Do Not Use For Navigation
The most recent NOAA Nautical Charts should be used for navigation. Data were created from many hydrographic surveys using different sounding techniques and conducted over different time periods. Hydrographic surveys are conducted for varying purposes, so soundings are collected at different spatial intervals and depth resolutions. Therefore, these data sets (DEM's) may contain temporal and spatial discontinuities, and should not be used for navigational purposes.
Reliability of Contours < 2m Generated from NOS' Estuarine Bathymetry Data
Users who generate vector contours from the DEM data, note the following information.
While it is possible to generate 0 and 1 meter depth contours using only observed soundings within estuaries, most of the 0 and many 1 m contours will dangle, in some GIS programs, without interpolations to the high water shoreline. The 2 meter contour is usually consistently closed and is thus the shallowest depth contour which is reliable throughout most estuaries.
Data were generated using both NOS soundings and a shoreline which is assumed to have an elevation equal to mean high water using triangulated irregular network (TIN) procedures. A portion of this TIN is based on a data model which is known to be incorrect. The TIN model assumes that each triangle surface is represented by a plane and that each edge and interior point can be estimated by linear interpolation from the corner point elevations. The profile of a beach or shore transect in a sedimentary estuary is typically non-linear with steep inter-tidal beach landward of a tidal flat each of which can be approximated by different but constant slopes. It is rare that the break between these two features is sampled with real soundings for a variety of operational reasons. The net result is that depths interpolated from triangles which include a shore vertex have a relatively large uncertainty for the horizontal location of any interpolated contour.
In addition to these uncertainties, there are three other types of errors which affect interpolated contours and may be present in some estuary's bathymetric data set. These are 1) surveys which were adjusted to different datums, 2) real temporal changes related to erosion and deposition, and 3) the dearth of very shallow soundings related to operational constraints on the surveys.
Often, estuaries have points which come from multiple surveys collected over intervals of several decades. Thus the soundings were not taken at the same tidal epoch throughout, and there will be some unknown variability in the datum related to real changes in sea level caused by long term sea level changes or earthquakes. Typically these errors are less than 0.1 m and they will be more noticeable in shallow areas.
Second, there is also a good possibility of higher errors near shore, caused by erosional and depositional changes in the near shore area over the years between the surveys. These changes are relatively less important as one looks at the deeper areas of an estuary simply because the relative change is less. A compounding source of error is that the location of the shoreline which existed at the time of the survey does not necessarily coincide with the shoreline used as the boundary for the TIN. For our computations we used a shoreline best representing its present day location (circa 1980 for the 1:250,000 NOS digital shoreline). In the extreme, accreting shores require that some soundings be discarded because they are landward of the high water shoreline. Conversely, eroding shores are fringed with an artificially expanding tidal flat.
Third, soundings less than 0.5 meters depth are systematically not collected because of operational constraints (small boats used to collect soundings run aground). Prior to acoustic soundings when poles and lead lines were used to measure depths and when sail or oars were used for the boat propulsion, it was not uncommon for the inter-tidal region to be reasonably sampled. However since the propeller it is rare to find observations shallower that 0.3 to 0.6 meters below the datum. In regions with extensive tidal flats, this translates to a wide region where interpolation based on a marginal depth profile model must be used to compute depths.