GPSFileDepot.com
 

More Mexico displacement problems-- searching for well-defined benchmark points

Started by hwstock, June 11, 2013, 06:53:07 PM

Previous topic - Next topic

hwstock

I'm working on a Garmin topo map for the region between lat 32 and 33, and long -115 and -116. I've used the USGS shapefiles for the streams, and have tested both the 1 arc-sec and 1/3 arc-sec DEMs from the national map server.  I process the DEMs from gridfloat files to GeoTiff, then use dem2topo to make the mp files, then GPSMapEdit to put it all together. (For the 1/3 arcsecond data, I used gdal_translate to split the 1degreex1degree file into 16 blocks, which I eventually reassembled in GPSMapEdit.).

The region is about 1/4 above the US-Mexico border, 3/4 below. Above the border (in USA), the topocontours agree well with the coordinates of benchmarks (on mountains), with GoogleEarth imagery, and the streams line up with the valleys highlighted by the contours.  But for every place I can check in Mexico, the contours seem to be shifted about -60 meters in the x direction, and -60 meters in the y direction. In Mexico, the nhd stream polylines data line up with google earth stream imagery (within the limits of simplification),. However, the streams are not in the valleys defined by the contours, and the few confirmed waypoints I have on mountains don't line up with the mountaintops as defined by the contours.  The same holds if I use the 1 arc-sec DEM or the 1/3 arc-sec DEM.

Since I downloaded these data all in one chunk from the National map server, this discontinuity seems a bit strange.

Questions:
1) Do you have any idea of the cause?  I would blame the software, but the agreement in the USA argues against that.  Unfortunately there are no border area ridges (that I've yet found) with enough regularity to show the discontinuity.

2) Is there a list of Mexican benchmarks -- on well-defined points, such as summits -- that I might use to quantify the problem better?

I last made a map for a mountainous area in N Baja, and the summits were so sharply defined in that area, that the misregistration of the contours was obvious.  There it was also an error of about 60 meters, but almost entirely along the x direction.

Thanks. I hope you can follow this!

Boyd

Quote from: hwstock on June 11, 2013, 06:53:07 PMI would blame the software, but the agreement in the USA argues against that.

I can't really help with most of that, but here's an idea for checking whether software is the issue. Download the free trial of GlobalMapper, open the DEM and see if that shows the same displacement. http://www.globalmapperforum.com/download.html