A physics engine, and a verification tool enforced on top of it. PhysWall inverts a closed, published, non-linear physical law at a single measurement point — and refuses when the inverse is not unique. Seven laws, one engine, and the same refusal in all of them.
A stage-discharge rating curve turns a water level into a flow. It is fitted from measurements, and the USGS is plain about the limit: each curve is "unique to that site and a particular period of time."
When the relationship moves, that is shifting control — "usually due to erosion or deposition of sediment at the stage measurement site." After a large enough flood, "the USGS may resurvey the channel cross section and produce a new rating curve."
This gauge went through Hurricane Harvey. The channel moved, and the curve moved with it.
peaks, before Harvey 8 annual peaks 24.663 ft
peaks, after Harvey 9 annual peaks 22.132 ft
field, before Harvey 11 measurements 22.771 ft
field, after Harvey 57 measurements 21.235 ft
at 10,000 cfs
Two eras and two methods. The spread at the flood mark is 3.43 ft, and it is 15 to 17 percent across the whole range.
⚠ This is the finding, not the failure. A tool that showed one curve would be hiding the thing worth knowing, and the exponent amplifies it further: a 5% stage reading is an 8.1% discharge, which is what an uncertainty budget is for.
A power law fitted to annual peaks says nothing about 1 cfs. Below half a foot of stage, and above sixty, the tool refuses rather than extrapolating — the record's highest is 49 ft, and a number past that would be arithmetic rather than hydrology.
Choosing between the four is a question about the river rather than about the arithmetic, and the tool will not make it for you.
USGS peak-flow and field-measurement records for one gauge, fitted as Q = C h^n. The coefficients are in the tool, and so are the sample sizes.
⚠ Where this stops: one gauge, one river, and two eras. It says nothing about a channel that has not been resurveyed, and nothing about a site with a different control.
It is not a flood forecast and not a design tool. It reads one gauge's own record back at it, and shows how much the answer depends on which record you use.
PhysWall was developed and architected by Gadi Zion.
Built on PhysWall — the same engine reads antenna bandwidth, conductor loss, bit erasure and heat limits. It answers what the measurement implies, and refuses when the measurement cannot say.⚠ Check this instead of believing it. Every number here reproduces from a source that is named, and the claims that turned out wrong are still printed next to what replaced them. The same engine runs all of these — it asks how much a measurement allows you to conclude, and refuses the same way in every field. The same engine runs all of these — it asks how much a measurement allows you to conclude, and refuses the same way in every field. How to check each one →