Pablo's Mission Planning Website

Home Links Tips Downloads MPSSF Tips

Greetings Mission Planners,

NGA has posted the raw DAFIF Next data on their DAFIF Webpage.  I expect  they'll post their single PFPS download zip shortly.  

In NGA News, a 7.3 Million Dollar contract was awarded to EarthData International LLC for imagery and data models describing the Earth's surface.  More details at http://www.washingtontechnology.com/news/1_1/contract-awards/23657-1.html. EarthData has particular expertise in obtaining high resolution terrain information (HRTI) using a fleet of aircraft, but the press release doesn't really explain what NGA expects them to do.  More on HRTI next week.

I've posted a new version of Excel2FV Mark II on the website.  I've added some error catching when you export a threat or local point file to an Excel spreadsheet.  Based on feedback it seems the export to Excel doesn't work with Office XP or 2003.  If you've had success with either version of Office please let me know.  I've also added the ability to update Local Point elevations from DTED.  When you add local points without available DTED you'll have records without elevations.  Now you can use Excel2FV's "Point Group Manager" tab to select point groups and update their elevations.  All points that don't have elevation values will be updated and you can choose to update elevations that were pulled from DTED or manual entry.  

Mission Planning Tip: DTED Part 4, SRTM Continued

Last time I discussed SRTM in general and the causes of data voids.  Today I'll explain why these data voids are so dangerous to anyone flying low level.  Simply put, using SRTM data without a clear understanding of its problems can lead to terrain impact.

Shiprock NM is the eroded core of an ancient volcano.  The base terrain elevation is approximately 5,500 ft and the top of Shiprock is at 7,178ft (as noted on the NGA ONC and USGS 1:24K charts).  The image below comes from http://www.geology.wisc.edu/~maher/air/air03.htm.  You can find many more beautiful pictures taken of Shiprock from the air there.

This is the way Shiprock looks in the SRTM DTED:

That nasty black area in the middle of Shiprock is a void.  This is the way the data looks in SkyView:

Standing closer to the edge in wireframe mode shows:

Without imagery overlapping the DTED posts you can see how the data void extends downward.  This is the way the contours look overlaid on the USGS 1 Meter DOQ imagery:

Instead of extending to 7,178ft, the peak elevation (that I could find) was 6,204ft - almost 1,000ft lower than the height of the actual terrain.

So what does this mean?  

  • Aircrew use terrain elevations to calculate numerous altitudes, including Minimum Safe Altitude (MSA), Emergency Safe Altitude (ESA), Recommended NVG Altitudes, and enroute Low Level Altitudes.  Any altitude calculation based on the SRTM DTED at Shiprock will be 1,000ft in error.  
  • Modern Terrain Following (TF) Systems use more than radar to direct the aircraft's terrain clearance altitude.  DTED elevations are used to plan for terrain ("hills don't move") with a bias for the current altitude as determined by the radar altimeter, and a short range sensor is used to detect unplotted obstructions like powerlines and towers.  A modern System like this can determine when to start a climb for upcoming turns and lets you fly TF without blasting more 'trons than a UHF television station.  Unfortunately many aircraft won't have the performance to out-climb a 1,000ft "tower" like Shiprock that's detected for the first time by a short range sensor because it fell in a SRTM void.
  • Many aircraft (military and civil) use a Terrain Avoidance and Warning System (TAWS) to alert them to terrain in their vicinity and to warn them of terrain that threatens their aircraft.  These systems depend on elevation databases loaded before flight and in this case won't be aware of a 1,500ft spike of terrain in the middle of a relatively flat valley.
  • Any terrain masking conducted in the vicinity of the void will be invalid. Without knowing all the terrain masking factors an aircraft or weapon might be forced to use a marginal ingress route because they were unaware of the terrain masking effects of the terrain in the SRTM void.

Amazingly enough, a US Government dataset exists that does include Shiprock.  The USGS National Elevation Database (NED) includes Shiprock with an elevation of 7.133ft - an error of less than 50ft from the elevation recorded on the NGA and USGS charts.  This is the way the NED elevation model plots in SkyView:

NGA has made no announced progress in filling in the SRTM voids.  Fortunately there are good elevation sources available for NGA's use in filling the voids.  The USGS NED high resolution dataset covers the United States and NGA has older DTED covering much of the earth.  SPOT Image's Spot 5 Satellite is busy gathering data covering much of the world to generate a Level 2 DEM (Digital Elevation Model).  So far they've collected almost 2,500 Million Square Miles of stereo imagery to use to automatically generate DEM's.

If you're interested in obtaining SRTM DTED on CD or DVD you can't get it from NGA as they haven't begun distribution, but NGA has provided the data to the USGS who is distributing the data.  You can find more details on the USGS website here.  The image below shows the Level 1 DTED coverage available from USGS and was updated on May 28th:

Each CD costs $45 and each DVD of Level 2 data costs $60.  I'll leave it at that.

A basic truism of flight instruction is it isn't the bad students that'll kill you, it's the good ones.  With a bad student you're always ghosting the stick and wondering what inventive way "Stan" will come up with to try and make you a mort.  With a good student you want to kick back, relax and enjoy the ride.  Even when the student makes a mistake you know they'll see it, diagnose it and correct it - that is up until the time they miss it, get confused and put you in a situation you're lucky to get out of alive. 

SRTM is a lot like that good student.  More than 99% of the elevations near Shiprock are accurate within 5 Meters, but the missing 1% are enough to kill you and everyone on your plane.  I have yet to find another example as egregious as Shiprock, but it seems unlikely there's only one such case in a worldwide dataset. I wouldn't use SRTM DTED unless I had no other choice.  Unfortunately for much of the earth you don't have any other choice.  SRTM is supposed to be our foundation for geospatial data, but it's a foundation riddled with holes.

Good Luck,

Paul