Elevation Certificate: Conflicting Heights, a Benchmark Note, and an Old FIRM Datum
You are preparing a FEMA Elevation Certificate for a residence near the Boise River. The effective FIRM panel, dated 1996, shows a Base Flood Elevation of 2,712.0 ft. Your differential level loop, run from an NGS benchmark on NAVD 88 and closing 0.03 ft over 1.6 miles, puts the first floor at 2,713.0 ft. An RTK-GNSS observation at the same building corner, using the current geoid model, gives 2,713.42 ft. The contractor's hand level reads 2,712.6 ft. The lender's underwriter calls: at 1.0 ft above the BFE the premium drops sharply, and at 0.6 ft it does not. You also notice that the FIRM's elevation reference marks were established in NGVD 29, that the nearest NGS benchmark datasheet carries a note about possible subsidence from groundwater pumping, and that the floodplain administrator wants the certificate Monday.
1. Which height do you use for the certificate, and how do you justify rejecting the others?
Read the worked answer
Use the differential level loop: 2,713.0 ft, subject to the datum check below.
Differential leveling measures the height difference directly and is the standard of practice for an Elevation Certificate. A 0.03-ft loop closure over 1.6 miles is comfortably inside third-order tolerances (roughly 0.05√M ft, about 0.06 ft for M = 1.6) and indicates the field work is sound.
Reject the RTK height. Geoid model error in this region is on the order of a few centimeters, and a 0.42-ft discrepancy is far outside that. When RTK orthometric height disagrees with leveling by that much, the cause is almost always an RTK blunder - a bad fix, wrong base position, antenna height entered incorrectly, or a datum/epoch mix-up - not the geoid. Re-examine the RTK session for your own understanding, but do not average the two heights or split the difference; that produces a number supported by nothing.
Reject the hand level. A hand level is good to perhaps 0.1–0.2 ft in ideal conditions and is not acceptable for a certified elevation.
One caution on the loop itself: a good closure proves precision, not correctness. A tight loop can close perfectly from a benchmark that has moved, which is why the subsidence note must be addressed before the number is certified.
2. How does the NGVD 29 notation on the FIRM affect the certificate?
Read the worked answer
It is potentially decisive, and it is the step most often skipped under deadline pressure.
The Elevation Certificate must report elevations in the datum of the effective FIRM. Your leveling is on NAVD 88 benchmarks; the FIRM's reference marks are NGVD 29. In Idaho the offset between the two datums is on the order of 3 to 3.5 ft and it varies across the state - it is not a constant, and applying a remembered statewide number is a classic error. Convert at the site with the NGS VERTCON tool or another published, documented method, and cite the conversion in the file.
Consequences here: if the FIRM is on NGVD 29, the stated BFE of 2,712.0 is not directly comparable to a NAVD 88 height of 2,713.0. Converting the BFE to NAVD 88 produces a materially different freeboard than the naive subtraction, and the underwriter's 1.0-ft versus 0.6-ft question cannot be answered at all until the datum is resolved.
Verify the datum from the FIRM's legend and the flood profile - do not infer it from the panel date, since 1990s panels were converted on varying schedules. If the panel datum cannot be established with confidence, say so in writing to the administrator and the lender rather than certifying a number on an assumption.
3. What do you do about the subsidence note, and what do you report to the underwriter?
Read the worked answer
Subsidence. Check the NGS datasheet for the date and basis of the note and its stability classification. Then run the loop from a second, independent benchmark with a stable setting (deep-set monument, bedrock, or a benchmark outside the pumping area) and compare the two results. If the two benchmarks agree, the subsidence concern is likely historical or localized and the height stands. If they disagree by an amount that matters, the loop must be tied to the more stable mark and the discrepancy documented. Either way, record which benchmark controlled the certified elevation, its published elevation, and the closure - the certificate requires that information, and it is the first thing challenged if the elevation is disputed.
To the underwriter. Report the measured first-floor elevation in the FIRM datum, the BFE as shown on the effective FIRM, and the difference between them. Do not adjust the number to reach the favorable bracket, and do not characterize the structure as compliant or non-compliant beyond what the numbers and the community's rules support. State the method, the benchmark, the datum conversion, and the uncertainty of the measurement. If the result lands inside the measurement uncertainty, say so - an honest range is more useful and far safer than a falsely precise single value.
Put the same reasoning in writing to the floodplain administrator with the certificate. The deadline does not change the datum; a certificate filed on Monday with the wrong datum is a certificate that gets reopened, and it exposes the surveyor's seal to the consequence.