOOD-ANY-FREQ-DEKAD-03D2"
item = collection.get_item(item_id)
asset = next(a for a in item.assets.values() if "image/tiff" in (a.media_type or ""))
with rasterio.open(asset.href) as src:
window = from_bounds(34.5, -19.8, 34.7, -19.6, src.transform)
values = src.read(1, window=window)
print(values.shape) # Check matching N_valid before interpreting values.
fao-gismgr:VISTA:raster:mapsets:N-FLOOD-ANY-FREQ-DEKADfao-gismgr:VISTA:raster:mapsets:N-FLOOD-ANY-FREQ-MONTHLYFor downloading the full range of products, use the bucket. STAC adds standard metadata that software can read: collection descriptions, spatial footprints and links to each file. It is useful when building QGIS or Python workflows around these frequency layers.
The FAO rolling monthly copies also come as single global rasters, whereas the main bucket distributes monthly frequency as tiles. Both routes let you download GeoTIFFs; STAC does not add new observations or metrics. Remote regional reading comes from the Cloud Optimized GeoTIFF format, which the bucket’s COGs also support.
Browse all VISTA products · About the STAC standard
Each file is a single-band Float32 GeoTIFF in EPSG:4326, with approximately 375 m pixels. The sampled files use 512 × 512 pixel tiles, eight overview levels and a cloud-optimised GDAL layout. HTTP range access is supported.
Dekadal rasters are 106,753 × 44,482 pixels, reaching approximately ±75° latitude. Monthly rasters are 106,754 × 53,378 pixels: CSI mosaics the source tiles onto a global ±90° frame. That frame does not establish additional observed coverage.
The FAO catalogue labels this “Band 9 - Any Flood”. It is freq_flood_any, band 13 in VISTA’s original 14-band frequency files. Each FAO file contains only this band. The full products also provide counts and other frequency bands, including freq_0_nodata.
The processing pipeline includes the previous, centre and following dekads, crossing December/January. Missing source dekads are recorded and available ones are still combined, so some archive-edge windows are incomplete. The recurring window label describes the intended window.
Data are licensed under CC BY 4.0. Credit VISTA, FAO and JRC, retain source acknowledgements and indicate changes. Use the selected catalogue record for metadata and citation, with your access date; no release version is supplied.
Dekadal catalogue record4fd2bcff-a546-411c-987f-073c7e6e6166
Rolling monthly catalogue record1f53a6d7-9d67-4943-9991-861659320a0a
See also the VISTA citation guidance.
The catalogue’s Download button opens Google Cloud and asks for a login. Use this page’s direct links instead. STAC metadata and raster asset URLs are anonymous. Downloading retrieves a global file; opening a small region remotely avoids that full transfer, although a broad map view can still request substantial data.
QGIS accepts the HTTPS URL under its HTTP(S) protocol source. GDAL tools can use the same URL prefixed with /vsicurl/. The original VISTA downloads also include dated observations, count products and dominant class.
FAO OpenEO and its Web Editor require authentication. FAO OpenEO STAC is restricted to FAO users. See the OpenEO documentation and tutorials.
Copy a URL below into QGIS or your browser. For another period, change the month and D1–D3 or P1–P3 suffix. Credit VISTA, FAO and JRC under CC BY 4.0. The attribution dialog requires JavaScript.
Dekadal · 03D2 · 312.0 MBhttps://storage.googleapis.com/fao-gismgr-vista-data/DATA/VISTA/MAPSET/N-FLOOD-ANY-FREQ-DEKAD/VISTA.N-FLOOD-ANY-FREQ-DEKAD.03D2.tif
Rolling monthly · 03P2 · 363.7 MBhttps://storage.googleapis.com/fao-gismgr-vista-data/DATA/VISTA/MAPSET/N-FLOOD-ANY-FREQ-MONTHLY/VISTA.N-FLOOD-ANY-FREQ-MONTHLY.03P2.tif
Links checked: 5 November 2026. Sizes are for the entire global file (decimal MB/GB).