26.06.2013 Environmental Data Enrichment
From IMarine Wiki
Time: Thu, June 26, 2013, 16:00 - 17:00 CEST
- Environmental Data Enrichment
- Anton Ellenbroek (FAO)
- Fabrice Brito (Terradue)
- Pasquale Pagano (CNR)
- Gianpaolo Coro (CNR)
- Edward Vanden Berghe
The objective of the meeting was to discuss progress with the environmental data enrichment.
Edward recalled what the objective of this activity was, and how it fits into the ‘big plan’ of the biodiversity cluster within iMarine.
The present environmental envelope modeling tools within iMarine (Aquamaps, openModeler) take environmental/oceanographic data from aggregated sources – mainly from the World Ocean Atlas; oceanographic parameters are taken from the 1 degree cell in which a biological observation falls (though Aquamaps has some layers on a quarter degree). Obviously, this is very crude, and definitely too crude to do any work in coastal areas. The idea is to replace the aggregated sources with the sources of the point data (in case of the World Ocean Atlas, for example, replace it with the World Ocean Database).
This necessitates an infrastructure where we can interpolate from 4D points of the raw data sources to the 4D points of the biological observations. This would result in a much closer coupling between oceanographic and biogeographic data, and so, hopefully, in better predictions that would also be valid in coastal areas. A gridded/aggregated version of the environmental data will still be needed for the extrapolation/mapping step of the modeling.
Fabrice explained that this was still very well understood by Terradue. Terradue has developed some experimental/demonstrator connectors with oceanographic data, including the datasets available within BioOracle. They are now ready to create links to other sources of oceanographic data. They are also ready to start providing services to extract data in a 4D box around 4D points of interest (e.g. 4D points of the biogeographic data set to serve as input in modeling). They will not, however, deal with 4D interpolation. In principle, the connection to all oceanographic data will be dynamic, without long-term local storage; this will bring advantages in terms of storage, and in terms of keeping data in sync with repositories of oceanographic data.
We need to compile a list of priority datasets to be connected by Terradue to the infrastructure; Edward will compile such a list, and where he’s aware of this, a potential source for the data; also an indication of which layers would be important to model which groups (e.g. bottom vs water column, and calcifiers vs non-calcifiers).
We need to be able to interpolate from the 4D points in the raw oceanographic data sets to the 4D points of the biogeographic observations. IRD might have expertise, but this is not clear. Edward will contact Julien Barde on this. NODC/WDC in Silver Spring clearly has this capacity, as they do it to generate the WOA out of the raw data in WOD; Edward will contact them.
Terms of reference for Edward’s contract include a study of the effects of the improved resolution of the environmental data enrichment on the predictions of the environmental modeling. It is clear that this is critically dependent on the availability of these improved-resolution data.