Field notes: a basin where three countries draw from one reservoir, one discloses
A trader I know keeps a spreadsheet tab labeled "guesses." It's for a basin shared by three countries, fed by a cascade of reservoirs, where exactly one of the upstream operators publishes a storage bulletin and the other two treat elevation data like a state secret. The downstream grid, which buys power generated off releases it can't see coming, has learned to read tea leaves instead.
This is close to the median case for shared basins, once you look past the handful that have treaty-mandated data-sharing annexes. Someone upstream controls the gates. Someone downstream needs to schedule thermal backup, price imports, or hedge a hydro-heavy position, and the only numbers available are whatever the upstream operator decides to release, on whatever schedule suits them.
The bulletin that does exist is useful, but it's also a lesson in what disclosure covers and what it quietly leaves out. It reports outflow at the dam face, updated most weeks, sometimes monthly when the rainy season gets busy. It does not report storage as a percentage of capacity. It does not report the other two reservoirs upstream of it in the cascade, which belong to a different ministry and a different politics. A downstream analyst trying to reconstruct basin-wide storage from that one bulletin is extrapolating from a third of the picture and hoping the ratio holds.
What makes this basin worth a field note, rather than just another entry in the guesses tab, is how the gap gets filled in practice. Nobody waits for disclosure reform. Rainfall proxies get pulled into a model and treated as a storage stand-in, which works reasonably well in a normal monsoon and falls apart the moment a reservoir operator draws down early for maintenance or holds back for a dry forecast nobody announced. Annual reports from development banks get mined for a single capacity figure, cited for years past its relevance, because nothing newer exists. Traders on the downstream side end up pricing basin-wide hydro availability off a number that describes conditions from three seasons ago.
What actually changes the picture is that the water itself doesn't care about jurisdiction. A reservoir's surface area shrinks and grows with storage regardless of who owns the gate or whether anyone's filing a report about it, and that shoreline shows up in a satellite pass the same way whether the operator wants it seen or not. Run the waterline through the reservoir's area-capacity curve and you get a volume estimate that doesn't depend on a ministry deciding to release a bulletin. It won't match the operator's own gauge reading to the decimeter, but it closes the gap between one published number for one dam in a three-dam cascade and a storage series for the whole system, refreshed on a schedule you control instead of theirs.
That's the shift worth noting for anyone pricing power off this kind of basin. The published bulletin stops being your only anchor and starts being one data point you can check against an independent series. When the upstream operator's outflow numbers and the satellite-derived storage trend start diverging, that divergence is information in itself, often earlier information than the next bulletin will give you.
Reservoir Monitoring was built around exactly this gap, converting each reservoir's water-surface polygon into a storage estimate across a watchlist, including the ones that never file a report at all. If your basin has the same lopsided disclosure problem, that's the entry point to see your own guesses tab replaced with something you don't have to wait on anyone else to publish.