Estimating reservoir storage when the area-capacity curve is stale
Most area-capacity curves on file for a given reservoir were built from a bathymetric survey done once, maybe twice, in the structure's lifetime. A dam commissioned in 1974 might be running on a curve surveyed in 1974, or updated once in the 1990s after an inspection flagged obvious shoaling near the intake. Everything since then is sediment nobody re-measured.
That matters more than people outside the basin realize. Sediment doesn't deposit evenly. It piles up near the inflow deltas and in the dead pool first, which is exactly where the curve's lower elevation bands live. So the error in a stale table isn't uniform across the reservoir's operating range. A curve that's fine at full pool can be badly wrong at 30% capacity, which is precisely the range where traders and grid operators most want to know what's left.
Why the curve degrades from the bottom up
Think about where inflow drops its load. A river entering a reservoir slows down, and the coarse sediment it's carrying settles out almost immediately, building a delta that grows toward the dam over decades. Fine sediment travels farther and settles in the deeper pool, including the storage band a utility counts on for late-season generation.
The result: the stage-storage relationship for low-to-mid pool elevations gets optimistic over time. The table says 2.1 million acre-feet at a given elevation; the actual basin holds less, because what used to be open water is now a few meters of accumulated silt. Nobody updates this annually because a full bathymetric resurvey is expensive and the dam owner has no regulatory trigger to do it more than once every decade or two, if that.
Working around a table you don't trust
If a new survey isn't coming, a few adjustments narrow the gap without pretending the curve is current:
Flag the age of the curve against known sediment yield. If regional data puts annual sediment accumulation at roughly 0.1-0.3% of original capacity (rates vary enormously by basin, so use what's documented for that specific watershed if it exists), multiply by years since the last survey for a rough capacity-loss correction. It's a blunt instrument, useful mainly to keep a 1998 table from being treated as gospel in 2026.
Weight recent elevation readings more heavily at low pool. If the curve is known to overstate storage below a certain elevation, cross-check any gauge reading in that range against operational history. If the utility drew the reservoir to the same stage last year and reported a different volume than the table predicts, that's a data point worth keeping.
Watch the surface area trend instead of trusting the volume number outright. This is the part that doesn't require anyone to commission a survey. A shrinking or growing surface polygon tells you drawdown or refill is happening even when the absolute volume conversion carries uncertainty. For a trading desk comparing week over week, the direction of change can matter more than the exact acre-foot figure, especially across a watchlist where getting a resurvey for every dam simply isn't going to happen. Reservoir Monitoring turns that weekly surface polygon into a volume estimate against whatever area-capacity curve is on file, so the storage series keeps moving even on reservoirs where nobody's done a sedimentation adjustment in twenty years.
When a resurvey is worth pushing for
Push for a full bathymetric resurvey when a 10-15% error in stored volume at this reservoir would change a real trading or dispatch call, the kind of error that shows up when a quarter of the operating range has silted in since the last survey. It's also worth pushing when the dam owner is already funding dredging or a capacity study for other reasons and a new survey can ride along on that budget. Outside those two situations, the cost rarely clears approval, and for most reservoirs in a basin-wide watchlist that bar isn't worth clearing, so the surface trend becomes the leading signal and the volume conversion stays directionally useful rather than precise to three decimal places.
None of this closes the gap between a stale table and a fresh one perfectly. It does mean a desk isn't flying blind on storage just because a dam owner hasn't funded a new survey since the last administration. If that's the situation across your watchlist, take a look at what a weekly surface-to-volume series looks like for your basin.