Datasetpublic

Rekin226/aquascope-gauges

AquaScope gauges The station catalog of every water-observation source AquaScope can reach, harvested on a schedule and published as GeoParquet. Last run 2026-09-23T23:27:06+00:00 with aquascope 0.18.0: 75,721 stations from 11 of 11 sources. Files: stations.parquet: GeoParquet 1.0 (WKB point geometry, WGS84). One row per station: source, station_id, site_id, name, latitude, longitude, variables, period_start, period_end, url (deep link to the agency page), river, country… See the full description on the dataset page: https://huggingface.co/datasets/Rekin226/aquascope-gauges.

sourceHugging Faceotherupdated 1d agoView on Hugging Face

0likes2.6kdownloads

Dataset Card

AquaScope gauges

The station catalog of every water-observation source AquaScope can reach, harvested on a schedule and published as GeoParquet. Last run 2026-09-23T23:27:06+00:00 with aquascope 0.18.0: 75,721 stations from 11 of 11 sources.

Files:

—stations.parquet: GeoParquet 1.0 (WKB point geometry, WGS84). One row per station: source, station_id, site_id, name, latitude, longitude, variables, period_start, period_end, url (deep link to the agency page), river, country, agency, license, redistributable, extra.
—stations.geojson: the same rows as GeoJSON for tools that don't read parquet.
—health.json: per-source status of the last run (station count, seconds, error if any).
—signatures.parquet: flow signatures of every mirrored discharge station (source, station_id, record span and completeness, mean flow, Q5/Q50/Q95, baseflow index, Mann-Kendall p and Sen slope of the annual maxima with amax_trend rising/falling/none, GEV L-moments Q100, mean day of the annual maximum, and notes saying why a value is empty). Join on (source, station_id).

site_id identifies a physical site within a source and defaults to station_id. Group by (source, site_id) to associate sub-stations while preserving every station record.

Query it in place

DuckDB reads the parquet over HTTPS without downloading it:

sql

INSTALL httpfs; LOAD httpfs;
SELECT source, count(*) FROM
  'https://huggingface.co/datasets/Rekin226/aquascope-gauges/resolve/main/stations.parquet'
GROUP BY source ORDER BY 2 DESC;

Or from Python: pandas.read_parquet("hf://datasets/Rekin226/aquascope-gauges/stations.parquet").

Sources and terms

Only sources whose terms allow redistribution will have their observations mirrored (Phase 1 of #188); every source appears in the catalog with a link back to the agency. Attribution for each source is in the AquaScope registry (aquascope.registry.SOURCES[key].attribution).

key	source	country	variables	licence	mirror observations	last run
`bom`	BOM Water Data Online	AUS	discharge, waterlevel, reservoirstorage, groundwaterlevel, precipitation, waterquality	CC-BY-3.0-AU	yes	16,673 stations
`brazil_ana`	ANA Hidroweb	BRA	discharge, water_level, precipitation	unknown	catalog only	11,406 stations
`greece_hydroscope`	Greece Hydroscope	GRC	water_level, discharge, precipitation	unknown	catalog only	704 stations
`greece_openhi`	Greece OpenHi.net	GRC	discharge, waterlevel, precipitation, climate, waterquality, evapotranspiration	CC-BY-SA-4.0	yes	65 stations
`hubeau_hydrometrie`	Hub'Eau hydrométrie	FRA	discharge, water_level	etalab-2.0	yes	4,158 stations
`ireland_opw`	Ireland OPW	IRL	water_level	CC-BY-4.0	yes	465 stations
`pegelonline`	PEGELONLINE	DEU	water_level, discharge	DL-DE-BY-2.0	yes	729 stations
`poland_imgw`	Poland IMGW-PIB	POL	discharge, waterlevel, waterquality	IMGW-PIB-open-data	yes	911 stations
`taiwan_cwa`	Taiwan CWA climate stations	TWN	climate, precipitation	OGDL-Taiwan-1.0	yes	1,265 stations
`uk_ea`	Environment Agency (England)	GBR	discharge, waterlevel, precipitation, groundwaterlevel	OGL-UK-3.0	yes	9,536 stations
`usgs`	USGS Water Services	USA	discharge, waterlevel, waterquality	US-PD	yes	29,809 stations

Observations (filling up week by week)

Daily values per station as obs/<variable>/<source>/<station_id>.csv.gz (date,value; discharge in m3/s, water and groundwater level in m, precipitation in mm/day), only for sources whose terms allow mirroring, plus one Parquet bundle per variable and source, obs/<variable>/<source>.parquet (station_id, date, value; join station_id to stations.parquet). obs/manifest.json lists every harvested station with its period, count and unit, and every bundle. Read one station:

python

pd.read_csv("hf://datasets/Rekin226/aquascope-gauges/obs/discharge/usgs/USGS-01646500.csv.gz")

a whole source in one go:

python

pd.read_parquet("hf://datasets/Rekin226/aquascope-gauges/obs/discharge/hubeau_hydrometrie.parquet")

or with DuckDB, joined to the catalog:

sql

SELECT s.name, o.date, o.value
FROM 'hf://datasets/Rekin226/aquascope-gauges/obs/discharge/uk_ea.parquet' o
JOIN 'hf://datasets/Rekin226/aquascope-gauges/stations.parquet' s USING (station_id)
WHERE s.name ILIKE '%thames%'

source	variable	stations harvested	bundle	licence
`greece_openhi`	discharge	6	`obs/discharge/greece_openhi.parquet` (11,570 rows)	CC-BY-SA-4.0
`greece_openhi`	precipitation	30	`obs/precipitation/greece_openhi.parquet` (73,039 rows)	CC-BY-SA-4.0
`greece_openhi`	water_level	40	`obs/water_level/greece_openhi.parquet` (66,140 rows)	CC-BY-SA-4.0
`hubeau_hydrometrie`	discharge	510	`obs/discharge/hubeau_hydrometrie.parquet` (4,412,883 rows)	etalab-2.0
`poland_imgw`	discharge	4	`obs/discharge/poland_imgw.parquet` (80,636 rows)	IMGW-PIB-open-data
`poland_imgw`	water_level	5	`obs/water_level/poland_imgw.parquet` (108,395 rows)	IMGW-PIB-open-data
`taiwan_cwa`	precipitation	35	`obs/precipitation/taiwan_cwa.parquet` (99,860 rows)	OGDL-Taiwan-1.0
`uk_ea`	discharge	660	`obs/discharge/uk_ea.parquet` (9,749,604 rows)	OGL-UK-3.0
`uk_ea`	groundwater_level	213	`obs/groundwater_level/uk_ea.parquet` (210,129 rows)	OGL-UK-3.0
`uk_ea`	precipitation	201	`obs/precipitation/uk_ea.parquet` (1,975,618 rows)	OGL-UK-3.0
`uk_ea`	water_level	223	`obs/water_level/uk_ea.parquet` (1,702,513 rows)	OGL-UK-3.0
`usgs`	discharge	591	`obs/discharge/usgs.parquet` (5,463,958 rows)	US-PD
`usgs`	water_level	102	`obs/water_level/usgs.parquet` (574,852 rows)	US-PD

Catchments (BasinATLAS, `basins/`)

The level-12 sub-basins of HydroATLAS v1.0 / BasinATLAS (Linke et al. 2019, CC BY 4.0) with their routing and attributes, so any point on land can be placed in its catchment and described: basins/lev12.fgb (simplified polygons, spatially indexed; point lookups over HTTPS read a few kilobytes), basins/lev12_topology.parquet (hybas_id, next_down, main_bas, sub_area, up_area, ... plus a representative point), basins/lev12_attributes.parquet (every BasinATLAS attribute per sub-basin, incl. the upstream-aggregated *_u* fields), and basins/lev12.pmtiles / basins/lev06.pmtiles for maps. Built by basins.yml; aquascope basins at LAT LON and the MCP tool describe_catchment read them. Cite: Linke, S., Lehner, B., Ouellet Dallaire, C., et al. (2019). Global hydro-environmental sub-basin and river reach characteristics at high spatial resolution. Scientific Data 6: 283. https://doi.org/10.1038/s41597-019-0300-6

How it is built

aquascope harvest stations --out archive --publish Rekin226/aquascope-gauges runs weekly from GitHub Actions. Every collector answers stations() from its own agency API; nothing is hand-edited. Add a source to AquaScope and it appears here on the next run.

Citation

Cite the AquaScope software (Zenodo concept DOI in the repository README) and the agency of any source you use, as listed in the table above.