Hydrology Metering

How full the nation's reservoirs actually are

Every reservoir on this page is measured by a federal agency — the US Geological Survey's gages, the Bureau of Reclamation's operations record, or the Army Corps of Engineers' project data. This desk reads those records directly and shows them dam by dam: the water line, the volume behind it, and how both compare to the reservoir's published capacity and its own history.

Delicate water zones

City desks over the reservoirs each metro actually drinks from — the leading percent is the LOWEST member, because that is the pool the city's drought rules key off.

River systems

Combined storage is a physical sum across operators. Each percent counts only the members that publish both storage and a capacity, on the basis each system's note names — hover a card for it.

Falling fastest over 30 days: Lake Oahe (-21,867 af/day), Lake Sakakawea (-15,867 af/day), Lake Oroville (-13,144 af/day). Rising: San Luis Reservoir (+6,735 af/day), Amistad Reservoir (+351 af/day), Elephant Butte Reservoir (+22.9 af/day). Measured storage change from each operator's own record — not a forecast.

Emptiest first

Ranked by storage as a share of the operator's capacity figure. The percent is storage divided by capacity, both in acre-feet — it is never inferred from the water line's elevation.

af = acre-feet; M af = million acre-feet. A date without a clock time is a daily value. Each percent divides the operator's storage by the operator's capacity on one stated basis — Corps and Texas lakes measure against their conservation pool (their pages name the basis), so a flood pool in use can read above 100%.

Measured by elevation only

These reservoirs report a water level, but no storage percent ships for them — either no public storage series exists, or the operator's capacity figure could not be verified on the same basis as its storage series. They are listed, not ranked: elevation alone does not say how full a reservoir is.

Sources

Live readings: USGS National Water Information System (15-minute gage data), Bureau of Reclamation RISE (daily operations record), the US Army Corps of Engineers CWMS data API (hourly project data, including each project's own pool-capacity levels), California DWR CDEC (daily state reservoir sensors), and TWDB Water Data for Texas (daily conditions for the Texas lakes). The map's context layers are committed at build time from public-domain sets — NASA/USGS elevation (AWS Terrain Tiles) hillshaded into the relief, Natural Earth river centerlines, the USGS Watershed Boundary Dataset's 18 regions — and the drought layer is the US Drought Monitor's current week (NDMC / USDA / NOAA) in its own colors, dated on the map. Each reading on this desk names its source and observation time in the reservoir's own time zone; when every live lane is dark, the desk serves its committed copy of the record with the date it was retrieved — it never fabricates freshness. Capacity figures are the operators' published numbers, each captioned with its source and vintage.