I am going to be teaching a one day class next week on GNSS in a small country with very little survey infrastructure. They will soon be getting a network of CORS. I wonder what users in countries without a good geoid model do for transforming heights?
EGM08??ÿ
When I did work for a water system in Ecuador, it was in an area with no existing control at all, so I used EGM08 to get orthometric heights. It was a small area?ÿ so I am sure it will work fine. However, when dealing with a large area, I would imagine the uncertainties and unmodeled biases in a worldwide model like EGM08 would cause problems. However, since they have no leveling network at all except along the coast, it may not matter. They certainly do not have the resources to create a leveling network or to compute their own geoid. Plus, the interior of the country is largely uninhabited jungle with few roads.?ÿ?ÿ
The following week I will be attending the SIRGAS 2018 symposium in Mexico, I will have plenty of opportunity to discuss this then, but unfortunately that will be AFTER the class.?ÿ
?ÿ
FWIW, I am not a ??non-US surveyor? but will nonetheless chime in.
I am curious whether most private-sector surveyors in South or Central America engage in ??control surveying.? Are these activities the domain of the national entities (e.g. IGM or Universities)? Do private entities/surveyors mostly work with site-specific ??assumed? elevations? Are the national-entity vertical control work done as a network? Has the vertical work ever undergone systematic analysis and adjustment?
Given the paucity of vertical control and lack of easy access to data (my personal experience in S. America) the resort to ellipsoid height and a model of the ellipsoid-geoid separation seems the most efficient approach. EGM08 is certainly an improvement over previous world-wide models. Note in the reference below that the model??s authors refer specifically to very limited GPS on benchmark data in South America. The absence of GPS on BM data does preclude some model validation over these areas. It also makes the issue of creating a hybrid geoid model allowing work in a previous vertical datum moot.
As you note, SIRGAS is addressing this issue at their October 2018 workshop. I hope the Conference and Workshop are fruitful. I will be in France at that time.?ÿ
Good article on EGM08:
https://agupubs.onlinelibrary.wiley.com/doi/pdf/10.1029/2011JB008916
UNAVCO??s EGM08 tool:
SIRGAS Vertical issues meeting site:
http://geoweb2.inegi.org.mx/sirgas2018/pdf/ProgramaTallerSIRGASv2.pdf
?ÿ
I think the quality and quantity of benchmarks varies greatly across South and Central America. A lot of it was done by (or in collaboration with) the Inter American Geodetic Survey (IAGS, part of AMS/DMA/NGA etc) from after WWII until it was disbanded in the mid 1980's. In Peru I was only doing azimuths in hysro tunnels, so I did not look for any horizontal or vertical control. In El Salvador I was mostly concerned with horizontal (recovered triangulation stations) but I did recover some BM's there and used them in my GPS network. The current work is in Guyana, the UK Directorate of Overseas Surveys (DOS)) did work years ago when it was a British colony, but pretty much just along the coast (much of which is below sea level). Horizontal control was extended inland by Aerodist (31 stations) done in 1967 and 1968 and HIRAN (5 stations). The water authority did some work also (90 BM's along the coast). Interesting to note that the Georgetown?ÿ Vertical Datum is set to -56 feet MSL. NGS set a few stations at the main airport in 1996, and published them as NAD83 (1996). I believe they will use SIRGAS in the future, which is basically ITRF
If you look at the country in Google Earth you will see that away from the coast is almost entirely uninhabited jungle, a few small settlements (some indigenous) and mining projects.?ÿ
I made a map of EGM08 in the country (0.1 m contour interval):
?ÿ
A Geoid is the best fit in a least square sense of the earth's gravity field at mean sea level. GPS cannot measure gravity thus it will not return an accurate geoid. It will only rely on the model it has to establish its height.
You cannot see it or measure it directly. Differential leveling can be utilized to model an area, but further investigation must be done through the use of local gravimetric observations. Then, through the use of Disturbing Potential, Bruns Formula, Height Relationships, Gravity Anomalies and Stoke's Formula can you then begin to model your local area with an accurate representation of the local Geoid.
For information about the 1997 US NGS/FAA GPS surveys and derived products see:?ÿ https://www.ngs.noaa.gov/PUBS_LIB/CARIB97_PAPER/carib97_paper.html
This was not only a GPS ?ÿsurvey. See, for example:?ÿ