FAA Part 107 CertifiedDrone Photography · Video · Photogrammetry$2M Liability InsuredCalifornia · Metro Atlanta · South FloridaNDAA-Compliant LiDAR on Request for Federal WorkCheckpoint Report With Every LiDAR JobAI-Planned Flights · Pilot in Command
← Back to Blog

Outdated FEMA Flood Maps: What a Lidar DEM Can and Can't Fix

March 18, 2026

Illustrative image: Outdated FEMA Flood Maps: What a Lidar DEM Can and Can't Fix

Better terrain data is not the bottleneck in most of Southern California and Metro Atlanta. Public 3DEP lidar already covers nearly every county in both regions. The bottleneck is the engineering study behind the effective flood map, and how long ago it was done.

That changes what a city should buy. A new lidar DEM can tighten a floodplain boundary. It cannot, by itself, fix the hydrology, the channel geometry, or the structures that decide where the water goes. This post covers where terrain enters FEMA's process, what a DEM changes, and what public data says about your county.

How current are FEMA's maps?

FEMA now tracks currency by engineering, not by map date. Its metric is New, Validated, or Updated Engineering (NVUE). FEMA's September 2025 Notice to Congress covers the third quarter of fiscal 2025. It reported 40.9% of 1,286,462 mapped miles at NVUE "attained." That means preliminary or effective maps are issued on current engineering.

Another 564,495 miles were "initiated," meaning funded but not yet issued. Attained plus initiated reached 84.8%. The same notice says FEMA reviews each study every five years and drops it from NVUE if conditions changed significantly.

For history, a 2017 DHS Inspector General audit found 42% of flood map miles "updated and valid" as of December 2016. FEMA's own goal then was 64%. Those figures are nine years old. The June and September 2026 notices do not include an NVUE appendix, so September 2025 is the latest published national figure found.

A map can be current on paper and still miss losses. FloodSmart reports that 29% of NFIP claims from 2014 to 2024 came from outside high-risk flood areas. That is a mapping-limits number, not proof that terrain is wrong.

How FEMA uses terrain in a Risk MAP study

FEMA's Elevation Guidance (Guidance Document No. 47, November 2023) is the document that governs terrain data. Its revision table lists November 2022 as the prior version. The link is a state agency's copy, so confirm on fema.gov that this edition is current before citing it in a scope.

Four points matter for a city engineer.

New collection is QL2. The guidance says FEMA "has standardized on Quality Level 2" from the USGS Lidar Base Specification for new lidar acquisition. USGS defines QL2 as 10 cm RMSEz and at least 2 points per square meter.

Required accuracy depends on flood risk and slope. Table 1 of the guidance sets vertical accuracy by risk decile and terrain. The tightest tier is high risk on the flattest ground. It asks for 24.5 cm at 95% confidence in non-vegetated areas and 36.3 cm in vegetated areas. The loosest tier, low risk, allows 147 cm and 218 cm.

Vegetation is tested separately. When vegetated accuracy fails, the data producer outlines low-confidence areas. The guidance names dense evergreen forest as a case where lidar points may not reach the ground.

Photogrammetry is not excluded. The guidance says photogrammetric solutions "are not used as widely" now. For localized areas they "might still provide the best source of data." Lidar is the prevalent method.

None of this says FEMA will reject your site-scale DEM. It says that for a regulatory study, the dataset needs documented accuracy, independent checkpoints, and metadata.

How terrain enters a LOMR or CLOMR

A Letter of Map Revision (LOMR) changes the effective map to reflect existing conditions. A Conditional LOMR (CLOMR) is FEMA's comment on a proposed project. FEMA's MT-2 guidance (December 2020, Guidance Document 106) lays out the submittal. Check fema.gov for a newer edition before relying on it.

The terrain appears in three places.

  1. The model chain. A LOMR needs a duplicate effective model. It may also need a corrected effective model, which "incorporates more detailed topographic information." After that come a pre-project model and a revised or post-project model. Updated topography in the corrected effective model "should reflect the physical conditions of the area at the date of the FIRM."
  2. The topographic workmap. It must be sealed, signed and dated by a registered professional engineer. It shows contours with elevation labels, the effective and revised boundaries, cross section locations, and the vertical datum. That datum must match the hydraulic analysis.
  3. The hydrology backup. When the revision changes hydrology, include a DEM if available. Add a drainage area map with contours and sub-basins.

There is also a path built for better topography alone. FEMA will process a LOMR that redelineates the floodplain on better terrain only if two conditions hold. No manmade modification nearby may change base flood elevations. And the area must be small and site specific, or the reach must be "Valid" in FEMA's Coordinated Needs Management Strategy (CNMS).

That second condition is why the CNMS status of your reach matters before you commission a flight. Per the same guidance, requests that improve the effective information are fee exempt, unless they incorporate manmade modifications.

What a lidar DEM changes, and what it doesn't

A DEM changes the geometry the model runs on. It does not change the flow estimate, the roughness, or the structures.

A published test shows the size of the effect. In a 2008 Purdue thesis by Aaron Cook, the same HEC-RAS model ran on four terrain sources on Strouds Creek, North Carolina. Inundation area was 0.121 mi² on a lidar-plus-surveyed-channel DEM. It grew 20% on lidar alone, 47% on a 10 m national DEM, and 64% on a 30 m DEM.

Look at the lidar-only result. It still overstated the area by 20%, because lidar does not capture the channel below the water surface. The thesis notes that surface datasets "do not include main channel bathymetry." On the larger Brazos River reach, the gaps were 20% for lidar and 26% for the 30 m DEM.

This is a thesis on two reaches, not a general rule. It shows why a lidar DEM is necessary and not sufficient.

What a better DEM changes:

  • Floodplain boundaries where ground is flat and the boundary was drawn on coarse contours.
  • Cross sections and overbank storage.
  • Drainage divides and flow paths in the hydrology backup.

What it does not change:

  • Hydrology. MT-2 allows gage statistics, precipitation-runoff models, and regional regression for new hydrology. A DEM feeds basin delineation but not the rainfall or the record length.
  • Channel geometry below water. That needs survey or bathymetry.
  • Structures. Bridges, culverts and basins need as-built or surveyed geometry. MT-2 asks for certified as-builts.
  • Coastal and levee factors. These have their own analyses.

What 3DEP already covers in SoCal and Metro Atlanta

Before paying for terrain, check what exists. USGS publishes work-unit footprints from its Work Unit Extent Spatial Metadata (WESM). Figure 1 uses the footprints from the USGS 3DEP index service, which carries the WESM fields, intersected with Census county outlines.

Dot chart of USGS 3DEP lidar work units by county for Los Angeles, Orange, Riverside, San Bernardino and ten Metro Atlanta counties, plotted by collection year and quality level
Figure 1. Each dot is a 3DEP work unit covering at least 10% of the county; area scales with share covered. Blue is QL1 or QL2 that meets the spec; hollow grey is older or sub-QL2. Source: USGS 3DEP index (WESM fields), accessed 2026-10-17; county outlines from the U.S. Census Bureau (cb_2023). Metro Atlanta here means the ten Atlanta Regional Commission counties.

Three findings stand out.

  • Coverage is near complete. QL1 or QL2 work units that meet the spec cover 97.9% of Riverside County. They cover 99.6% or more of each of the other thirteen. These shares are computed here as the union of footprints.
  • Collection years are recent. Los Angeles has a 2016 QL2 set and a QL1 set whose collection ended in 2024. San Bernardino has four QL1 work units from 2024 named for FEMA Region 9. The name suggests a FEMA-funded collection, but that is unverified.
  • Old sets still overlap. A sub-QL2 Orange County set, with collection ending in 2012, covers 87% of the county. Fulton has a 2006 set and DeKalb has a 2010 set. Newer QL2 lidar from 2018 to 2020 overlays them.

So for these counties, the question is rarely whether lidar exists. It is whether your effective study used it. USGS reports the national baseline was 98.3% complete or in work as of FY2024, with higher quality levels and faster refresh next.

One location: an effective flood zone on 2019 terrain

The second figure puts a real FEMA flood zone on a real 1 m surface. It shows a creek corridor in north-central Atlanta, Fulton County.

Hillshade of a 1 m lidar DEM of a creek corridor in Atlanta with the FEMA Zone AE floodplain in translucent blue, the regulatory floodway hatched, and the 0.2 percent annual chance zone in orange
Figure 2. FEMA NFHL flood hazard zones (FIRM panels 13121C0233F, 0234F, 0241F, 0242F, effective September 18, 2013) over a hillshade computed from the USGS 3DEP 1 m DEM, project GA_Statewide_2018_B18_DRRA, collected January 28 to April 24, 2019. Sources: FEMA National Flood Hazard Layer and USGS TNM, accessed 2026-10-17.

The floodway follows the channel and the Zone AE spreads across the low ground. The hillshade also resolves roads, berms, a rail embankment and building pads at about 1 m. Those are features a 2006-era or 30 m DEM would blur.

Be careful about what this proves. It does not show the 2013 zone is wrong. The panels were effective about five and a half years before the lidar was flown. Which terrain the 2013 study used is not established here. The 2006 Fulton County set in WESM is a candidate, but that is an inference. The effective Flood Insurance Study cites its terrain source, and that is where to check.

That check is the useful exercise for a city. Pull the FIS, find the terrain source and date, then compare against the WESM list for your county.

The DJI and federal funding constraint

If federal money pays for the flight, the platform matters. The American Security Drone Act (Pub. L. 118-31, sections 1821 to 1833) bars federal funds for procuring or operating drones from covered foreign entities. FAR 40.202-3(c) applies the ban to federal funds from December 22, 2025. The statute sunsets December 22, 2028.

FEMA mitigation and mapping money is federal money. Check the award terms before choosing a vendor. Check the manufacturer's current covered-entity status on SAM.gov rather than relying on a blog post, including for DJI. Twinn Digital rents equipment per job and is equipment-agnostic. For federally funded work, NDAA-compliant sensors are available on request.

The ban concerns the platform used for a funded flight. It does not limit how you use a 3DEP dataset someone else collected.

What to ask for

For terrain that may support a LOMR, CLOMR or local drainage model, put these in the scope:

  • A classified point cloud (LAS or LAZ) with ground class 2, and a bare-earth DEM at stated resolution.
  • Vertical datum, geoid model, and units stated. The workmap datum must match the hydraulic model.
  • An accuracy report per ASPRS Positional Accuracy Standards, Edition 2. It needs at least 30 independent checkpoints and RMSEv. Vegetated accuracy is reported, not graded pass or fail.
  • Checkpoints measured by someone other than the processor, as the Elevation Guidance recommends.
  • Breaklines or channel survey where the channel matters.

Licensing is not optional. In California, topo, contours and DTMs for design or a permit need a licensee in responsible charge. That is a PLS, or a civil engineer legally authorized to practice land surveying (BPELSG Bulletin 50, B&P 8726). In Georgia, it is a PLS (Ga. Comp. R. & Regs. 180-7-.04).

FEMA names its own signers too. Under MT-2, new topographic data must be certified by a licensed surveyor or registered professional engineer. The LOMR workmap must be sealed by a registered professional engineer. An Elevation Certificate is certified in its Section D by the surveyor, engineer or architect the FEMA form calls for. Twinn captures and processes the data. A licensed partner stamps it.

A practical next step

Pick one flood-prone reach. Look up its effective study, its terrain source and date, and its CNMS status. Then check WESM for newer QL2 coverage over the same reach. If a newer dataset exists and the reach is "Valid," a terrain-only LOMR may be possible. If not, the gap is in the engineering, and a new DEM is only one input.

If you want help scoping the capture and accuracy report, see Twinn Digital's page for surveyors and engineers.

Ready to See Your Terrain?

Get a quote built around your project

Whether you manage farmland, watersheds, municipal infrastructure, or development sites — tell us about your site and we'll prepare a scope and estimate around the required outputs and schedule.

Get a quote

Prefer texting? Text (424) 577-5461

(424) 577-5461Get a quote