abberrations
Geo- for grown-ups
 
 FAQFAQ   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   RegisterRegister 
 ProfileProfile   Log in to check your private messagesLog in to check your private messages    Log inLog in 
  GamesGames   


Bookmark this site!
abberrations
Goto page 1, 2  Next
 
Post new topic   Reply to topic     Forum Index -> SMC contour maps
View previous topic :: View next topic  
Author Message
lucas
Member


Joined: 12 Nov 2005
Posts: 9

PostPosted: Sat Nov 12, 2005 6:43 pm    Post subject: abberrations Reply with quote

I have noticed a couple of strange things using the SMC contour maps.

Firstly when viewing certain contour maps with MapSource (map detail set higher or above and zoom level 500m or more) the 10m contours disappear. I have noticed it in two areas so far; around Beinn a'Bheithir near Ballachulish, Glen Coe (map N56W007, Garmin file 13207056.img) and The Black Mount, Rannoch Moor (N56W005, 13205056.img). It only seems to affect viewing in MapSource but not data that's uploaded to a GPS.

The other thing I've found is the appearance of "phantom" peaks and other features. The peaks are little diamond shapes of very concentrated contours. I presume that have been caused by the inclusion of spot heights in void areas when the original SRTM data was interpolated. I've spotted one on the Aonach Eagach Ridge (map N57W007, Garmin file 13207057) at N56 40' 44" W5 00' 54" and also several along the Cullin Ridge, Skye (N57W007, 13207057.img). I have also found a brand new corrie on Skye at N57 11' 07" W6 11' 01". This isn't a MapSource viewing issue - these features are uploaded to a GPS.

Has anyone come across similar things in other regions?

I have reinterpolated the SRTM data and compiled new Garmin images for the areas I have discovered. The new files cure the zooming problem in MapSource. The "phantoms" have also gone and the contours look closer to those on an OS map.

I have put the three new Garmin files (13205056.img, 13207056.img and 13207057.img) in a zip file which can be downloaded from: -

http://www.btinternet.com/~lm0128/GPS/contours.zip

Here are some before and after, reinterpolation screen shots.

Black Mount - loss of contours: -



Aonach Eagach - "phantom" peak: -



Skye - new corrie: -



Last edited by lucas on Sun Nov 13, 2005 12:08 pm; edited 3 times in total
Back to top
View user's profile Send private message
Wildlifewriter
Founder member


Joined: 04 Aug 2005
Posts: 948
Location: Norn Iron

PostPosted: Sun Nov 13, 2005 11:42 am    Post subject: Reply with quote

OK.

I've spent some time (most of the morning, in fact) looking at this.

The "disappearing contours" problem, as you probably realise, is to do with labels - or rather, the absence of them. It seems as if some of the contour line segments on the standard map, though they look contiguous, are actually segments without labels. Because of the way Mapsource works, this affects how they are displayed. My guess is that this would also be seen on the GPSr, but dependent on the detail settings (and display capabilities) of the GPSr.

The anomalous "features" are (of course) due to voids in the original data. I'm guessing that the original interpolation programme which was used to generate the contours made an attempt to handle the voided areas, producing mysterious corries and such as an artefact.

Clearly, the presence of these artefacts and voids is undesirable. more so since they tend to affect the very areas where people need the contour maps, for walking or climbing!

This then, is good work. Out of interest, can you give us a little more information on how you generated the corrections?

Subject to checking, it would be good if these corrections were applied to the current map files. Some additional work would be needed. For example: the areas where voids have been corrected would need to be edited out of the "voids" file AND the overview map.

I believe that you should take this matter up with Dave Storey (who produced the original sets) both as a courtesy and in case there are any technical issues with the changes.

I will send you a PM with his e-mail address.

-Wlw.
Back to top
View user's profile Send private message Send e-mail
lucas
Member


Joined: 12 Nov 2005
Posts: 9

PostPosted: Mon Nov 14, 2005 12:30 pm    Post subject: Reply with quote

Can I firstly just acknowledge the great work of Dave Storey. I have merely been copying his techniques.

To explain how I produced the "corrections" it is probably easiest to first explain what didn't work. I'll use Skye as an example.

Looking at the complex terrain of the Cullins and the sheer number of null points in the SMTR files it would be assumed that some additional data would be needed to fill these voids. I tried running an interpolation merging NIMA DTED0 with SRTM and also SRTM30 with SRTM. The results were quite disappointing. I did however amost reproduce the "Storey Corrie" Smile : -



I tried processing some preinterpolated DEMs. I used the "Hole-filled seamless SRTM data V1, 2004" available from the International Centre for Tropical Agriculture (CIAT): -

http://gisweb.ciat.cgiar.org/sig/90m_data_tropics.htm

The results were very good but the DEMs are 5 degree x 5degree tiles in GeoTiff format; a bit unwieldy to work with.

What actually seemed work most successfully was just running a straight interpolation of the orginal SRTM tiles with no additional data (I upped the iterations a little) Why this seems to give a good end result is probably down to luck! The void areas, while large in number, are relatively small in area.
Back to top
View user's profile Send private message
Wildlifewriter
Founder member


Joined: 04 Aug 2005
Posts: 948
Location: Norn Iron

PostPosted: Mon Nov 14, 2005 12:51 pm    Post subject: Reply with quote

lucas wrote:
What actually seemed work most successfully was just running a straight interpolation of the orginal SRTM tiles with no additional data (I upped the iterations a little) Why this seems to give a good end result is probably down to luck! The void areas, while large in number, are relatively small in area.

Which interpolation tool were you using?

I know that some of them (for example Blackart) have been improved since the DEM first became available.

-Wlw.
Back to top
View user's profile Send private message Send e-mail
lucas
Member


Joined: 12 Nov 2005
Posts: 9

PostPosted: Mon Nov 14, 2005 1:00 pm    Post subject: Reply with quote

Hi

BLACKART 4.02

I did wonder if it might be down to this. Reading through Chris Child's help file made me realise he's improved things a lot.
Back to top
View user's profile Send private message
GSV3MiaC
Member


Joined: 14 Nov 2005
Posts: 12

PostPosted: Mon Nov 14, 2005 11:51 pm    Post subject: Reply with quote

I did look at the newer version of Blackart, but I get basically the same
results as I did with 3.99 (which is what the SMC mapping was produced
using). I ought to rememebr the smoothing parameters, but I don't. Laughing

Here is the area of Skye around 57.189N, 6.192W, as shown by

GPSMAPEDIT (and these =should= be the same files which are loaded
on the SMC website).

I can't actually see 'my Corrie', but I can see that there is a break in the
ridge, (which may or may not exist on the ground) where there are quite a
few areas of missing height data. There are also a =heck= of a lot of
contours being generated per SRTM DEM data point (the smallest 'void'
rectangles are the size of each SRTM grid 'square' - i.e. about 90m N/S
by 60m E/W.



Last edited by GSV3MiaC on Tue Nov 15, 2005 9:43 pm; edited 1 time in total
Back to top
View user's profile Send private message
Wildlifewriter
Founder member


Joined: 04 Aug 2005
Posts: 948
Location: Norn Iron

PostPosted: Tue Nov 15, 2005 12:30 am    Post subject: Reply with quote

GSV3MiaC wrote:
Here is the area of Skye around 57.189N, 6.192W, as shown byGPSMAPEDIT (and these =should= be the same files which are loaded on the SMC website).

In that case, Dave, something bad has happened - because that ISN'T the mapping that's currently being hosted by SMC. (I just downloaded and extracted that Zip again, to make sure.)

This is the same region from the SMC files as displayed by GPSMapedit:



Click HERE for fullsize (200k) image. Clearly, this is not the same as your example.

-Paul
Back to top
View user's profile Send private message Send e-mail
GSV3MiaC
Member


Joined: 14 Nov 2005
Posts: 12

PostPosted: Tue Nov 15, 2005 11:38 am    Post subject: Reply with quote

Oh Rats, that version is 1.0 or 0.99 or something - quite possibly I zipped
the wrong file(s) when sending the CD to the webmaster (not having any
broadband here I'm doing everything at once removed).

I guess a 1.2 update version needs to go out - hopefully it will be uploaded
sometime in the next month or so. It's really only the heavy 'void' areas that
suffered from that probem (inadequate smoothing iterations in Blackart, iirc). The revised version if just as imaginary, but more geologically likely.
Wink
Back to top
View user's profile Send private message
Wildlifewriter
Founder member


Joined: 04 Aug 2005
Posts: 948
Location: Norn Iron

PostPosted: Tue Nov 15, 2005 12:09 pm    Post subject: Reply with quote

OK.

It could be that the website is wrong - perhaps an out-of-date link or something. Whatever, it would be good is this can be sorted out.

If there's anything we can do to help, please say so.

-Wlw.
Back to top
View user's profile Send private message Send e-mail
lucas
Member


Joined: 12 Nov 2005
Posts: 9

PostPosted: Tue Nov 15, 2005 1:41 pm    Post subject: Reply with quote

I second that. I'd be glad to help in anyway as it was me that caused the bother Embarassed
Back to top
View user's profile Send private message
GSV3MiaC
Member


Joined: 14 Nov 2005
Posts: 12

PostPosted: Tue Nov 15, 2005 4:33 pm    Post subject: Reply with quote

The website is wrong because there was ancient data on the CD that I sent
the webmaster .. I have perfectly working data (matched set, including the
right voids files, overviews, etc.) on my own PC, so I just need to send him a CD.

If someone else wants to host a copy, that'd be Ok with me, but we ought
check with the SMC webmaster (Ken) first.

The CD is burned - I just need to post it (and then Ken has to upload it). It is
several hundred MB in total, and has been generating a lot of download traffic (and will again, as soon as someone twigs there is an update) - just mention that because I'd not want someone to volunteer and then drown.

[p.s. why can't this wretched bulletin board wrap lines in the right place
(or an I doing something daft?).]
Back to top
View user's profile Send private message
Wildlifewriter
Founder member


Joined: 04 Aug 2005
Posts: 948
Location: Norn Iron

PostPosted: Tue Nov 15, 2005 6:17 pm    Post subject: Reply with quote

GSV3MiaC wrote:
[p.s. why can't this wretched bulletin board wrap lines in the right place
(or an I doing something daft?).]

The column width has been pushed out by the large (1165x741) graphic in an earlier post.

It happens all the time, in our digital photography section.
Shocked

-Wlw.
Back to top
View user's profile Send private message Send e-mail
lucas
Member


Joined: 12 Nov 2005
Posts: 9

PostPosted: Tue Nov 15, 2005 9:17 pm    Post subject: Reply with quote

Dave I've got a wee web space I don't use for anything so would be happy to host a copy of the files as soon as you get the nod from Ken.
Back to top
View user's profile Send private message
GSV3MiaC
Member


Joined: 14 Nov 2005
Posts: 12

PostPosted: Tue Nov 15, 2005 9:31 pm    Post subject: Reply with quote

Wildlifewriter wrote:
GSV3MiaC wrote:
[p.s. why can't this wretched bulletin board wrap lines in the right place
(or an I doing something daft?).]

The column width has been pushed out by the large (1165x741) graphic in an earlier post.

It happens all the time, in our digital photography section.
Shocked

-Wlw.


Ah, Ok, I'll fix the size of the graphic file then, since it no longer needs to be so big. 8>.
Back to top
View user's profile Send private message
GSV3MiaC
Member


Joined: 14 Nov 2005
Posts: 12

PostPosted: Tue Nov 15, 2005 9:57 pm    Post subject: Reply with quote

lucas wrote:
Dave I've got a wee web space I don't use for anything so would be happy to host a copy of the files as soon as you get the nod from Ken.


OK, I shall ask Ken if he has any concerns, and how much of the SMC webpage(s) you can/should mirror as well as just the data. Obviously I need the disclaimers, and preferably the thanks, attached to the data, in case someone falls down 'Storey's Corrie' .. OK, so it's not there in the 1.2 contour maps, but that doesn't mean it ain't there on the ground - a void is still 'here be dragons' in mapping terms. I guess we aren't going to get another SRTM mission to fill the holes in any time soon!!

I'll mail you the 1.2 CD so you can check that it's fixed, it works OK, and see how much web space it would consume. I can't guess how much bandwidth it'll chew up, that depends on number of downloads. ISTR it took the SMC download rate up by 30x to about 0.5 GBytes/day - but that was before most people had it, if they all want an update at once ...
Back to top
View user's profile Send private message

Display posts from previous:   
Post new topic   Reply to topic     Forum Index -> SMC contour maps All times are GMT
Goto page 1, 2  Next
Page 1 of 2

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


Create your own free forum now!
Terms of Service Purchase Ad Removal Forum Archive Report Abuse