A group around Landcare, NIWA, GNS, HillTop Software and Horizons Regional council worked the last 2 years to develop an interoperable web-based data access and sharing standard. Working group formed as results from Hydrological Society Annual Conference Data Access Workshops. Subsequent meetings and workshops at Horizons regional Council office in Palmerston North even featured visits and input from Dr David Maidment as well as from delegates from BOM and Kisters, Australia. Further workshops were held at GNS in Wairakei and NIWA offices in Auckland and Wellington.
The NEMS-endorsed Environmental Observation Data Profile (EODP) is based on proven open standards of the Open Geospatial Consortium (OGC) and comprises a profile of WFS that allows the discovery of timeseries data associated with monitoring stations. This forms part of a larger project to enhance the interoperability of Environmental Observations throughout New Zealand.
Organisations such as NIWA, Landcare, GNS and Regional councils maintain a range of monitoring stations making environmental observations. These include Climate, Hydrology, Air Quality, Soils, Water Quality, and Marine data.
We wish to make this data available for research, analysis and reporting through open data protocols such as the Sensor Observation Service (SOS), Web Feature Service (WFS), Web Coverage Service (WCS) and Web Mapping Service (WMS). We also want to publish the existance of such data sets and services through online metadata catalogues (CSW).
In order to make use such data services users must first be able to discover their existence and then accurately select the subset of data that is relevant to their needs. The EOPD is focussed on this selection process in essence addressing the Where, What and When questions.
WFS provides a rich filtering capability, but typical WFS sources use a flat property/value schema that makes it difficult to express the many to one relationship of the timeseries measurements taken at a station. The EODP is an Application Schema for WFS that encodes the station metadata as a nesting of SF_SpatialSamplingFeature and OM_Observations using the language of Observations and Measurements to describe the characteristics of the timeseries.
The profile is extended by the use of an external Vocabulary service used to dereference measurement identifiers into familiar timeseries names such as rainfall or temperature.
Alistair Ritchie from Landcare Research did the main formal write-up and published the documentation on GitHub: https://github.com/EODP-NZ/eodp-dev
A implementation testbed is currently underway in joint collaboration with NIWA on the NGMP database at GNS and the Climate Database (CLIDB) at NIWA.
http://portal.smart-project.info/sos-smart/service?service=SOS&version=2.0.0&request=GetDataAvailability&observedProperty=http://vocab.smart-project.info/ngmp/phenomenon/1679
http://portal.smart-project.info/gs-smart/wfs?request=GetFeature&service=WFS&typename=sams:SF_SpatialSamplingFeature
Showing posts with label Horizons. Show all posts
Showing posts with label Horizons. Show all posts
Tuesday, 10 February 2015
Friday, 16 November 2012
Starting to play with spatial-temporal data
Thinking the big picture is a different thing to actually implement it :-) Well, who doesn't know that. Having some hydrological time-series (groundwater levels) in place in a sensor observation service (SOS), the hydrogeological all-in-one-wonder-portal is going to get a glance of the next level :-p
The last weeks I started to play around with the R environment for statistical computing and visualization (The R Project). 52°North developed a neat R toolkit to access and digest SOS time-series - sos4R.
It is well documented and pretty easy to connect to a SOS server, and query observations. So for the fun of it and to demonstrate the general feasibilty, I quickly queried the groundwater levels of the Horowhenua area in New Zealand, where I got some sample data (courtesy by the regional council).
With the R sos4R, fields and akima packages from the CRAN R packages archive I (quite coarsely) interpolated the groundwater surfaces for the years 1991-2009 and put the images together as an animated gif (meters above mean sea level over time).
Discussion
I am aware of the total uselessness of this particualr way presenting :-) No years, the scale changes slightly, and the exact spatial extent and north orientiation are not reliable :-p
Nevertheless, for just playing around, this was a motivating simple first shot to easily visualise changes over time.
I would like to play around with the gstat and spacetime R package, integrate a more sophisticated script as a 52°North WPS process and have those things happening automagically in the interwebz.
The last weeks I started to play around with the R environment for statistical computing and visualization (The R Project). 52°North developed a neat R toolkit to access and digest SOS time-series - sos4R.
It is well documented and pretty easy to connect to a SOS server, and query observations. So for the fun of it and to demonstrate the general feasibilty, I quickly queried the groundwater levels of the Horowhenua area in New Zealand, where I got some sample data (courtesy by the regional council).
With the R sos4R, fields and akima packages from the CRAN R packages archive I (quite coarsely) interpolated the groundwater surfaces for the years 1991-2009 and put the images together as an animated gif (meters above mean sea level over time).
I am aware of the total uselessness of this particualr way presenting :-) No years, the scale changes slightly, and the exact spatial extent and north orientiation are not reliable :-p
Nevertheless, for just playing around, this was a motivating simple first shot to easily visualise changes over time.
I would like to play around with the gstat and spacetime R package, integrate a more sophisticated script as a 52°North WPS process and have those things happening automagically in the interwebz.
Subscribe to:
Posts (Atom)