CHAHâ€“ALA collaboration: integrate the ALA gazetteer service into herbarium information systems
Maybe it is my imperfect knowledge of English, but I think 'integrate' is entirely the wrong word, so I am changing the brief to making the best possible use of the ALA gazetteer service in herbarium information systems. We will take herbarium information systems in the widest possible sense, so AVH, which does not belong to an individual herbarium is also a herbarium information system. Also, herbarium information services are not just collections databases: FloraBase, Plantnet, APNI, APG and AusMoss are herbarium information services, although not all of them are necessarily relevant in this context.
Below I will briefly summarise the services the ALA gazetteer project will provide. Please feel free to comment and indicate which of these services you think would be useful in your herbarium. Also, if you would like to use a feature, but are not able to, please indicate what the impediments are. This will provide me with an overview of the possibilities and challenges at the different herbaria that I can then dump on CHAH.
I would also like us to think about ways the Australian herbaria can contribute to the ALA gazetteer and further enhance its usefulness for the botanical community. That is in the next section.
Finally, there is a 'Comments' section, where you can put anything that may be relevant, however marginally, to this topic. This being a wiki page, feel free to put your comments anywhere else on this page that you think is more appropriate.
Features of the ALA gazetteer service
The services the ALA gazetteer provides are described in the wiki on the project's Google Code site and more in layman's terms below. Note that the examples on the project's wiki actually work, so you can try them out yourself. Also, when you follow the links on the results pages you get the detailed records.
List layers, retrieve layer data and feature lookup
Respectively give an overview of the layers the ALA gazetteer supports, list the features in a layer and give details of a feature. Probably not so useful for herbarium information systems and more relevant to the environmental map layers topic , but I imagine one could load the features into a spatial database. Interesting to have a look at anyway to see what is there.
This is really cool, as it returns all the polygons a set of coordinates is in. In the example in the wiki, a point search on 148.866°E 36.170°S returns New South Wales, Snowy River (not sure if that is the actual river, the catchment, or an administrative region) and South Eastern Highlands (IBRA region). Great for quality control.
You can also add a radius, or (im)precision if you like, to the query string. Adding a radius of 30 km to the above query adds Australian Alps to the results.
- It would be nice if you could add the layer (or layers) you want to search in the query string. [ Niels Klazenga, 29 December 2010]
This is most like the gazetteer searches we are used to. The ALA gazetteer does not offer resolution of locational text such as '29 furlongs SE in gulley from Oodnadatta Pub' (example not mine) as tools like BioGeomancer (I wouldn't bother looking that up, as it does not work and the site hasn't been updated since 2006) and GEOLocate (definitively look that one up) do and Lee Belbin, ALA contact and project lead, advises to use these tools, rather than the ALA gazetteer term search. ALA is prepared to consider providing its gazetteer data to these services.
- MEL will use GEOLocate. A GEOLocate plugin ships with Specify 6. [ Niels Klazenga, 29 December 2010]
- The BioGeomancer workbench is supposed to provide uncertainty (imprecision) as well, which would be interesting. At the moment, however, the BioGeomancer workbench doesn't do much of anything.
- I vaguely remember (from before I joined HISCOM) that at some point AVH was going to send its locality data to the BioGeomancer project for them to play with. What happened with that?
Closest feature search
This is basically a reverse geocoding service: you put in a set of coordinates and a radius you want to search within and it will provide details of the feature that is closest to the point indicated by the set of coordinates, as well as a distance and bearing. Optionally you can include a number of features to be returned and the layer you are interested in to the query string.
- In MEL we have been using reverse geocoding to obtain a generalised locality that we deliver to AVH for just over a year now (NearNamedPlace the MEL way). Generalised locality is provided with the public query results in AVH, while locality is not. It was decided at the last HISCOM meeting that the reverse geocoding would be used AVH wide for records that are not delivered with a generalised locality. MEL still uses the 1995 gazetteer, so if ALA provides a service that uses an up-to-date gazetteer we might as well use that. [Niels Klazenga, 29 December 2010]
- It would be nice if we could restrict the query result by feature class by adding that to the query string. Currently, to obtain the generalised locality we only query populated places. [Niels Klazenga, 29 December 2010]
How the Australian herbaria can contribute to the ALA gazetteer
Now here comes the really exciting bit. The best way for CHAH herbaria to contribute to the ALA gazetteer and at the same time provide a service to themselves and the botanical community, I think, would be to feed spatial information from their collections into the ALA gazetteer. I am thinking about collecting localities from, for instance, the Burke and Wills expedition, that can not necessarily be found in the GeoScience Australia gazetteer. When I was a PhD student in Leiden I frequently came across collections from, say, '..<super>th</super> Archbold Expedition, Lake Habbema, Camp B' and had to spend hours in the literature to find out where Camp B was (in Leiden we had the good (or bad) fortune that the literature was actually there). It would have been great if, after I had done that, I could put the information I found into a database that was available to everyone, so that the next person would not have to go through the same exercise.
It would be great to have as part of AVH (or VH?) a database with collecting localities and itineraries of Australasian collectors (or people that have collected in Australia and New Zealand rather), along with a database of collectors and expeditions, and feed the spatial information from this database into the ALA gazetteer. I think this would make a great addition to AVH.
When I suggested this to Lee, he was very interested. In fact, he is working on something like this for the limnological community. You can imagine existing gazetteers are not as useful for people working on marine and fresh-water organisms as they are for people that work with terrestrial organisms. Apparently, spatial data for fresh-water localities is even scarcer than for marine localities.
Stuart Pillman from the SA Department of Environment and National Resources (DENR) wrote this (in an email about the GEOLocate workshop):
The immediate areas we (DENR) have in mind are:
- Adelaide Herbarium specimen locations which are provided as text strings - validation of geocoordinate locations where these have also been provided; checking/validation of our existing manually determined
- geocoordinates (interpolated from descriptions); more efficient capture of location/coordinate information in the future
- community science based observations, such as images, which frequently have locations which are provided as text strings
- text string searches of existing databased scientific research descriptions which have no geocoordinates but have location descriptions
- extraction/generation of coordinates from digital literature.