Talking to local peers and co workers, there seems to be some disconnect between which geoids should be used when performing a FEMA flood certificate. The newer flood maps are referenced to NAVD88 datum, but what about the particular geoid? Each geoid can obviously obtain differing elevations on a single benchmark. Any insight as to which particular geoid used when obtaining elevations via GPS would be greatly appreciated!
The property in debate is within a flood map with an effective date of March, 2009.
Hold the bench marks referenced on the FEMA map. I hold the NAVD88 first order bench marks from the FEMA maps and apply Geoid 18. Gets me close to all of them.
We occupy the local HARN point (first order BM) fixing the 88 number, locate the Airport Bench Mark (first order BM about 4 miles away) using Geoid18 and there is .01'-.02' vertically each time. Using Cors/OPUS and Geoid 18 there is .06-.12' each time depending on the Cors points that are held.
Not bad, but not as good as you can do. Holding the bench marks is a CYA issue.
If you don't have bench marks or they are all gone I would suggest using the latest epoch say NAD83(2011) and Geoid 18. Geoid 18 is an improvement over earlier ones, but you need to match it to the correct epoch.
When working in the current realization of the NSRS, use the current geoid.
It's that simple. So if you are working in NAD83(2011), you should be using GEOID18.
I don't know of anyone who is consistently working in earlier realizations when doing flood cert work. That would be odd.
The only reason to NOT use the current geoid with the current NSRS is if there is a known and documented issue with it in your area. Those places are few and far between - it's far more likely for a benchmark to have been disturbed or subsided than it is for the geoid to be incorrect.
This article was written before GEOID18 was released - at the time the current model was 12A. Still an excellent resource:
The property in debate is within a flood map with an effective date of March, 2009.
If the datum is NAVD88, the dates don't matter.
The hybrid geoid models developed by the NGS are designed to approximate NAVD88 as much as possible. The various iterations over the years are not different vertical datums, but improvements to that approximation of NAVD88.
The GPS on Benchmarks campaign (lots and lots of new observations) and a lot better gravity modelling contributed significantly to the development of GEOID18.
The bottom line is that GEOID18 is the best available official geoid model today, and it is the correct model to use if working in the NSRS right now.
I side with Rover. I have done several of these and I simply state exactly what i held in the paperwork. Most of the time it’s simply nad83(2011) navd 88 via geoid 18. I held a couple that i knew BM were not disturbed and had been cked between by me. And new that the BM’s were good and had been held in the navd88 geoid 18. Gps on bench marks just supply the meta data. What and how. Seems that covers your rump pretty good.
You need to be sure whatever you're doing matches the Bench Marks noted as control on the maps. If 2011/2018 will accomplish that then great. Clearly it doesn't in my area. Also it's helpful to know how the original mapping was created. I was the person who did the control for our mapping so I'm very familiar how that was done, and it wasn't Geoid 18 since it happened in 2003. Bench marks (NAVD88) were held for all targets and the mapping used that elevation basis. I know other communities went through the same process and that is reflected on the 2010 mapping panels. You need to dance with who brung ya. That's the CYA way, otherwise it's EYA, not cover, expose.
Start by reaching out to the floodplain manager in your state, county or municipality. This could save you a great deal of time now and in the future. As MightyMoe suggested, matching the BMs that were referenced on the current iteration of that area's flood map is the procedure recommended by FEMA.
After taking the certified floodplain surveyor course, I can assure you that trying to intuit the meaning of the language used by FEMA is an exercise in futility. Don't guess, just start with a internet search for the floodplain manager at the state level, email them with a reference to your AOI and the flood panel or map ID. Even states with few water resources such as Kansas ( https://agriculture.ks.gov/divisions-programs/dwr/floodplain/contacts ) are readily accessible.
In my case, everyone I must work with who has been awarded the title, Flood Plain Manager, doesn't have a clue about the flood plain. I am dead serious about this. One county gave that title to the fellow who was in charge of general maintenance of the buildings, lawns, etc. Not once but on three separate occasions as the maintenance job had turnover. Another county added this to the duties of the County Attorney. Another placed it with one of the flunkies in the County Appraisers office who worked with GIS maps. Another put it on a Health Department flunkie.
Murphy has mentioned a great aid. The Division of Water Resources in Kansas is my go to source for what they call a BFA in the rural areas with no true flood maps. FEMA has never questioned a BFA elevation on any of my LOMA or EC forms. What it really amounts to is LIDAR taken during major floods in recent years. Our 2007 floods in southeast Kansas exceeded our official BFE's by as much as eight feet of elevation.
If we don't have a "real" flood zone on a FEMA map and only one of the Zone A guesses, we run cross-sections, submit the data to a local engineer, they create a report for the BFE and only that satisfies FEMA. Otherwise it's a guess, the Zone A maps sometimes are 60 feet in elevation from the actual 100 year flood location.
It's worth noting that the instructions for Item C2 on an Elevation Certificate indicate that an OPUS-derived elevation is acceptable. Note the CORS used and attach the OPUS report, and Bob's your uncle.
It must be noted that many FIRM maps lack bench mark control. In those cases it's not possible to use them for elevations. If there are no BFE plus no Bench Marks then CORS/Geoid Models are the only option. Of course if there are no BFE's your cross section survey is going to be the standard and all the metadata information gathered needs to be in the report submitted.
I know how the data was collected for the cross-sections that created the BFE's and it wasn't great. College students handed some GPS equipment and sent out to the field. That had to be circ. 2004-2006. However, with the topo data and the cross-sections on the local drainages the flood zones are matching the FIRM maps very well, in Zone AE.
Zone A is a disaster and should be removed from all GIS data bases.
Then of course there's the private GIS'ers that have caused havoc and should be sued.
Like any areas, it's important to understand the relationship between CORS/Geoid Models and NAVD88 bench marks. That should be SOP for any surveyor working with flood zones and local infrastructure.
As previously stated if possible always use the passive bench marks published by NGS to do your work. If not realistic and you use OPUS please make sure you pay attention to the positional accuracy that tool provides. If you're doing something else then do yourself a favor first, run the GEOID18 tool on the NGS site -- https://geodesy.noaa.gov/GEOID/GEOID18/ and check the accuracy that NGS defines for your area of interest. While the tool will spit out a geoid height to the nearest mm that is nowhere close to how accurate it really is. In many parts of the country the 95% confidence that NGS describes is often well in excess of .04m/.13 ft. You need to then account for the uncertainty in your observed ellipsoid height which if you're not following the NGS guidelines for GPS-derived heights can often be in excess of .03m/.10 ft.
"You need to then account for the uncertainty in your observed ellipsoid height which if you’re not following the NGS guidelines for GPS-derived heights can often be in excess of .03m/.10 ft."
I agree that one needs to pay attention to the OPUS accuracy, but the measurements and assumptions that go into floodplain mapping make 0.1 foot look pretty insignificant when it comes to predicting flood heights.
Finding that elusive court case where there is a dispute between the Geoid Model derived orthometric NAVD88, and the level run from Bench Mark NAVD88 elevation probably isn't going to happen. The local FIRM mapping isn't up to the challenge of finding the tenth of separation between the two results. I've been very impressed by the location of the 100 year flood line when a big event happens and where the mapping shows it. But a tenth is a bridge too far for the mapping.
Imagine the glossed over eyes of the judge and or jury in the court as two experts argue about an inch where a building was submerged by a couple of feet of water.
Using 2011/Geoid2018 should work to get close, but I'd still want to know how it relates to local Bench Marks. I do know in a number of communities and I figure most surveyors know if they've been working with it for a while.
Within my fair city, which is about 20 square miles in area, there are exactly 2 NGS benchmarks with levelled elevation. Both were set prior to 1960. I very much doubt that either has been reobserved by the NGS since. One is in a sidewalk near a building doorway (RD0310), the other in a culvert headwall (RD0313), so neither is impervious to settlement. Nevertheless, they look great.
The local county surveyor also has several benchmarks in the area, a couple of which are within a mile of one of the NGS marks. These were probably set in the 1970's or earlier. But nobody seems to know for sure. I've levelled between one of the the NGS marks and a few of the county marks and found good correllation - so far. So I'm confident that these marks are reasonably locally stable.
Meanwhile, elevations determined by long duration OPUS sessions consistently come out about 0.1' higher than those from the levelling relative to those passive marks. So I'm left wondering if the whole area hasn't subsided a bit in the last 60-90 years. This is the PNW, so we know that the whole thing is moving horizontally at a substantial rate. Why not some in the vertical as well? On a clear day I can see Mt. St. Helens from here, and Mt. Hood in the other direction. And what about ground water depletion?
I'm left wondering if these passive marks are really stable - compared to .... what? And if I wanted to really, finally determine the true elevation of my area how would I go about that? Is relying on a 90 year old passive mark -not in solid bedrock - really better than long duration GPS observations relative to CORS? What are the implications for mass of as-built documents of my city's utility infrastructure? Realizing that the 0.1' difference I've seen isn't all that significant for that last, but what if it was?
"What are the implications for mass of as-built documents of my city’s utility infrastructure?"
Those of us in subsiding areas have long since gotten used to the fact that legacy elevations aren't worth much. If the vertical movement is substantial and widespread, a program that monitors critical infrastructure is a best practice. But most discrete engineering projects can be managed successfully as long as the critical vertical elements are identified and measured prior to design.
Pick whatever vertical reference seems most appropriate, document it clearly, and move ahead. Personally, I prefer to reference NAVD88 rather than a 50-year-old NGVD29 mark, but the way you realize NAVD88 can vary as long as it's documented and reproduceable.
I often use a single RTN observation on small-scale standalone projects, and show a few control points on the site for those who follow later. For large-scale jobs I like to use OPUS-Projects for the RTK base stations, picking CORS that I know to be (relatively) stable. I call if NAVD88 in both cases, but I also describe how I got there.
Apparently there is an easier way. The Washington Post had a laudatory article on a young man in Miami who is raising awareness of climate change by posting signs in neighborhoods with the flood elevation on them. And how does he get this elevation? He "downloads" it. No survey crew needed. Ha, who knew?
"Personally, I prefer to reference NAVD88 rather than a 50-year-old NGVD29 mark..."
Fun fact - Early indications are that NATRF2022 elevations in my fair city are going to very nearly equal NGVD29. The difference looks to be well under 0.1 feet, probably close to ±0.01 feet.
Both city and county has continued to specify the use NGVD29 up to the present day because of the legacy as-built data. When we switch to the new datum it will be like '88 never existed. The only agencies anywhere in Oregon using '88 that I know of is the DOT and, I think, the Port of Portland. And FEMA, of course.
