Hydrology¶

Overview¶

The POWER hydrology component complements the existing solar (Energy Fluxes) and atmospheric (Meteorological) products with surface and subsurface water-cycle information. Soil moisture, inland water-body height, and precipitation are essential terrestrial water-cycle variables for monitoring land– atmosphere exchange and hydrologic state. Where the meteorology data describes the atmospheric state above a grid cell, the hydrology data describes the land and water-body state at or below the surface, i.e. the height of inland water bodies, the depth of soil moisture, and (via partial integration) the high-resolution precipitation that forces the hydrologic system. Satellite altimetry provides height for inland water bodies while passive microwave radiometry constrains near-surface soil moisture, two independent observing strategies for the land water store.

Hydrologic data products differ from POWER's traditional gridded fluxes in several ways:

1. Feature-based, not gridded. NASA Global Water Measurements (GWM) products are tied to specific lakes, reservoirs, and rivers rather than a uniform global grid.
2. Mission-bounded. Most hydrology sources have hard temporal start dates tied to satellite launches. Soil Moisture Active Passive (SMAP) was launched on 31 January 2015 with science operations beginning later that year, while the Global Precipitation Measurement (GPM) Core Observatory launched on 27 February 2014 and anchors the IMERG record (Huffman et al., 2020). IMERG is reprocessed back to 2000 via the multi-satellite era with TRMM satellites.
3. Higher native resolution. Hydrology sources are typically delivered at spatial resolutions finer than POWER's ½° × ⅝° meteorological grid, which is why they are exposed via POWER's high-resolution data pathway. IMERG is provided at 0.1° and SMAP L4_SM at ~9 km, both finer than the MERRA-2-based POWER meteorological grid.

GWM¶

The NASA Global Water Measurements (GWM) portal at NASA Goddard Space Flight Center is an online source for satellite-derived water products covering inland water bodies and global mean sea level. The GWM portal distributes operational radar-altimetry time series for inland water bodies and is maintained at NASA GSFC. GWM products integrate observations from a heritage of radar and laser altimetry missions (TOPEX/Poseidon, Jason series, OSTM, Envisat, SARAL/AltiKa, Sentinel-3, Sentinel-6, ICESat-2, SWOT) into operational time series of water surface heights and related variables for individual lakes, reservoirs, river reaches, and wetlands. Multi-mission satellite altimetry has been demonstrated to retrieve lake and reservoir surface heights to decimeter- level accuracy over large water bodies, and SWOT extends these capabilities with wide-swath interferometry.

Source and Reference¶

TODO[SOURCING|High]: verify the AWS Open Data path s3://nasa-power/gwm/ against the canonical POWER/GWM distribution endpoint and cite the bucket landing page.

POWER does not ingest GWM data from the main GWM portal. The GWM team provides a curated package of multi-resolution radar altimetry products specifically for Hydroelectric Power (HEP) reservoirs, updated regularly at a dedicated distribution point. This curated data set is a subset of the broader GWM product suite, filtered to reservoirs with identified HEP dams.

POWER's HEP Reservoir Focus¶

POWER currently serves GWM data for approximately 340 reservoirs that have been identified as having Hydroelectric Power (HEP) stations in their basins. GWM publishes per-reservoir altimetry time series for hydroelectric and other major reservoirs through its lake/reservoir index. TODO[DATA|High]: confirm the current count of HEP reservoirs served by POWER (≈340) against the GWM/POWER catalog and cite the source; the number is unverified here. The identification process involves two steps:

1. Ground-track collocation — The GWM team examines satellite altimeter ground tracks in tools such as Google Earth and identifies which tracks cross an inland water body.
2. Hydroelectric classification — Reservoir catalogs and databases are used to determine whether the water body hosts an HEP station. Only those reservoirs with confirmed HEP infrastructure are included in the curated POWER subset.

Water-Body Size Limitations¶

Due to the measurement geometry of nadir-pointing radar altimeters, only reservoirs with a surface area of approximately ≥10 km² are reliably acquired (Birkett, et al., 2022). The nadir radar altimeters sample along narrow ground tracks, so reliable retrievals are generally limited to larger open-water targets while wide-swath SWOT interferometry is designed to extend coverage to smaller water bodies. Ongoing instrument refinement and data exploration techniques continue to push through this size limit, but smaller water bodies may not appear in the catalog.

GWM Data Structure¶

Unlike POWER's primary energy-flux and meteorological products, GWM is not a uniform global grid. The native data are organized as time series per water-body feature (per reservoir) because:

1. Satellite altimeters do not produce a continuous gridded field over land — they produce point measurements along their orbital ground tracks.
2. Each water body is sampled at its own cadence (driven by repeat-orbit geometry and ground-track location).
3. Heights are most useful as feature-relative anomalies rather than absolute grid values.

This means GWM data are available by feature identifier rather than by latitude/longitude grid lookup, and the temporal axis is irregular (per visit) rather than fixed (hourly/daily). The GWM system maintains a strict one time series per one ground track approach — it does not merge data sets from satellites of varying temporal resolution or combine multiple ground tracks of the same resolution over a single water body.

Satellite Missions and Repeat Frequencies¶

Each radar altimetry mission is placed in a short-repeat orbit with a fixed exact repeat period. A reservoir is observable by a given mission only if the mission's ground track passes directly over it — if the track does not cross the water body, that reservoir is unobservable by that mission. A summary table of the missions is available at the GWM web site.

Source Prioritization¶

The 10-day resolution product files may contain two elevation values on the same day (spaced approximately one minute apart) when continuity missions overlap. The GWM team does not automatically select one value over another; POWER's data pipeline applies the following mission-priority rules:

These rules are applied during POWER's ingest pipeline so that users receive a single, prioritized water-level value per reservoir per observation date.

Parameters¶

GWM provides (per feature) parameters such as:

• Water surface height anomaly (relative to a feature-specific reference).
• Absolute water surface elevation (where reference geoid is established).
• Time stamp of each observation (irregular per orbital revisit).
• Quality / validity flag.

Derived products may include lake-area, reservoir-storage, and river-discharge proxies; those are mission- and feature-dependent and documented per product on the GWM portal.

Possible Use Cases¶

• Reservoir-storage monitoring for hydroelectric power planning.
• Lake-level anomaly tracking for drought / flood early warning.
• River-stage time series for transboundary basins lacking in-situ gauges.
• Calibration / validation of land-surface and hydrologic models.

Validation¶

POWER has not independently validated GWM parameters. Please see FAQs, documents, and references at GWM for more information on accuracy and validation.

SMAP¶

For a higher resolution soil moisture product, POWER provides SMAP Level-4 (L4_SM) data. The L4 product assimilates SMAP L-band brightness temperatures into the same NASA Catchment land-surface model used by MERRA-2, but running on a much finer grid (~9 km) and directly constrained by satellite observations. The L4_SM algorithm ingests SMAP L-band brightness temperatures into the Catchment model via an ensemble Kalman filter to produce ~9 km surface and root-zone soil moisture (Reichle et al., 2017).

Users can obtain SMAP data by enabling the high-resolution toggle in the DAV or request SMAP-aliased parameters through the API, POWER substitutes the SMAP Level-4 (L4_SM) product for MERRA-2 soil moisture.

SMAP Processing¶

1. Source product: NASA produces the L4_SM global daily analysis on a 9 km Equal-Area Scalable Earth (EASE) grid.
2. Regridding: The POWER pipeline reprojects L4_SM fields from the EASE grid onto POWER's standard latitude–longitude grid at approximately 0.09° spacing.
3. Temporal aggregation: POWER serves SMAP at the daily UTC temporal level. Monthly and annual aggregates are computed by averaging daily values.
4. Storage: Processed fields are stored as Zarr arrays in the POWER datastore on AWS (s3://nasa-power-smap/smap/).

TODO[SOURCING|High]: verify the AWS bucket/path s3://nasa-power-smap/smap/ and the 0.09° regridded spacing against the POWER datastore configuration; cite the canonical source. The SMAP L4_SM native grid is the 9 km EASE-Grid 2.0 [Reichle et al., 2017] (https://doi.org/10.1175/JHM-D-16-0204.1; accessed 2026-06-22); [SMAP L4_SM, 2026] (https://nsidc.org/data/spl4smgp; accessed 2026-06-22).

Parameters¶

Beyond the three wetness-fraction parameters shared with MERRA-2, the SMAP pathway adds volumetric moisture and climatological percentiles:

Additionally, SMAP provides six soil-temperature layers, evaporation, and land-surface radiation fluxes.

Comparison to MERRA-2¶

Both products use the NASA Catchment land-surface model at their core. This table highlights the difference in the two products:

SMAP set a target unbiased RMSE of 0.04 m³/m³ for surface soil moisture, and L4_SM validation has reported skill near this requirement against in-situ core validation sites (Entekhabi et al., 2010).

• Strengths:
  • ~6× finer spatial resolution than MERRA-2
  • Extensive product assessments completed by SMAP team
  • Percentile products can be used for drought classification.
• Limitations:
  • Record begins in 2015 (MERRA-2 present for full POWER availability)
  • Not quite global coverage: ~85°N/S latitude coverage
  • Active radar lost July 2015

Because both products share the same underlying model physics, transitions between MERRA-2 and SMAP in a time series are physically consistent in character, though absolute magnitudes may differ due to the satellite constraint and grid resolution.

Caveats¶

• No cross-calibration — POWER does not adjust MERRA-2 or SMAP values to force agreement across the 2015 boundary. Users building long time series that span both sources should apply their own bias correction.
• Resolution mixing — requesting soil moisture alongside ½° meteorological fields in a single API call will return data at different effective resolutions depending on the parameter. The grid coordinates are interpolated, but the information content differs.
• Frozen ground — both sources report reduced skill over frozen or snow-covered soils. SMAP brightness temperatures over frozen ground are excluded from assimilation.

SMAP Selection Logic¶

POWER applies a straightforward priority when resolving a soil-moisture request:

1. If the user has enabled high-resolution mode and the request falls within the SMAP spatial and temporal domain (2015–present, land, ±84.66°), return SMAP.
2. Otherwise, return MERRA-2.

Users requesting long climate records (pre-2015) will always receive MERRA-2 values regardless of resolution toggle state.

Validation¶

POWER has not independently validated SMAP parameters. Please see documentation at SMAP for quality assessments.

IMERG¶

IMERG merges precipitation estimates from the GPM constellation of passive- microwave and infrared sensors into a 0.1° × 0.1° (~10 km) half-hourly gridded product, providing the high-resolution precipitation that forces hydrologic models. More information on IMERG is available on the Meteorological page.