Viewing 15 posts - 1 through 15 (of 24 total)
  • Author
    Posts
  • #25940

    geoffmozz
    Participant

    Is there any chance that the Ordnance Survey 10m interval contours (OpenData) could be used for the Great Britain OpenAndroMap?
    https://www.ordnancesurvey.co.uk/business-and-government/products/terrain-50.html

    It’s freely available and free to use. The Ordnance Survey say…
    “You can use OS OpenData map datasets in any way and for any purpose.”

    I think the only licensing requirements are that the OS are appropriately acknowledged.
    https://www.ordnancesurvey.co.uk/business-and-government/licensing/using-creating-data-with-os-products/os-opendata.html

    WalkLakes have created a map (WayMap) made with the OpenData Terrain 50 contours. Much more detail than SRTM/NASA contours.

    https://www.walklakes.co.uk/maps/

    Attachments:
    #25957

    Sonny
    Participant

    Thank you!

    I compared the OS Terrain 50 with SRTM 1″ data. In my diagnosis the details of OS are about those of SRTM. Maybe even marginally less. So I think no reason to use OS 50 instead of SRTM 1″.

    But since you wrote: “Much more detail than SRTM/NASA contours”. Can you provide a comparison of contours of both models in a certain area, which backs your statement?

    1 user thanked author for this post.
    #25959

    Sonny
    Participant

    I created 2 screenshots of maps of the hilly area around Pillar.

    contours based on OS50:

    contours based on SRTM1:

    1 user thanked author for this post.
    #25963

    geoffmozz
    Participant

    Sonny, I agree that the differences between the two screenshots you’ve shown are small.

    The WayMap that I pictured uses the OS50 contours. However the OpenAndroMap of Great Britain that I have uses a 20m contour interval and there’s obviously less detail in the contours than any of the examples that you and I have posted. eg Look at the contours around the crags SW of ‘Steeple’. The OS50 WayMap contours in my first post look quite good.

    Do you know why the OpenAndroMap contours don’t have the 10m interval and detail shown in your screenshot? The maps would be better if they did. Here’s a screenshot of my OpenAndroMap of Great Britain…

    • This reply was modified 2 weeks ago by  geoffmozz.
    • This reply was modified 2 weeks ago by  geoffmozz.
    • This reply was modified 2 weeks ago by  geoffmozz.
    #25970
    ChristianK
    ChristianK
    Keymaster

    Do you know why the OpenAndroMap contours don’t have the 10m interval

    The answer is simple: Too much data.
    The size of the map would be too big with 10m lines.

    With OS Maps this is no problem ’cause these maps are rastermaps – there is no differencce in size using 20/25 or even 10m Lines.
    The OAM of GB uses 3″ Elevation Data for the same reason.

    I will check if it is possible to use 1″ SRTM for UK_Scotland.map in future.
    So far I avoided SRTM and used Viewfinder Elevation Data cause SRTM is notorious unreliable.
    So using SRTM1 / 10m would mean to add LOT of Data/Mapsize and a fine appearence but questionable accuracy – in my experience.

    @sonny
    So far as I know the NASA Data still is based on the 99th SpaceShuttle mission with some German data X-SAR ??
    So with some math enhancement – but still based on faulty data ??
    Or happened there a new scan of the earth in the meanwhile.
    So: Is the “new” dataset (2015) really new and reliable or the n_th math/interpolation “enhancement” it was till 2015?

    Best regards
    Christian

    1 user thanked author for this post.
    #25972

    geoffmozz
    Participant

    Christian….thanks for replying. I think your maps are great and use them at home in Britain and for trips abroad.

    How much extra data would 10m interval contours require?

    eg If you have an OpenAndroMap with 20m contours that has a file size of 1GB, approximately how big would the file become with 10m interval contours?

    It would be great to have the contour detail provided by the OS50 data as seen in the WayMap in my first post. It’s what micro SD cards are for.

    • This reply was modified 1 week, 6 days ago by  geoffmozz.
    #25976

    Sonny
    Participant

    @ChristianK

    Yes, there’s just been one SRTM mission, the one in 2000.Btw: There are 2 intersting videos of a speech of the German astrounaut Gerhard Thiele who was part of the Space-Shuttle crew in this mission:
    https://www.youtube.com/watch?v=MK5Q0s0JDJM
    https://www.youtube.com/watch?v=UYrUAcD1wCQ

    These SRTM data have been published in differnt versions and resolutions in the past. The latest one is Version 3.0 Global 1 arc second. Compared to former version, this one is void-filled using elevation data from other sources. And for the first time 1″-resolution of the SRTM data is freely available outside the USA (before just 3″ models for e.g. Europe):
    https://earthdata.nasa.gov/nasa-shuttle-radar-topography-mission-srtm-version-3-0-global-1-arc-second-data-released-over-asia-and-australia

    Finally I created 2 screenshots to compare 20m contour lines of the Pillar-region based on Viewfinderpano’s 3″ model with SRTM 1″ model:

    SRTM1:

    Viewfinderpanorama (=SRTM3):

    We can see that SRTM1 data has more details, more corners. And is able to map steep ridges or gorges better than the former SRTM3 model. That’s not because the data is “better” (both have the same source), it’s just a matter of the sample resolution of both models (about 25 meter comparted to 75 meter)

    2 users thanked author for this post.
    #25981

    Sonny
    Participant

    @geoffmozz

    There’ve been some discussions in the past regarding 10m vs. 20m contour lines, for exmple:
    https://www.openandromaps.org/oam-forums/topic/neues-genaueres-hoehenmodell-fuer-oesterreich/page/6/#post-12291

    It’s not just a matter of more data as ChristianK already stated. It’s also an optical matter, depending on region and zoom. 10 m contour lines could quickly cover the map layout, disturbing the optical recognizing of other important features like ways, waymarks etc. For example even in such a topographic special country like Austria, the main Topomaps (Amap, Kompass, Alpenverein) have “just” 20m contour lines printed on.

    1 user thanked author for this post.
    #25983

    geoffmozz
    Participant

    Couldn’t the 10m contours just be revealed at higher zoom levels? The Garmin vector maps supplied by Talky Toaster do this and work well.

    Having read the Google translated version of the thread you linked to, it seems that 20m interval contours make up about 25% of the total file size. Let’s say 250MB of a 1 GB file. Presumably 10m contours would require double the contour data? ie 500MB/10m instead of 250MB/20m. That would make a 1GB map go to 1.25 GB with 10 m contours. That seems a tolerable increase in file size for better contours.

    I’d like to try a mapsforge map with 10m contours (revealed only at high zoom levels) and see how it performed. How much slower would it be to render on screen etc. Does a 10m contour interval mapsforge map exist?

    It may be that you’ve tried all this already and found it unworkable but your replies so far haven’t made me give up entirely…

    • This reply was modified 1 week, 5 days ago by  geoffmozz.
    #25986
    JohnPercy
    JohnPercy
    Participant

    I don’t find the 20m lines sufficient and have had to swap to OS maps to see the slopes when walking recently.
    i, too, would support 10m contour lines appearing at higher zooms. Would that reduce the amount of data packed into the maps themselves?
    However, there are two different issues being discussed here: contour spacing and resolution. Increasing the resolution from 3″ to 1″ would result in 9x the data unless there is some clever data packing. And I suspect that without increasing the resolution, reducing the contour spacing would result in more unsatisfactory artefacts.

    2 users thanked author for this post.
    #25988

    geoffmozz
    Participant

    Yes, I wasn’t paying too much attention to the technical implications of 10m interval contours. Partly because I don’t know enough about how contour detail and data are related. My queries are simply motivated by a desire for better contours.

    However, if getting the kind of contour detail seen in the WayMap that I first posted requires an unworkable amount of extra data then I’d be interested in seeing improved detail/resolution in the existing 20m interval contours. Perhaps that would have a tolerable data penalty?

    As a mountain walker/climber I’m aware that steep craggy ground is poorly represented and better contours would help overcome this. I picked the Pillar region in the Lake District as the 20m interval contours in the OpenAndroMap don’t give any clues as to the steep craggy nature of the area. The crag markings alone don’t represent the ground very well. See my screenshot of the OpenAndroMap. Here’s a picture of what Pillar Rock and it’s surroundings look like in real life…

    • This reply was modified 1 week, 4 days ago by  geoffmozz.
    • This reply was modified 1 week, 4 days ago by  geoffmozz.
    • This reply was modified 1 week, 4 days ago by  geoffmozz.
    Attachments:
    #25994
    ChristianK
    ChristianK
    Keymaster

    The source for the NASA 3.0 Files is
    https://e4ftl01.cr.usgs.gov/MEASURES/SRTMGL1.003/2000.02.11/
    .. is this the right dataset?

    As for the 10m lines: This will dramaticly increase the file size of the maps.
    And I have to increase tagmapping for indication of 50m- intervals and not only 100m as its actually the case.

    As for 1″ instead of 3″: Increases filesize but I have implemented an algorithm that eliminates unnecessary nodes – still an increase but not by factor 9.

    In general: OS Landranger are among the very best maps available round the globe – I have _lots_ of these maps here right at hand from the time I spent every summer vacation in Scotland.
    Vector maps can’t really match OS Maps – thats a fact.

    OS / MemoryMaps are still top an unmatched.

    I will give SRTM1/003 a try, at least for 20m lines.
    However this is cumbersome cause phyghtmap on windows cant bulk_download the tiles I have to do this by hand, so this will last a while….

    BTW: I’v seen that Shetland Islands are not covered by the 1″ Dataset.

    Best regards
    Christian

    Hmm.. If someone could download the GB SRTM 1″ Dataset and provide it on GDrive he would do me a great favor…

    #25996

    Sonny
    Participant

    Yes Christian, this is the right dataset. Maybe you could ask Derek of http://dwtkns.com/srtm30m/ for help who’s running a private download service of those tiles.

    Shetland islands are missing, cause it’s already slightly outside the SRTM-coverage (from 56°S to 60°N). In these regions the japanese ALOS-data could be a global alternative.

    #25998
    ChristianK
    ChristianK
    Keymaster

    Hi @sonny,

    Thanks again for support, I downloaded the GB_Dataset from Derek’s site.
    The Interface is perfect fine – just a few clicks to do.
    I will try to feed phyghtmap with this tiles.

    Would it be possible for you to provide missing (three?) tile of the shetlands as 1″ hgt’s?
    If not I will use the view3 data for this area.

    BTW: As far as I see there are serious problems with elevation data and melting glaciers.
    Best to be seen on South_Georgia.
    That’s really scary comparing (not so old) elevation data and nearly up2date aerial images.

    Thanks and best regards
    Christian

    #26000

    geoffmozz
    Participant

    Christian….thanks for trying this out.

    Just to be clear the WayMap I first posted is NOT an Ordnance Survey map although it does look like one. It just uses OpenData provided by the OS (crags and contours).

    These guys make the raster WayMaps using a mix of OpenStreetMap data and OS OpenData. They provide the WayMaps for WalkLakes. See the WayMaps radio button top right.

    • This reply was modified 1 week, 3 days ago by  geoffmozz.
    • This reply was modified 1 week, 3 days ago by  geoffmozz.
Viewing 15 posts - 1 through 15 (of 24 total)

You must be logged in to reply to this topic.