Cliff edge walk - very wrong profile

philpugh

I was walking a small part of the Cleveland Way from Robin Hoods Bay to Whitby over Easter (as a light break from the NYM). I planned the walk on SatMaps Routeplanner and the profile looked a bit dodgy - but I wasn't overly bothered. Its a 7.4 mile walk (actual logged distance - the planner has it as 1/2 mile less) but the Active10 reckoned it had 2500 ft of ascent! ( I would have thought no more than 1200-1500 ft) I guess the altitude based of the mapping is a bit suspect when on the edge of cliffs. The plotted tracklog was clearly inside the edge when plotted onto MemoryMap - so it s not down to positional errors. I guess the DEM they have has problems with sheer drops. Nice walk though - even though parts of the marked path are now in the North Sea and they are preparing even more diversions in expectation.
_________________
Phil

LostMike · Posted: Wed Apr 22, 2009 6:22 pm Post subject:

The A10 is renowned for getting total ascent far higher than most people would believe. It has been commented on here several times along with a hope that they may come up with a better algorithm.

So if you mean the maximum height you reached less the minimum height then I am surprised but if you mean the A10's estimate of the total height climbed on the walk then I would say it was about par for the course.
_________________
LostMike
Satmap A10. Platform 21
Software version 1.5.9193
Satsync 1.525

super_claret · Posted: Wed Apr 22, 2009 10:01 pm Post subject:

I am also having problems with the total ascent readings being ridiculously high.

I plotted a route on the Satmap which read total ascents 2100ft but when I actually followed the route the final readings were 4450ft!!!!

I'm sure it is a lot worse after installing the recent update. I personally think this is unacceptable, therefore I'm going to confront Satmap about this. I have one of my data boxes set to tell me the total ascent as I follow my route but I can't rely on this at all.

philpugh · Posted: Thu Apr 23, 2009 9:26 am Post subject:

I'm pretty sure the problem lies with the use of Digital Elevation Model that is supplied with the mapping. Other walks I have done with the Active 10 have the total ascent matching that provided by (for example) a GARMIN Oregon (to within 10ft in a 4300ft total ascent) - so I think it works most of the time. I suspect that the algorithm used can't handle sheer cliffs. The walk is along 7 miles of 150-200ft cliffs (within a few feet of the edge) and I suspect that it sometimes gives a 0ft (or very low) elevation when you get very close to the edge!

On the Satmap route planner the profile doesn't look correct either. I've published the route (Whitby to Robin Hoods Bay - look for those by philpugh) if you want to check it! Some of the 'incorrectness' is caused by the straight line route between the waypoints - but even so it just doesn't seem to match what the map is telling you.
_________________
Phil

Physicist · Posted: Sun Apr 26, 2009 8:00 pm Post subject:

I agree with Phil's cliff analysis (I have walked that Cleveland Way path with similar daft ascents). It isn't difficult to eliminate the problem in software; just discard any 'ascents' that are over say 5 m when the GPS updates itself every second. If your GPS height has changed by 100 ft in 1 second you haven't (unless you fell off the cliff!)

satmapian · Occasional Visitor Joined: Jan 14, 2009 Posts: 59

When planning a route on the Active 10 the profile uses only those heights that were plotted as a waypoint. Therefore to get a closer approximation to the actual total ascent to be covered you need to plot your route using waypoints at every dip and rise.

You can see this for yourself by plotting a route using two only waypoints placed either side of a mountain. You will see that the mountain does not appear on the profile. (I'm used to Anquet software, on which the mountain does appear, therefore waypoints are only needed at changes of direction).

philpugh · Posted: Tue May 05, 2009 1:43 pm Post subject:

Yes but....

This was the profile of the actual walked route - which when plotted onto a MemoryMap map was completely contained on the land side of the cliff and showed a better height profile.

I don't pay much attention to planned walk ascents as I don't litter a walk with waypoints - it gets too cluttered if you print off a map. You can judge the ascent of a route from the contours. I tend to use MemoryMap for route planning but decided to try out the Satmap version as it is simpler to share routes that way.
_________________
Phil

satmapian · Occasional Visitor Joined: Jan 14, 2009 Posts: 59

Sorry Phil I was really aiming my observation at Super_Claret's comments, and I should have been clearer.

I accept what you say. I also take height data with a pinch of salt when near cliffs and crags while out on the hill.

philpugh · Posted: Wed May 06, 2009 4:51 pm Post subject:

4ndynorfolk · Posted: Thu Jun 18, 2009 9:22 pm Post subject:

I climbed Helvelynn from Thirlsmere Swirl on Tuesday (fantastic visibility - could see Blackpool).

My Active 10 was set as usual to read elevation from maps. As I climbed the elevation would not change for ages, and would then jump 80 or 90 metres. Looking at the profile as I went up, the first 600m had just 5 or 6 steps. I changed the setting to take the gps reading and continued.

When I spoke to Satmap yesterday, I discovered that the height fom maps is not an interpolation of the contour lines as I had assumed. Instead, it calculates a weighted average height from the nearest 4 points on an OS supllied grid with 200m point spacing. This was what was causing my strange readings. I was also surprised to find that this data in part of thr firmware of the Active 10, rather than supplied on the relevant map card.

I was told bt Satmap that they were switching to NASA supplied data using a 90m grid. This will 'probably' be in the next update.

I know this does not explain the iffy ascent calculations, but it does explain phipugh's cliff edge issue.

Tony_P1 · Occasional Visitor Joined: Dec 30, 2008 Posts: 34

I had heard that the height grid was very wide, which explains why I can stand next to a trig point and be 10s of metres out when using altitude from maps. Consequently I always use altitude from GPS. Contrary to what the SatMap manual says, I find this is very accurate in most circumstances giving heights very close to the trig point/contour values. The base of cliffs and the bottom of ravines are not so good, but the map altitudes would be lousy there as well. Cliff tops and hills are spot on.
Tony

FrequentFlyer · Posted: Fri Jun 19, 2009 4:51 pm Post subject:

Tony_P1 · Occasional Visitor Joined: Dec 30, 2008 Posts: 34

	PocketGPS
	Web