› ›

Maps with datum Hong Kong 1929

wmiswmwmiswm Global Mapper UserPosts: 220Trusted User
edited September 2011 in Georeferencing Imagery Discussion
I have scans of some 50 years old maps that I need to rectify so as to compare them with modern maps of today. These maps are supposed to be using the Hong Kong 1929 datum. Please find attached the legend of the maps to show the datum, projection, spheroid, false Northing and Easting, origin etc.

map legend.jpg

I dictate some relevant sentences below:
1. The graticule on this sheet is in terms of The Hong Kong 1929 Datum and is in sympathy with the graticule on the Admiralty Charts of the area
2. Grid data: Universal Transverse Mercator Grid Zone 50
Projection
Transverse Mercator
Spheroid
International
Origin
117° E
Equator
False co-ordinates of origin
500,000 m.E. 0 m.N.


Immediately after loading a scanned tif file, the [Image Rectifier (automatic)] sheet appears.
On [Image Rectifier (automatic)], I click on "Select Projection".
On [Select Control Point Projection], I select
Projection: Transverse Mercator
Datum: since GM does not support the datum "Hong Kong 1929" directly, I suppose I have to click "Add Datum" Planar Units: ??

The Create New Datum sheet appears. Then, I am sorry that I have to ask stupid questions as the parameters showed in the map legend cannot be translated to the parameters that I have to fill in GM directly. Please advise on how to fill in the required parameters.

Datum Name: Hong Kong 1929
Abbreviation (optional): HK1929
Prime Meridian (Degrees): ??
Ellipsoid (Speroid) selection: International 1909 (Hayford/Intl 1924)
Datum Transformation Method:
3-parameter transformation (??)
7-parameter transformation (??)
Custom shift based on control point file (??)
Shifts to WGS84 (meters)
X shift: ??
Y shift: ??
Z shift: ??
Rotation to WGS84 (arc-seconds)
X: ??
Y: ??
Z: ??
Scale correction to WGS84 (parts per million)
Scale (ppm): ??

Please help. Thank you very much.

Comments

  • global_mapperglobal_mapper Administrator Posts: 17,238
    edited August 2011
    I can't find the datum transformation parameters for Hong Kong 1929 anywhere either. I would suggest trying Hong Kong 1963 as it is based on the same ellipsoid, so perhaps they are different names for the same datum.

    Let me know if I can be of further assistance.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • wmiswmwmiswm Global Mapper User Posts: 220Trusted User
    edited August 2011
    Thx Mike, but the map was produced in 1957, earlier than the introduction of "Hong Kong 1963". I suppose there must be something different between "Hong Kong 1929" and "Hong Kong 1963", otherwise need not the change. By the way, if I ask the Hong Kong Lands Department on this, how should I ask them? What parameters of the HK1929 should they provide to make GM work in my case?
  • global_mapperglobal_mapper Administrator Posts: 17,238
    edited August 2011
    I would think it's different too, but I can't find a description of the transformation parameters anywhere. Getting the ellipsoid right will get you somewhat close (i.e. probably 10m or so), but if your map is detailed enough you will want better. You would need to ask for the XYZ shifts for a 3-parameter Molodensky transformation.

    Let me know if I can be of further assistance.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • wmiswmwmiswm Global Mapper User Posts: 220Trusted User
    edited August 2011
    Thx Mike. I am asking the Lands Department on the parameters. One further question, if Lands Department cannot provide the shift parameters (to WGS84), what other alternative can be available? Say, is it possible to create a HK1929 projection file, what parameters will be required instead?
  • global_mapperglobal_mapper Administrator Posts: 17,238
    edited August 2011
    Without the parameters your only other reasonable option would be to get at least 3 or 4 points in the HK1929 datum and in WGS84, then you could add a custom datum from a control point file. You could do this using your map and some other map in WGS84, like digital imagery. It would be best to just get the XYZ transformation parameters though.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • wmiswmwmiswm Global Mapper User Posts: 220Trusted User
    edited August 2011
    Thx Mike. What is the best number of control points you recommend that I would have to pick? I can actually pick as many as you recommend (of course with accuracy limited to that of map reading).
  • global_mapperglobal_mapper Administrator Posts: 17,238
    edited August 2011
    It really depends on how large of an area you are covering, but I'd say a minimum of 3, with 5 or 6 being better. You could maybe get away with just 2 for a small area.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • wmiswmwmiswm Global Mapper User Posts: 220Trusted User
    edited August 2011
    Thx Mike, they are 1:20000 maps covering the whole Hong Kong of about 1,100 sq km, cut into 24 pieces. So how many control points would you recommend? Actually, is it the more control points the better?
  • global_mapperglobal_mapper Administrator Posts: 17,238
    edited August 2011
    Generally the more the better, but for just the Hong Kong area probably more than 4 is overkill.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • wmiswmwmiswm Global Mapper User Posts: 220Trusted User
    edited August 2011
    hi Mike, in search for the HK1929 datum information, just come across this piece of information:

    http://www.geodetic.gov.hk/smo/gsi/Data/pdf/explanatorynotes.pdf

    Do you think this can help?
  • global_mapperglobal_mapper Administrator Posts: 17,238
    edited August 2011
    Unfortunately I don't see anything about the HK1929 datum in there.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • global_mapperglobal_mapper Administrator Posts: 17,238
    edited August 2011
    Unfortunately I don't see anything about the HK1929 datum in there.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • wmiswmwmiswm Global Mapper User Posts: 220Trusted User
    edited August 2011
    hi Mike, I got a reply from the Lands Dept.
    Please refer to Grid and Datum section of our website http://www.geodetic.gov.hk/smo/gsi/programs/en/GSS/grid/refdoc.htm to download the 7-parameters "Geodetic Datum Transformation and Map Projection Parameters" for the transformation of Hong Kong 1980 Geodetic Datum to WGS84.
    We have to know the reference ellipsoid in working on geodetic / geographical (latitude and longitude) datum. Referring to the footnote of your old map, the Spheroid (i.e. ellipsoid) is International, it is the same reference ellipsoid "International 1910" as the Hong Kong 1980 Geodetic Datum.
    So, you can use our 7-parameters shown in "Geodetic Datum Transformation and Map Projection Parameters" for the transformation of your old map to WGS84.

    Any comment?
  • global_mapperglobal_mapper Administrator Posts: 17,238
    edited August 2011
    Just because the datums use the same ellipsoid doesn't mean the same transformation parameters will work, that is not correct at all. There are lots of datums based on the International 1909/1924 ellipsoid, all with different transformation parameters to complete the transformation description. Sounds like maybe they don't really know either.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • wmiswmwmiswm Global Mapper User Posts: 220Trusted User
    edited August 2011
    This is also my thinking, quite disappointing.

    I am thinking an alternative way to do this more accurately. The problem is that I only have a scan of the HK1929 datum map. There are referencing grids printed on the map, but to obtain the co-ordinate (in HK1929) of those control points (that I can define the WGS84 co-ordinates accurately), it is very inaccurate just by map reading. I am thinking of using datum of the same ellipsoid as HK1929, say the HK1980, first rectify the HK1929 map into the shape of WGS1984 (using grid intersection as control points). Since HK1929 and HK1980 used the same ellipsoid, the shape of the image of the map, i.e. the projection and scaling, will be the same for the two datum, though the co-ordinate will read differently. After getting the HK1929 map into the shape of WGS84, I can then remove the georeference data from the generated tif image. Then, I can use accurately determined control point file to re-calibrate the co-ordinate of the image. Do you think that this approach can work?
  • wmiswmwmiswm Global Mapper User Posts: 220Trusted User
    edited August 2011
    wmiswm wrote: »
    control point file to re-calibrate

    I mean control points
  • global_mapperglobal_mapper Administrator Posts: 17,238
    edited August 2011
    There won't be any significant shape difference between HK1929 and WGS84 or HK1980, mainly just a shift for such a small area. I would suggest loading some other known data for that area (perhaps something from the File->Download Online Imagery/Data menu command), then rectify your scanned image against that by selecting common points, then just let the warp of the rectify take care of things.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • wmiswmwmiswm Global Mapper User Posts: 220Trusted User
    edited September 2011
    The Lands Dept supplement me with further data.
    I tabulate below the difference between the two datums.
    Datum________________Ellipsoid_________________Semi-Major Axis____Flattening
    Hong Kong 1929 Datum__Clarke 1880 (RGS)_________6378249.145m_____1/293.465
    Hong Kong 1980 Datum__International Hayford 1910__6378388m_________1/297
    Any clue for an accurate rectification?
  • global_mapperglobal_mapper Administrator Posts: 17,238
    edited September 2011
    According to that the Hong Kong 1929 datum is based on the Clarke 1880 (RGS) ellipsoid and not the International/Hayford ellipsoid. You could create a new custom datum that uses the Clarke 1880 (RGS) ellipsoid and just no transformation parameters (i.e. all 0), so at least the ellipsoid would transform correctly, but you really still need the additional transformation parameters, which I cannot find anywhere.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
  • wmiswmwmiswm Global Mapper User Posts: 220Trusted User
    edited September 2011
    thx Mike, I'll try. Should I re-calibrate after the ellipsoid transformation?
  • global_mapperglobal_mapper Administrator Posts: 17,238
    edited September 2011
    What do you mean by re-calibrate? I would really suggest just rectifying against some known imagery and really removing the source datum from the equation altogether.

    Thanks,

    Mike
    Global Mapper Support
    support@globalmapper.com
Sign In or Register to comment.