Pablo's Mission Planning Website
Greetings Mission Planners,
Hopefully this hasn't bummed Digital Globe out too much. On their plus side, DG was recognized by "Frost & Sullivan" for the "Aerospace and Defense Product of the Year". Details are available in this press release:
As for NGA they were "present at Geospatial World in a big way." Not sure what that means but details can be found here.
I've posted a new version of Excel2FV Mark II here. I've made some fixes based on reports from the PFPS 3.3.1 testing and added the ability to export a Threat File to an Excel Spreadsheet.
Mission Planning Tip: Digital Terrain Elevation Data Part 3, Shuttle Radar Topography Mission (STRM)
NGA and NASA teamed up to launch the space shuttle Endeavour on STS-99 carrying the Shuttle Radar Topography Mission "Instrument" (catchy name eh?). Through a combination of radar interferometry and BFM technology the shuttle gathered over 12TB of data from 56S to 60N - approximately 80% of the earth's landmass. A problem with a nitrogen thruster threatened to dramatically reduce the amount of data gathered, but the flight crew and big brains at Mission Control were able to regroup and gather almost all the planned data. At mission completion 99.968% of the targeted landmass had been covered at least once, 94% had been covered by two passes and 50% covered by three or more passes. The image below shows the sensor coverage and number of passes. A more detailed version can be viewed here.
Of course with 99.968% coverage there must be 0.032% that wasn't covered. The image below shows where those holes are. You can click here for a much bigger version.
The image below shows where the holes are on a US Map:
Despite the statement that all the holes are over the US, you can see that one hole extends into Mexico and a small hole (not depicted above) covers a portion of Columbia.
What does a coverage hole look like in FalconView?
As you can see, these holes are large and no attempt was made to fill them.
One characteristic of SRTM DTED is it measures the reflective surface on the day of the mission, not the elevation of the ground. Nothing shows this better than the image below of the Washington Monument with SRTM contour lines (10ft interval):
Of course since the height of the Washington Monument is 555ft you can also see that the data isn't perfect.
In addition to the large data voids, SRTM holes occur when the sensor was blocked on all passes. The image below of shows the rugged terrain of Canyon de Chelly National Monument:
Looking at a more micro example, here's a very small SRTM DTED hole in Colorado:
This hole covers 33 DTED posts. You'll remember that holes covering 15 posts or less are filled in by interpolation during processing. Here's what the DTED hole looks like on 1 Meter Imagery:
The void is in a low spot guarded on three sides by rising terrain. Without an lucky orbital pass it's difficult for the SRTM radar to see down the valley.
Another problem area for the SRTM is areas of low reflectivity, i.e. "stealth ground". If the radar waves are absorbed or diffused instead of being bounced back to the shuttle then an elevation can't be pulled from the data. This image of Edwards AFB shows large areas where the SRTM signal wasn't sufficient to determine an elevation. You'll also notice the area near the holes is mottled which shows a degradation of the radar signal.
Finally, it appears that the SRTM data or processing has trouble with high rise urban areas. This image shows the SRTM data over downtown Chicago:
Here's the area on a USGS 1:100K chart:
No similar data voids exist nearby, and since similar holes exist in Baltimore, Philadelphia, Atlanta, Dallas, Seattle and Los Angeles it seems clear that the multiple high rise structures are causing problems.
Next time I'll discuss what makes SRTM data so good, so bad and what US Government Agency can provide it to you today (hint: it's not who you think).