1. Integrating the Climate Science Modelling Language with geospatial software and services Dominic Lowe (d.lowe@rl.ac.uk), British Atmospheric Data Centred.lowe@rl.ac.uk A.Woolf, B.N. Lawrence, S. Pascoe

2. Context Developed and deployed a metadata schema with associated services for the Atmospheric/Oceanographic application domain. Summary of issues encountered Almost exactly the process INSPIRE is mandating.

3. INSPIRE Data Specifications: 34 themes including Atmospheric Conditions, Oceanography, Meteorology “standardisation of formats and nomenclatures so that the data sets can be combined seamlessly and without manual intervention” “Interoperability is the ability of two or more autonomous entities, like web services or applications, to communicate and co-operate with each other in a meaningful way. This interaction should not require special efforts by the data producer or consumer -be it human or machine.(www.inspire-geoportal.eu)”

4. Aim to engage with GIS standards... NetCDF, HDF, PP, GRIB, NASAAmes, ASCII, Databases... ? DataCentre WMS WCS WFS Metadata ISO TC211 for Geographic Information GML Application Schema CSML – Climate Science Modelling Language OGC

5. Climate Science Modelling Language UML model & GML Application Schema

6. CSML Feature Types AVHRR satellite imageryTwo-dimensional grid of data along a satellite ground-path SwathFeature numerical weather prediction model, ocean general circulation model Time-series of gridded parameter fieldsGridSeriesFeature gridded analysis fieldSingle time-snapshot of a gridded field.GridFeature weather radarBackscatter profiles along a look direction at fixed elevation but rotating in azimuth ScanningRadarFeature marine CTD measurements along a ship’s cruise track Series of profiles of unequal length along a trajectory in time and space RaggedSectionFeature shipborne ADCPSeries of profiles from positions along a trajectory in time and space. SectionFeature repeat daily balloon soundings of atmospheric temperature from the same location Time-series of unequal-length profiles, but on fixed vertical levels, at a fixed location RaggedProfileSeriesFeature vertical radar timeseries, thermistor chain timeseries Time-series of profiles on fixed vertical levels at a fixed location ProfileSeriesFeature wind sounding, XBT, CTD, radiosondeSingle ‘profile’ of some parameter along a vertical line in space. ProfileFeature 2m temperatures measured at weather stations across the UK at 0600z. Collection of distributed single datum measurements at a particular time PointCollectionFeature surface salinity along a ship’s cruise track; atmospheric aerosols along an aircraft’s flight path Measurement along a discrete path in time and space. TrajectoryFeature tidegauge, rainfall timeseriesTime-series of single datum measurements at a fixed location in space. PointSeriesFeature raingauge measurementSingle point measurement.PointFeature ExampleDescriptionFeature type

7. Problem solved? NetCDF, HDF, PP, GRIB, NASAAmes, ASCII, Databases... DataCentre WMS WCS WFS CSML Application Schema OGC ? ?

8. Infrastructure components

9. XML Handling Custom parser built for CSML Application Schema to read and write instance documents. Investment needed in generic GML Application Schema parsers CSML Documents... Python Classes class GridFeature:... toXML() fromXML() Provides API to CSML documents

10. Creating instance documents CSML Scanner PointSeries GridSeries Profile Section Radar Swath.... Features Which feature type? which files? which file type? which variables? which coordinate reference system...?

11. References to stored data CSML contains FileExtracts – pointers to data. Goes beyond current GML practice of inline storage. Essential for large volume data. Extensible for different formats Data access schema development in parallel with application schema development.

12. Harmonised data access layer Sensible handling of multiple data formats requires robust plug-in data interface architecture. CSML Code Prototype developed, but more generic/robust system needed for other application domains. Harmonised Data Interface layer CDMS NaPP Y PIL SQL Data APIs

13. Xlink in GML Application Schema Describes the nature of the remote resource in relation to the referring GML element (“local resource”) Describes the nature of the remote resource <gmlElement xlink:arcrole="urn:ogc:def:contentprovider:OGC:1.0:direct" xlink:href="file:///myfile.nc#temperature" xlink:role="urn:ogc:def:fileformat:OGC:1.0:netcdf" xlink:show="embed"/> See OGC Document: 07-083 Proposal to OGC – using XLINK to reference data. Requires more research. Beyond current GML practice for App. Schemas.

14. Vocabularies/Catalogues/Standard Terms Coordinate Reference Sytems in our application domain are not well catalogued. Hybrid Sigma Pressure Levels, etc. Catalogues and tools required to assist with identifying Coordinate Reference systems Ongoing work with standard vocabularies e.g. CF Standard Names. Assist greatly with axis identification. All these things aid automation and reduce human intervention. Using common vocabularies and catalogues is critical for interoperability.

15. Summary Followed rules for application schema development Problem only partly solved! Many implementation issues to consider Opportunities for development of re-useable component INSPIRE is coming!