Transcript EcoGRID: A problem solving environment for ecological - VL-e

EcoGRID: A problem solving environment for ecological modeling
J.Shamoun-Baranes, W. Bouten, N. Caarls, H. Sierdsema, F. Sluiter,
E. van Loon and G. van Reenen
Computational Biogeography and Physical Geography, IBED, University of Amsterdam, Nieuwe Achtergracht 166,
1018WV Amsterdam, The Netherlands
Introduction
Ecogrid case study
Over the centuries, a wealth of information has been collected by public and
private organizations around the world on the presence and abundance of local
flora and fauna. These data are typically distributed over many different
organisations. An important recent development is that this observation data is
becoming available via national, European and global biodiversity information
facilities. Simultaneously, massive amounts of spatially explicit data on site
attributes (e.g. climate, soils, topography, features recorded by remote
sensing) have become available via the Internet.
EcoGRID is a GRID and virtual lab technology based information system
and research environment for the management, integration and analyses of
observations and model results of the spatial and temporal dynamics of flora
and fauna distribution. It aims at incorporating novel technologies, easily
accessible data and models to enable the discovery of new ecological insights,
in order to cope with the ecological challenges that we are facing such as global
change, the spread of disease, decreasing biodiversity, and waning resources.
It has the ultimate goal to promote a breakthrough in the effectiveness of
decision and policy-making.
The spatial modelling tools developed for the EcoGrid PSE and the EcoGrid
database have been used to model the spatial and temporal distribution of 62
species of birds in the Netherlands for the Netherlands Bird Avoidance
Model (http://meridian.science.uva.nl/bambas).
The spatial distribution model, with GIS functionality is part of the NL-BAM
decision support system and will be used as a strategic planning tool for training
schedules as well as habitat management regimes on airfields to improve flight
safety in the Royal Netherlands Air Force. The methodologies used to develop
this system can easily be adapted to the requirements of civil aviation and the
development of small-scale local BAMs.
The Ecogrid framework
Within the EcoGrid framework, web services are being developed for a) the
input, management and quality assurance of observational data (>10.000
users), b) for scientific research, and c) for public use of processed data. Within
this framework the National Database of Flora & Fauna in the Netherlands was
developed, integrating the distributed and previously non-uniform databases of
nine private organisations responsible for data collection, allied in the
Association for Research of Flora & Fauna (VOFF). The EcoGrid database already
contains 90% of all national spatio-temporal observations of species
(10.000.000 records) from different taxa, for example fungi, lichens, plants,
mollusca, insects, mammals, birds, reptiles, fish and amphibians. It has the
flexibility to expand by including other organisations or institutes in the
Netherlands and abroad.
Data Portal
National flora & fauna
database
Data entry
Data QA
plants
dragonflies
fungi
butterflies
mollusca
moss &
lichens
Fish +
Web services
Public data access
Mapping tools
Summary statistics
Science Portal:
Vl-e science problem solving environment
Ecogrid
Virtual
Database
mammals
Weather
Taxonomy
Social
economic
The spatial modelling toolbox, used to develop the BAM, enables the spatial
modelling of species abundance using several modelling techniques such as
regression statistics, spatial interpolation, autoregressive models as well as an
integration of various methodologies. The toolbox includes a calibration and
validation phase. Model output is visualized in different ways at various points
during the modelling procedure to facilitate interpretation and feedback from
the user. Input variables are selected by the user and can include diverse
spatial data, for example abiotic and biotic environment data including
anthropogenic features. Therefore, the toolbox is easily applied to spatial
modeling of other groups of flora and fauna.
Visualization
GIS
Data mining
birds
Geostatistics
Modelling
Scale
conversions
Landscape
Spatial modeling toolbox
Processed
data
Products
The problem solving research environment is developed to streamline both the
methodological and computational process of analyzing and modelling spatial
observational data. Within the PSE various methodologies and generic tools are
developed and applied in case studies. The PSE includes:
1. Methodologies and generic tools for statistical analyses and visualization of
spatio-temporal data.
2. Methodologies and efficient algorithms for spatial data-mining
3. Methodologies and generic tools for simulation modelling, sensitivity
analyses, and combined interpretation of observations and models results
through data-assimilation and inverse modelling.
Vl-e and GRID technology
EcoGrid shares the characteristics of other data intensive sciences where the
data is collected and stored in geographically distributed databases and is
analysed while using remote resources such as cluster computers or Grid
facilities. The use of distributed databases requires fast network connections
and huge temporal storage capacity. Furthermore, modelling and data
exploration tools such as spatial data-mining on large datasets, simulation and
inverse modelling needs extensive memory allocation and computational power.
There is a growing need for efficient tools that can visualize, during the
modelling process, the multiple perspectives of dynamic model input, parameter
evolutions, model results and fusion of measurements and model output.
J.Shamoun-Baranes et al.
CBPG, IBED, University of Amsterdam,
Nieuwe Achtergracht 166, 1018WV
Amsterdam, The Netherlands
Email: [email protected]
URL: http://www.vl-e.nl/
This work was carried out in the context of the
Virtual Laboratory for e-Science project. This
project is supported by a BSIK grant from the
Dutch Ministry of Education, Culture and Science
(OC&W) and is part of the ICT innovation
program of the Ministry of Economic Affairs (EZ).