If you have built enough HEC-RAS models, you’ve probably discovered that having data in different vertical datums is a bit of a pain. RAS lets you adjust the vertical datum of individual cross sections (under “Geometric Data”, “Cross Section”, “Options”, “Adjust Elevations”) or portions of the model (“Geometric Data”, “Tools”, “Datum Adjustment”). However, I have to confess this: datum adjustments are always a little bit puzzling to me. I need pen and paper and a little numbering conversion to make sure that I’m doing things right when moving from one vertical datum to the next. And I recently had a “eureka” moment when I finally realized that the sentence “NAVD88 is higher than NGVD29 (for most of the locations we work on)” is actually wrong. Let me step back a moment and I’ll explain why.

The two most common vertical datums used in the U.S. are the North American Vertical Datum of 1988 (NAVD88) and the National Geodetic Vertical Datum of 1929 (NGVD29). NOAA’s VERTCON program can be used to compute the conversion factors from one datum to the other.

Tidal datums are, in general, confusing. Vertical elevations can refer to NAVD88, the station datum, Mean High Water, Mean Low Water, Mean Range of Tide, and a variety of other datums. In recent years, NOAA has been updating its reference tables with excellent graphs showing how the different datums relate to each other (like the one below for Astoria, OR – http://tidesandcurrents.noaa.gov/datums.html?id=9439040 -, where all elevations presented are related to the station datum, which is the Columbia River Datum or CRD).

Let’s say that you download data in the Astoria station datum and you want to convert the data in NAVD88. Since the NAVD88 datum is 2.02 ft above the station datum, you need to subtract 2.02 ft from the station datum to obtain elevations in NAVD88. Or, to make it easier to remember, if the data you obtain is in datum X and you want to convert it to datum Y, you need to ADD the vertical distance between the two datums if X is ABOVE Y (or SUBTRACT if X is below Y). In our case, X is the Astoria station datum, Y is NAVD88, X is below Y, and you subtract 2.02 ft to the Astoria datum to obtain elevations in NAVD88 (the distance between the two). Makes sense, right? This also explains why for most locations on the West Coast USA, the elevation in NAVD88 of a certain location is always higher than the elevation in NGVD29 of the same location (that’s because the NGVD29 datum is actually above the NAVD88 datum). For example, in Astoria, you need to subtract 3.4 ft to elevations in NAVD88 to obtain elevations in NGVD29.

Now, to confuse things even more, gage locations along the Columbia River follow an intrinsic datum that actually changes as you move upstream from the river’s mouth. For example, at Astoria, the station datum is BELOW NGVD29, while at Vancouver, the station datum is ABOVE NGVD29 (see graph below, at http://www.thsoa.org/hy05/09_1.pdf).

The USGS sent us a nice graph (see below) that helped me clear things up when reading datum-related station description information. This one is for the Columbia River at Vancouver gage in WA. The short station description reads: “Datum of gage: 1.82 ft above NGVD29”. The long version reads: “Datum of gage is Columbia River Datum (add 1.82 ft to correct to NGVD29)”. Remember: X is the station datum, Y is NGVD 29, X is ABOVE Y, so you need to add the distance to go from X to Y. To add something more to this graph, I would say that in CRD, the 0 of the NGVD29 datum is 1.82 ft below the 0 of the CDR datum, which reflects the previous graph. Hope this helps!