---------- Forwarded message ----------
Date: Wed, 23 Sep 1998 16:35:48 +0100
From: Darius Bartlett <[log in to unmask]>
To: Maps and Air Photo Systems Forum <[log in to unmask]>
Subject: Re: details to mean sea level, Alicante: reply
 
At 09:36 23/09/98 -0400, Andrea Hausold wrote:
>---------- Forwarded message ----------
>Date: Wed, 23 Sep 1998 09:16:39 +0200 (DFT)
>From: "A. Hausold" <[log in to unmask]>
>To: [log in to unmask]
>Subject: details to mean sea level, Alicant
>
>Dear members of the list,
>
>first of all thanks a lot to all who have already responded to my request
>and gave hints.
>
>On reading the answer of Darius it seems that I might be on the wrong
>track. If you do not mind, I would like to explain want I am planning to
>achieve. Maybe you then have even more advise for me?
 
I'll do my best. What follows is quite long: those not particularly
interested in basic geodesy may hit the delete button now! :-)
 
>I have airborne scanner data of a testsite in spain and I want to do a
>geometric correction. For doing this I need the flight path, derived from
>differential GPS measurements/calculations, and some ground control
>points. The flight path and the GCPs have to be in the same geodetic
>system (ellipsoid, datum) concerning position and height.
>
>The DGPS data refer to WGS84 in position and height (so these are
>ellipsoidal heights, am I right?).
 
Yes. But WGS84 is just one of many possible ellipsoids (see below). It
happens to be the one that has been adopted as "standard" for GPS and
similar applications.
 
>The map is based on Hayford Ellipsoid, European datum, height reference is
>mean sea level in Alicante. So what does this mean for the heights? Maybe
>I got this wrong.
 
The Earth, as we all know, is not a perfect sphere. One approximation is
that it is an oblate spheroid - an ellipsoid, with a major and a minor
axis, and a defined centre that is some distance removed from what would
be the centre if the Earth was a true sphere.
 
There are many different ways of defining these ellipsoids, depending on
where you place the origin and where the resulting axes intersect the surface
of the earth. In general terms, conversion from one ellipsoid to another is
normally achieved by using what is known as the Molodensky Datum Transfer
Proceedure, or Molodensky Constants.  These constants comprise three
values - the Delta-X, Delta-Y and Delta-Z shifts required to convert
from WGS84 to any thus-defined new datum.
 
If you have access to the manual for IDRISI for Windows version 1.0 or
above, Appendix 2 has a comprehensive list of these Constants for a
very long list of alternative geodetic datums. Unfortunately, though,
the Hayford Ellipsoid does not appear to be one of these, and I have no
source of reference for these myself. The writers of the IDRISI manual
actually cite their own source as being "a data file accompanying the
MAD-TRAN V.9109.04 datum conversion software (1992) available from
the US Defense Mapping Agency (DMA Stock No MADTRANIBMPC Edition No. 002)"
as the source of their data, so it may be possible that the parameters
you require are in this original document.
 
>So what can I do to get the height informations in the same system?
 
In a nutshell, if you can somehow track down the relevant Molodensky
Constants for the Hayford Ellipsoid, European Datum, this should give
you the conversion details you need. Then, most GIS software can perform
the required adjustments - Idrisi and Arc/Info both have facilities
whereby you feed in the constants and the software does the rest.
 
>And - I hope not to bother you too much - one more question, maybe I have
>not understood this so far:
>
>is the "geoid level" the same as "mean sea level". Doubts arose, because
>there exist so many "mean sea levels", depending on the country.
 
No, the geoid level is very different from mean sea level. As I said above,
the Earth is not a sphere. For most applications, including most
topographic mapping, it is sufficient to approximate the shape of the
Earth to one of the ellipsoids mentioned above. The actual choice depends
on whereabouts on the planet you happen to be. However, there are times
when a very much more detailed, "precise" definition of the shape of the
Earth is required: this is the geoid (the name actually means
"Earth-shaped" :-)
 
A formal definition of the geoid is that it is a sea-level gravity
equipotential surface for the planet. Robinson et al (Elements of
Cartography, 6th Edition, 1995, p44) say of the geoid that it is "the
three-dimensional shape that would be approximated by mean sea level
in the oceans and the surface of a hypothetical series of sea-level
canals crisscrossing the continents". They also add the information
that the geoid surface can deviate from the ellipsoid by up to 100m
in certain locations, so it is clearly a factor to be taken seriously.
 
Also, as you say, there are many "mean sea levels" - depending on the length
of the time series of measurements is used for computing the "mean", on the
actual tidal range (very different in the micro-tidal location of Alicante
and the more macro-tidal range of the North Sea for example), and depending
on which phases and states of the tide are used as minima and maxima (spring
tides, neap tides, ordinary tides?).
 
>Maybe these questions are too basic, then I am sorry for that and I would
>be grateful for a hint on good literature concerning this.
 
No, not too basic at all - indeed this is something I think is very
often overlooked and misunderstood. I would only claim to know the very
basic principles myself. If you do want to know more, I recommend the
book by Robinson et al cited above as a good introduction to some of
the issues. They also cite the following, which I have not seen myself,
but which looks as if it could be useful:
 
Langley, R.B. "Basic Geodesy for GPS". GPS World 3 (1992) 38-43
 
Finally, "Photogrammetric Engineering and Remote Sensing", the journal
of the American Society for Photogrammetry and Remote Sensing (ASPRS)
has a very useful regular column on the subject of geodetic datums by
Clifford Mungier at the University of New Orleans.
 
Hope this helps!
 
Darius
************************************************************************
Darius Bartlett                                          Darius Bartlett
Department of Geography                           Roinn na Tireolaiochta
University College Cork                   Colaste na hOllscoile Corcaigh
Cork, Ireland                                             Corcaigh, Eire
 
Phone: (+353) 21 902835                             Fax: (+353) 21 271980
Mobile: (+353) 86 8238043
E-mail: [log in to unmask]     Web URL: http://www.ucc.ie/ucc/depts/geography/djb
-------------------------------------------------------------------------
This message was transmitted using 100% recycled electrons....
*************************************************************************