Georreferencing of collecting localties

[Ted Oliver]

Hi all

I agree with what has been said about the problems of georeferencing of old collections ..... it is not an easy task and requires LOTS of patient personal checking. 

We are currently databasing our herbarium collections and putting co-ordinates on the old collections, many of which are rather vague. We have a 'degree of accuracy' box which tells us whether we are using the centroid of a 16th degree (1:50 000 ordinance map), a district or a local cadastral unit or are using full GPS readings or ones worked out by the collector from a 1:50 000 map.

We have the full cadastral/physiographic electronic map in ArcView linked to the database and putting the georeferences with each collection is relatively easy.

Here we have the choice in GPS systems to use the Cape datum. All the South African ordinance maps are based on the Cape system. There was talk of changing it to the international system which could cause problems for localities. I have not worked out what the differences will be.

In the original PRECIS system of the National Herbarium, Pretoria, we were faced with hundreds of thousands of records in the database with no georeferences.  A system was developed some 20 years ago using a country gazetteer with quarter degree grids to update all major localities with the grid square. This saved an enormous amount of time and was reasonably successful. Needless to say there were/are some anomalies but these are gradually being resolved as time goes by .... once again lots of personal effort.

Ted Oliver 



----------------------------------------------------------------
Dr. E.G.H. Oliver
Compton Herbarium  (NBG)
Kirstenbosch Research Centre
National Botanical Institute
7735  CAPE TOWN

e-mail:  oliveregh at nbict.nbi.ac.za
Tel.  021 799 8724
http://www.nbi.ac.za/research/comptonherbarium.htm

The NBI accepts no liability for unauthorized use of its e-mail facility nor for corrupted or virus-infected messages.