Pozzi - what can be done for the Beijing GEO ministerial summit? It takes place first week in November 2010. Implies demo preparation in October.
Pozzi email: Both the EDO and the Princeton soil moisture and soil moisture anomaly fields (and drought index) are prepared using models, the Variable Infiltration Capacity (VIC) model in the case of Princeton. Both sets of models take inputs, such as precipitation fields, vector wind, specific humidity fields, solar and terrestrial radiation, soil porosity, vegetation stand height, root zone depth, and other variables to calculate soil moisture and soil moisture anomalies.
Under Stefano's scenario, EDO provides drought indexes (step 01), and in step 02 the drought hazard detection model is used to estimate drought hazard: if a threshold is exceeded (or a hazard is detected), an alert is issued (to the "decision support system").
Under Stefano and Stefan's scenario, in step 03, a user submits a query to discover drought hazard related information. The augmented search component includes an ontology service.
For this system to work, there will have to be an ontology for the hazards (I'm speaking of socio-economic hazards here, not agricultural crop vulnearbility), an ontology for crops and agriculture, and an ontology for water. The term for precipitation (and its sub classes) will have datasets registered to it, and these are the same datasets that are used as driving datasets in the EDO, the Princeton, (and other) drought models. The agricultural ontology will include phenology (growth stage of the crop) and will be crop specific.
There is several other issues here, however. Since both the EDO system and the Princeton system (as well as NIDIS and the Australia BoM) are near-real-time, these values have to be updated. Hence a catalog updater is required (as well as a catalog publisher, required to extend the catalog (and variables). To meet this task, "web services" are suites of code for accessing remote data repositories, in order to update these values, not only the secondary data from the models (VIC, etc), but the primary data (precipitation, streamflow for validatiion, etc). The amount of effort to set up a system is huge and will require multiple data transfer standards (including the water data transfer standard, Brad), in order to automate the process of updating the holdings.
Stefano - European Drought Observatory should be able to publish data on specific areas, e.g. Spain and Italy. Proposal to focus on Spain for an AIP-3 scenario
François - EO2HEAVEN will work on identifying datasets and refining scenarios, but does not have scope within AIP-3 to stand up their own software. Agreement with ESA to access data sets (Veronica joins).
Veronica - Discussion with Damien - Neventropic... about correlation work between water quality and cholera. Suggestion to make use of ESRIN SSE infrastructure for data / processing / model services.
Josh - It would be a good focusing point to publish a water community catalog, possibly as a Web Accessible Folder, but preferably as a CSW catalog. Good to distinguish that from deployment of access / processing services. So, issues include identification of relevant data types, available datasets, available IT infrastructure for cataloging, serving, processing, visualizing / analyzing / deciding.
WIll - Third issue is development and publication of applications / portals to leverage these services. Veronica: most of visualization may have to wait for AIP-4.
Josh - There can be a spectrum of "models" between exploratory statistics and policy decision support. A scenario of publishing and exploring data for drought / health correlations is entirely appropriate for a November target.
Will - A very common denominator is the development and publication of a model processing use case. Josh - topic for E2E call on Monday. Good to distinguish between GEOSS publication of modeling (inputs, outputs, algorithm, metadata) and deployment of model processing to the Web itself for general use. The former is essential. The latter will be nice to have in the future.
Stefano - Vocabulary systems for drought, also a research topic. Will - work on CUAHSI ontology to expand it for drought, drought risk, and health risk terms.
Josh - Piasecki will discuss development of CUAHSI HIS ontology at the Geosemantics Summit in Silver Spring - http://www.ogcnetwork.net/geosemantics
Josh - Water Quality (Gulf of Maine) scenario still under development, but will be vetted a bit at NeCODP meeting next weekend.
Stefano - ontology logistics: SESAME stack for managing knowledgebases, SPARQL for query. RDF and SKOS files (e.g. GEMET) but which tools are used to create / manage GEMET is not known.
Veronica - will bring up maritime perturbation as a possible scenario addition (email and topic for next week)
Telecon next week.
Further work pulling together scenarios, defining a scope in each for a demonstration by Beijing meeting.
Research on infrastructure (catalog, access, processing services) which can be used to support water quality-drought-health community
Time: 13h00 UTC (14h00 CET, 09h00 EDT)
Telecon: 702-473-3463 passcode 567454#