Evaluating VDatum in coastal national parks: Fire Island National Seashore, Gateway National Recreation Area, and Assateague Island National Seashore

Ullman, David; Babson, Amanda; Bradley, Michael

The accuracy of tidal datum elevations provided by NOAA's VDatum software package (version 3.2) is investigated for 3 National Park Service (NPS) coastal parks, Fire Island National Seashore (FIIS), Gateway National Recreation Area (GATE), and Assateague Island National Seashore (ASIS). All available water level observations that could be referenced to the orthometric datum NAVD88, from NOAA as well as other sources in the vicinity of the three parks, were used in the comparison. For observational sites lacking published NOAA values, datum elevations were computed and corrected to the 1983-2001 National Tidal Datum Epoch (NTDE), the Epoch used for VDatum, using standard NOAA methodologies. The estimate of the uncertainties in computed datum elevations, arising from uncertainties in the observations, instrument elevations, and the computation of datums from short (less than 19 years) records are approximately 0.075 m. The corresponding uncertainty in datum elevations from VDatum, estimated by NOAA, is approximately 0.10 m. Comparisons between observed and VDatum elevations were performed for mean high and low water (MHW and MLW), mean higher high and lower low waters (MHHW and MLLW), as well as mean sea level (MSL). For 2 of the 3 regions, VDatum estimates for MSL were high in relation to observed MSL. Mean differences were 0.035 m at GATE and 0.039 m at ASIS and were significantly different from zero. Root mean square (rms) differences over all tidal datums and all sites were 0.032 m for FIIS, 0.064 m for GATE, and 0.046 m for ASIS. These rms differences are smaller than the estimated uncertainties for either VDatum or the observationally-derived datums. This suggests that the uncertainties themselves are overestimates and/or the VDatum and observational uncertainties are highly correlated. It also indicates that VDatum is more accurate than its uncertainty estimates suggest. For NPS applications in which datum elevations need only be known to an accuracy of 0.05-0.10 m, VDatum is sufficiently accurate. Although the estimated uncertainty of observationally-derived datums (using NOAA estimates for NAVD88 errors) is also in this range, the analysis in this report suggest that this could be an overestimate. Thus, it is likely that observationally-derived datums are more accurate and that targeted sea level observations can provide more accurate datum elevations than the existing VDatum. A critical issue, which will require consultation with experts in geodetics, is how accurately can instrument elevations be determined relative to NAVD88 as it is this uncertainty that dominates the uncertainty in an observationally-derived datum elevation. Maps of VDatum-derived tidal datum elevations in the regions with barrier islands separating the ocean from shallow embayments (FIIS and ASIS) show large differences between ocean side and bay side levels. For management applications, such as estimating sea level rise inundation risk for a land based location, users of VDatum elevations on barrier islands must decide whether datum elevations from the ocean side or the bay side are most relevant. Managers must also be cognizant of the fact that VDatum elevations and observational datum elevations computed and corrected using NOAA methodologies are relative to the 1983-2001 NTDE and thus may not be representative of actual water levels given recent sea level rise.

Evaluating VDatum in coastal national parks: Fire Island National Seashore, Gateway National Recreation Area, and Assateague Island National Seashore

Series