|
|
SwissEx:EScience Related ProjectsFrom SwissExperimentEScience Related Projects
eScience LinksIt is vitally important that within SwissEx, we maintain an overview of other projects and technologies under development throughout the world and do not try to reinvent the wheel. the following list of links should be kept updated. Please email us if your project is missing.
Environmental Data NetworksThe following is a first impression, generated by only an hour or two of browsing and warrants further investigation: GEOSS is THE multinational network for earth observation data, backed by governments worldwide and pushing the standards system to make datasets from across the world accessible from a single web interface. The project is now in its 9th year and has plans to continue until 2015. So far, their list of conforming datasets is impressive, though their web interface is lacking way behind. The GeoPortal is their single access portal for GEOSS information, which uses the GeoNetwork system, so all the comments below apply. As a search database, it functions very well, although the solutions that appear usually redirect you to the originators website. The datasets that are available are viewable from a WebGIS interface, though this is slow and lacks way behind IKA's system. The system is so far very focused on geospatial information, though the in-situ sensors are due to be added this year. There seems to be only access to a single timestamp of information and no possibility for either stepping through timestamps or on-the-fly aggregation into the required time span. No real metadata source was found.
GeoNetwork is a geospatial database of links to datasets. It is used globally and has a good user base. As a search tool, the system is good, allowing the user to know what is available, but is it very geospatially biased and does not really allow the user to combine or visualise datasets. A small amount of visualisation is allowed (in the WSL instance, the image on the GeoNetwork website shows what looks to be a much more comprehensive system). Metadata is handled using RDF with Dublin Core ontology. This makes the metadata exchangable in a standardised format.
Neither the GEON nor GEOGrid portals were working at the time of writing this. NEON is a different type of project to most other data networks, where field sites have been created specifically for the NEON project across the US, with the aim of answering specific scientific questions. 'Standardised' measurement techniques will be used across the fieldsites to produce a high quality dataset, which will be available for public use. At the time of writing, these fieldsites are being constructed and no data systems or datasets exist online. CUAHSI contains a hydrologic-specific data network with a very well specified standards scheme. Timeseries data seems to be stored as WaterML, an OGC approved domain specific variant of SensorML. The system architecture seems to be similar to that used by SwissEx, where distributed database nodes are registered in a central metadata system, though this metadata/data system interacts with desktop-based data mining software instead of web-based software. As far as I can tell, this was a predecessor project to CUAHSI, from which some of the technologies and fieldsites have been integrated into the CUAHSI system. Snowtweets is more of a 'fun' crowdsourcing data project, aiming to get snow heights from Twitter. The technology is good, the data is representative.
Data made available as either a pre defined one week online query (returning results directly into the webpage) or an Excel download. No standards adhered to and no metadata provided. Uses the 'standard' BAFU webGIS system to display static (updated maybe once per year) geospatial data. Data only made public in static Hydrogeological maps. No public platform as yet. Small database with web query interface. Outputs are non-standard (image/html/csv) and no metadata is provided. Online search of document archives. A preliminary study has been made for a digital archive of geospatial data. A web interface for queries over 7 federated databases. This is a great service, although it is simplified by the small data volumes available. The results of the queries can be downloaded in non-standard formats or plotted using pre-defined plot types. The metadata is available, although not as comprehensive as the SwissEx system. Planetary skin is a very promising infrastructure given the collaborators and should use standard formats and processes. There is currently almost no public information on the progress of this infrastructure. The NASA global change master directory is a similar system to geonetwork, i.e. a database of data sources. As it also contains data sources from GEOSS and OGC, I would be surprised if it was not related or did not use the same standards. Fluxnet is a database of soil carbon, developed in collaboration with MS Research. The database is protected, so it is difficult to see what is behind it, but it probably runs using SenseWeb. The Global Seismographic Network is a global network of a specific type of seismograph. As such, the data is very homogeneous and does not pose the same challenges faced by SwissEx. The data is well controlled and has online access to data and metadata. The GSN is part of the GEOSS, hence is probably capable of standardised data exchange. A wiki based platform for aerosol data. The functionality in this platform is actually quite good although the user interface could do with some work. The WMS and WCS provide some good aggregation tools which GSN should have (specific months of the year etc.). The functionality for exchanging data is there, but it doesn't seem to adhere to W3C standards. Again, this platform is ahead of SwissEx on the data federation side. Part of GEOSS, hence data should be exchangable using standard formats. They seem to be focusing a lot on modelling, though no public interface is available yet. GIcat provides the catalogue services for dispirate catalogues in dispirate formats. It brings all of these WMS, WCS, NetCDF etc. sources into a single catalogue. It is used within GEOSS.
Other Environmental PlatformsOntologies/Standards
The SwissEx infrastructure has all the components in place for exchanging SensorML data and indeed has the basics of such an output in the GSN GML output. In order to provide an output in SensorML, we need to adopt and enforce a single ontology within the metadata wiki. Once this is done, there would be 2 options: 1. The GSN instance queries the wiki for the metadata whenever GSN is queried (the best option, but involves some development) 2. Add some extra components to the GSN XML sensor description file, which could be produced from the metadata (easier, but not as dynamic). The GSN databases themselves do not need to conform to the ontology as we can use the 'id' tags. It is vital for the future of SwissEx that we can provide a SensorML output SensorML would be easy to read with the current implementation of the GSN csv wrapper (with a tweak to ignore the footer). An extra wrapper would allow it to read 'binary' SensorML files, though the GSN description file would still be manually created, treating the SensorML file as a csv file with a large header. Automated creation of the sensor description file from the SensorML file would however be an advantage. NB: SensorML is fairly flexible and can take anything from sensor voltages to fully calulated and processed variables. They consider that everyone should save their raw sensor data to file (e.g. voltage) and provide the conversion to a measurement value as a SensorML Process Model. In theory, this is great and some projects such as PermaSense achieve this, but in many cases, this is not achievable due to communications bandwidth or cost. The lowest possible level of data processing should therefore be saved to any SensorML file. Providing process models prevents data from being duplicated (and hence the probability of hidden errors from previous processing), provides the widest range of possible applications for the data and provides a clear record of processing history.
Infrastructure
WorkflowsInteresting Technologies
ProcessingModelling Frameworks
PolicyData Visualisation |
