Lake area vs. lake volume: why extent isn't stage

A reservoir can lose a third of its storage and barely look different from above. That's the part nobody warns you about the first time you start pulling weekly extent layers into a basin report: area and volume are not the same number, and they don't move together.

Area tells you the footprint, not the bathtub

A lake's surface area is just the shape of the water where it meets dry ground at that moment. Volume is how much is actually in the bathtub. For a reservoir with steep sides, like most impoundments built in a canyon or gorge, a big drop in stage barely moves the shoreline. The surface shrinks a little, the storage number falls off a cliff, and if you're only watching extent you'd never know.

Flip it around for a shallow, saucer-shaped natural lake or a floodplain wetland, and the relationship goes the other way. A small rise in stage spreads the water across a lot more ground, so extent swings wildly while the volume change underneath is modest. Neither direction is "wrong." It's just the hypsometry of that particular basin, the area-elevation curve that's been there since the lakebed was carved, and it's different for every water body.

This is the area-extent vs storage problem in one sentence: without knowing the shape of the basin below the waterline, a bigger or smaller polygon on a weekly mask tells you almost nothing about how many acre-feet moved. You need the stage-area-volume curve, usually built from bathymetric survey or a digital elevation model of the lakebed, to convert one into the other. A satellite pass doesn't carry that curve with it. It just draws the edge.

The bathymetry gap in a satellite mask

Every surface water mask, ours included, is a wet/dry classification at the resolution of the pass. It answers "where is water this week" and, stacked against the archive, "where has water appeared, vanished, or moved since." What it can't answer is "how deep," because optical and radar instruments looking down from orbit see the surface, not the bottom. There's no bathymetry in the pixel.

That gap matters most for exactly the water bodies basin planners worry about: reservoirs operated against a storage target, and natural lakes where a few centimeters of stage change the usable yield. If your allocation decisions run off a storage number, extent alone won't get you there. You still need a gauge, a staff plate, or a stage-storage table specific to that basin to turn a shoreline into a volume.

What extent is good for is different, and it's not nothing. A weekly mask run against years of archive shows you which reaches are drying out that shouldn't be, which off-channel water bodies are filling after a flood pulse, and where a wetland boundary has crept or retreated over a decade, basin-wide, in places you don't have a gauge at all. That's a genuine gap in most monitoring networks. Gauges sit on the mainstem and a handful of named lakes. Everything between them, the oxbows, the irrigation-fed ponds, the side channels, goes unmeasured until someone drives out to look.

Using extent and stage together, not instead of each other

Extent and stage answer different questions, and a basin plan needs both. A gauge or reservoir operator's stage record gives you the number that goes into the water budget. A wide-swath weekly extent layer gives you the spatial picture of everywhere else in the basin, the reach-by-reach and lake-by-lake view of where water has shown up or dropped out between the stations you can measure.

Surface Water Monitoring was built around that second half: a weekly extent layer and a multi-decade time series for every lake and river reach in your basin, meant to sit alongside your gauge network rather than stand in for it. If you're trying to see the basin-wide picture between stations, that's the gap it's built to close.