IWMIHQ/incomati_water_accounting

Incomati Water Accounting

Water accounting for the Incomati basin at ~1 km: rainfall, total ET, green and blue ET, water yield and runoff, each as a yearly average and split into dry and wet seasons, averaged over 2003-2021. Water Accounting Plus output rebuilt from IWMI's Africa GeoPortal tile services, which serve no downloadable raster.

• Region Incomati
• Period 2003-2021
• Theme Water accounting
• Grid EPSG:4326, 357 x 235 at 0.009521933 deg (~1060 m)
• Bands 18
• Files 18 Cloud-Optimised GeoTIFFs under data/, named <band>.tif, one per band

Bands

Band	Type	Units	Nodata
yearly_rainfall	float32	mm	-9999.0
yearly_et	float32	mm	-9999.0
yearly_rainfall_et	float32	mm	-9999.0
yearly_incremental_et	float32	mm	-9999.0
yearly_water_yield	float32	mm	-9999.0
yearly_runoff	float32	mm	-9999.0
dry_rainfall	float32	mm	-9999.0
dry_et	float32	mm	-9999.0
dry_rainfall_et	float32	mm	-9999.0
dry_incremental_et	float32	mm	-9999.0
dry_water_yield	float32	mm	-9999.0
dry_runoff	float32	mm	-9999.0
wet_rainfall	float32	mm	-9999.0
wet_et	float32	mm	-9999.0
wet_rainfall_et	float32	mm	-9999.0
wet_incremental_et	float32	mm	-9999.0
wet_water_yield	float32	mm	-9999.0
wet_runoff	float32	mm	-9999.0

Read it

These are COGs, so GDAL reads them in place over HTTP -- no full download, and a windowed read fetches only the bytes it needs:

from huggingface_hub import hf_hub_url
import rasterio

url = hf_hub_url("IWMIHQ/incomati_water_accounting", "data/dry_et.tif", repo_type="dataset")
with rasterio.open(url) as src:
    print(src.width, src.height, src.crs, src.nodata)
    patch = src.read(1, window=((0, 256), (0, 256)))

Or pull the whole repo:

from huggingface_hub import snapshot_download
path = snapshot_download("IWMIHQ/incomati_water_accounting", repo_type="dataset")

Read this before using it

A 2003-2021 AVERAGE, not a time series. The period comes from the publisher's own sibling items, "Incomati Seasonal Maps (2003 - 2021)" and "Incomati Yearly Maps (2003 - 2021)". Every cell is a 19-year mean, so no year is described and no trend can be read. The dataset time spans the period only to record it. Season months are undeclared. Nothing in the service metadata or the item records says where dry ends and wet begins, so these cannot be aligned to a calendar or compared with another dataset's seasons. What is known is measured: the wet season averages 648 mm of rainfall against the dry season's 120 mm, and the two sum to the yearly layer to within 0.0003 mm. The yearly bands are redundant, not independent. Because dry + wet reproduces yearly exactly, the yearly layer carries no information the seasonal pair does not. Do not treat agreement between them as corroboration of anything. Dry-season water yield is negative on average, about -45 mm. That is P - ET with ET exceeding rainfall, which is physically ordinary for a dry season, but it means the band is not a "yield" in the sense of water available. Only 55.0% of the frame carries data -- the rest is outside the basin -- so any whole-frame statistic is meaningless. One grid, with one wrinkle: 17 of the 18 source services report their corner coordinates at full precision and Incomati_Water_Yield_yearly_avg reports its rounded to six decimals, a difference of 4e-6 degrees or about 0.44 mm on the ground -- 0.0004 of a pixel. The dataset declares the precise grid the other 17 agree on. Nothing is resampled and the discrepancy is far below anything that could matter, but it is recorded so nobody rediscovers it as a bug. Rebuilt from a tile cache rather than downloaded, so these values are whatever was published into the LERC pyramid, read at the one pyramid level that matches the services' native resolution. The ArcGIS service in each file's source tag stays the authority. The basin is transboundary -- South Africa, Eswatini and Mozambique -- so this is not a South African dataset even though its headwaters are in Mpumalanga.

Provenance

Rebuilt from IWMI Africa GeoPortal ArcGIS image services, which serve no downloadable raster: they answer exportImage with HTTP 400 because their capabilities are Image,TilesOnly. The full-resolution grid was recovered by fetching every LERC tile at the pyramid level matching the service's own native resolution, decoding, and mosaicking. Each GeoTIFF carries the originating service in its source tag, and that service -- not this copy -- is the authority for its values.

The same product is indexed in IWMI's Open Data Cube, which is the route to its footprint, time extent and band metadata as STAC:

https://explorer-production-0070.up.railway.app/stac/collections/incomati_water_accounting

Citation

International Water Management Institute (IWMI), Incomati Water Accounting. Published via the IWMI Open Data Cube. Licence CC-BY-4.0.

@misc{iwmi_incomati_water_accounting,
  title        = {Incomati Water Accounting},
  author       = {International Water Management Institute},
  howpublished = {Hugging Face Hub, IWMIHQ/incomati_water_accounting},
  note         = {Derived from IWMI Africa GeoPortal services; see Provenance},
  year         = {2026}
}