Based on your experiences, when post-processing your GPS results, what vertical datum do you use to get the closest ellipsoidal heights to MSL values?
My default setting is EGM96 but I think there is a newer model EGM2008. I switch between these 2 models but sometimes one or the other would give a closer value when I check with a MSL benchmark.
Any other model that would give consistent closer values to MSL?
What is Google Earth using for their vertical datum, it seems pretty close to MSL.
Ryan, when in doubt you may want to supercede your GNSS derived heights with a level and bench mark. Do you have passive marks available? Elevations can be critical, especially in the Netherlands.
Isn't Zeevolde a product of continuous water pumping? The town would be lost/under water without it.
First, it is important to make sure of what you are trying to accomplish. When you say MSL to you really intend it to be Mean Sea Level. MSL is a specifically defined tidal datum and exists only at or with a short distance of the shore line. Or do you mean to your or some other national geodetic vertical datum. My understanding is that in the Netherlands this would be either the Normaal Amsterdams Peil (NAP)/Amsterdam Ordnance Datum or possibly the European Vertical Reference System (EVRS). I would hope that the Netherlands national survey office would have a geoid model that would best fit these datums.
The most current and best global geoid model is EGM2008 but remember that it is trying to relate to the best global definition of Mean Sea Level, not local mean sea level and can easily be different by upwards of a meter. For example using EGM08 to compute the height at the tide gauge in Sandy Hook, NJ which is one of the oldest in the United States it misses the local mean sea value by about 40 cm.
For surveyors in the conterminous United States the National Geodetic Survey (NGS) provides the North American Vertical Datum of 1988 (NAVD 88) and accurate geoid models that are combined in a tool called VDatum that will allow uses to transform ellipsoid heights into either approximate NAVD 88 or local tidal datum values at the few cm level.
I usuallt have a cinnection done to a MSL Benchmark via differential leveling. However for points that are difficult to reach via differential leveling, I want to check on the carried elevation values using geoid models.
Actual MSL elevations are not critical in most of my surveys unlike maybe in the Netherlands or coastal areas.
Apply the geoid model to the elevation point, forget trying to apply it to the height, you will be pleasantly surprised how well your elevations will check...