Re: [Qgis-user] about some ellipsoïdal distances

[G. Allegri]

Probably I've found the bug.
In Qgis the semi-minor axis of the ellipsoid is calculated with:

b = a - (f/a)

where
b = semi-minor axis
a = semi-majot axis
f = inverse flattening

while it should be:

b = a - (a/f)


In Qgis the WGS84 semi-minor axis is 6378136,xxx
while it should be 6356752.xxx

giovanni





2012/4/6 G. Allegri <giohappy at gmail.com>

> AFAICS, the Qgis formula [1] to calculate distances on ellipssoid is
> Vincenty's.
> I've reproduced it for the first set of points and I can confirm tha QGis
> gives 540.9973919726176 m
>
> The online Vincenty calculator [2] gives 540.045 m
>
> I suppose it's a rounding issue.
> I will investigate it more...
>
> giovanni
>
>
> [1]
> http://trac.osgeo.org/qgis/browser/trunk/qgis/src/core/qgsdistancearea.cpp#L540
>
> [2] http://www.movable-type.co.uk/scripts/latlong-vincenty.html
>
> 2012/4/6 blavet <didier.blavet at ird.fr>
>
>> **
>> Dear QGis Developpers and Users
>>
>> First of all  many thanks to the Developpers of Qgis for this wonderful
>> and user friendly Sofwaret. I'm sure that Qgis will become more and more a
>> "standard" in the next few years
>>
>> So, just a small point  (and I hope to be  wrong with this, but if it is
>> the case i don't see where) : as i needed to find quite exactly some
>> ellipsoïdal distances, I'm afraid to have also found that  in some cases
>> the ellipsoïdal distance could have some problem in QGIS 1.7.4 (tested
>> versions Windows and Debian). This could depend on the algorithm used and
>> the geographical position of the points.
>>
>> The following table suggests for few studied points that the ellipsoïdal
>> distances computed in Qgis 1.7.4 was  overestimated and quite different
>> than ellipsoïdal distances computed with the help of two geodesic
>> estimations
>>
>>   *point name in the study* *East EPSG:27572
>> (m)
>> * *North EPSG:27572
>> (m)
>> * *Lat WGS84 (obtained from deprojection in latlon WGS 84 in QGIS 1.7.4)
>> deg. dec.
>> * *Long WGS84 (obtained from deprojection in latlon WGS 84 in QGIS 1.7.4)
>> deg. dec.
>> * *Simple Pythagore dist
>> (m)
>> * *Simple Pythagore dist QGIS 1.7.4
>> (m)
>> * dist ellips Clarke 1880 estimated from linear alteration given by
>> geodesic soft 1 (IGN Circé v4)
>> (m)
>>  dist ellips WGS84 from geodesic Vincenty formula recommanded by proj4
>> (found in soft 2 at
>> http://www.movable-type.co.uk/scripts/latlong-vincenty.html
>> (m)
>>  *dist ellips QGIS 1.7.4 Clarke 1880
>> (m)
>> * *dist ellips QGIS 1.7.4 WGS84
>> (m)
>> *
>>  *
>> * *
>> * *
>> * *
>> *
>>
>>
>>
>>
>>
>>   *B (-)9* 733080.2641 1847716.595 *43.61880918* *3.983106135* 540.795
>> 540.795 540.048 540.045 *541.016* *541.000*  *AC (-)9* 732806.9273
>> 1848183.228 *43.62305376* *3.979845905*  *
>> *
>>
>>  *
>> * *
>> *
>>
>>
>>
>>
>>
>>   *A 0* 733245.932 1847790.425 *43.61944142* *3.985174410* 560.824
>> 560.824 560.049 560.046 *561.054* *561.038*  *AC 0* 732962.4713
>> 1848274.34 *43.62384326* *3.981793539*
>>
>>
>>  *
>> * *
>> *
>>
>>
>>
>>
>>
>>   *B 8* 733374.0699 1847888.697 *43.62030034* *3.986784862* 540.794
>> 540.794 540.047 540.043 *541.016* *541.001*  *AC 8* 733100.7324
>> 1848355.328 *43.62454501* *3.983524812*
>>
>>
>>  *
>> *
>>
>>
>>
>>
>>
>>
>>   *C (-)9* 733070.1404 1847733.878 *43.61896639* *3.982985394* 380.560
>> 380.560 380.034 380.031 *380.715* *380.692*  *C 10* 733398.5116
>> 1847926.227 *43.62063297* *3.987096932*
>>
>>
>>  *
>> * *
>> *
>>
>>
>>
>>
>>
>>   *AC (-)13* 732737.7966 1848142.733 *43.62270285* *3.978980306* 420.617
>> 420.617 420.036 420.034 *420.790* *420.765*  *AC 8* 733100.7324
>> 1848355.328 *43.62454501* *3.983524812*
>>
>>
>> Best regards
>>
>>
>> _______________________________________________________________________________________
>>
>> *Didier B**lavet*
>>
>> Institut de Recherche pour le Développement - IRD (http://www.ird.fr)
>>
>> UMR Eco&Sols - Ecologie Fonctionnelle & Biogéochimie des Sols &
>> Agroécosystèmes
>>
>> (Montpellier SupAgro-CIRAD-INRA-IRD) (
>> http://www.montpellier.inra.fr/ecosols)
>>
>> Bâtiment 12, 2 place Viala, F-34060 Montpellier Cedex 2 , France
>>
>> Tel bureau : +33 (0)4 99 61 21 33
>>
>> Secrétariat : +33 (0)4 99 61 21 01 / Fax : +33 (0)4 99 61 21 19
>>
>> Courriel : didier.blavet at ird.fr
>>
>>
>> _______________________________________________________________________________________
>>
>> _______________________________________________
>> Qgis-user mailing list
>> Qgis-user at lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/qgis-user
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-user/attachments/20120406/6376cab8/attachment.html>