2–6 Mar 2009
Le Ciminiere, Catania, Sicily, Italy
Europe/Rome timezone

On the D4Science Approach Toward AquaMaps Richness Maps Generation

5 Mar 2009, 16:40
20m
Michelangelo (120) (Le Ciminiere, Catania, Sicily, Italy)

Michelangelo (120)

Le Ciminiere, Catania, Sicily, Italy

Viale Africa 95100 Catania
Oral Planned or on-going scientific work using the grid New Application Areas

Speaker

Dr pasquale pagano (CNR-ISTI)

Description

AquaMaps are resources resulting in Earth maps enriched to show the likelihood of a certain species or a combination of species to live in specific regions. The procedure leading to such maps is computationally intensive and prone to be executed multiple times to enhance the quality of the predictions. The results obtained by exploiting the EGEE production infrastructure through the D4Science Virtual Research Environments (VREs) to produce collections of such maps are presented.

Detailed analysis

AquaMaps are very important resources for species assessment. These products are requested to document the large-scale and long-term presence of species in certain regions of the world. To produce such predictive maps two complementary procedures are needed. One is dedicated to estimating the probabilities of species occurrence by matching the species environmental envelope. The other is dedicated to the actual production of the maps by plotting the estimated values on Earth maps. Both procedures require computationally intensive crunching activities that fit well with the properties of the EGEE grid infrastructure. This contribution reports on the crunching activity for supporting the second procedure of AquaMaps production. It exploits a data source capturing the data of more that 9000 species and consisting of more than 50 million probability entries to produce millions of products. Each generated product consists of several images and is stored and accessed through a VRE.

Impact

AquaMaps is a service widely used within the biodiversity community as a basis for species distribution inference. The service is currently used by other large-scale services like FishBase serving approximately one million users per month. One of the AquaMaps bottlenecks is the time needed to generate the richness maps a scientist can be interested in by crunching data per species (or set of species) to produce aggregated view per Kingdom, Phylum, Class, Order, Family or Species. D4Science proposes a different approach consisting in (i) the exploitation of EGEE grid facilities to produce such maps and (ii) by relying on D4Science data management facilities, the provision to scientists of tools for searching the maps they are looking for. Actually, a new large community is starting to appreciate EGEE potentialities while scientists are provided with compound information objects containing additional data and multiple views of the maps, thus improving the resulting quality of service.

URL for further information

Project Web Site: http://www.d4science.eu
http://www.aquamaps.org/

Keywords

Richness Species Distribution Map Generation, D4Science, AquaMaps, Biodiversity

Conclusions and Future Work

The service resulting from the synergic use of D4Science and EGEE technologies to supply scientists with species distribution maps represents an effective enhancement for the biodiversity community since its users are provided with richer information in a shorter time period. This approach sets the scene for other experimentations including the exploitation of grid facilities to generate alternative estimation data (and maps) resulting from other (more complex) predictive models.

Author

Dr pasquale pagano (CNR-ISTI)

Co-authors

Mr George Kakaletris (University of Athens) Mrs Leonardo Candela (CNR-ISTI) Mr Pedro Andrade (CERN)

Presentation materials