NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
I had been using black box code to compute the geoid offsets in creating netcdf files and reading my netcdf files to get the MSL heights shown in http://www.suominet.ucar.edu/data/PWV_LatLon.table and http://www.suominet.ucar.edu/data/Global_LatLon.table and thinking that while the results did not match what I would get with manual run with typing in latitude and longitude of some EGM96 spherical harmonic Geoid model software I have that the differences were under a meter. That was only true in North America interior areas.
I tracked down the inputs to the black box code and found it was using a binary gridded table of geoid offsets that had been computed with an antiquated jgm2 model with only 70 coefficients. Wednesday I replaced that gridded table with one I created using EMG96 with 360 coefficients. Some of the Caribbean Island sites also had up to 8 meters of height change.
So if you use the netcdf files you may notice this offset. If you use the ascii PWV files you need to get the upgraded LatLonHeight files if 5 to 8 meters change on the station height will impact your results.
I'm very sorry for the inconvenience of this. Regards, Teresa and the Suominet team.
suominet archives: