How to join the environmental data revolution

Intro

This page accompanies a talk by JonBlower, given as the ESSC seminar on 22nd February 2007. All the materials from this presentation are held on this page so you can examine them at your leisure.

Presentation slides: Click here.

The talk presents a number of simple ways to share and explore environmental data. All these technologies are well within the grasp of scientists, irrespective of their preferred programming language or operating system. The key is to encourage scientists to share more data, rather than static plots of data. The talk mainly concerned data formats that can very easily be visualized and "mashed up" to examine relationships between data from different sources.

(UPDATE: An article related to this Wiki page has been published on the NewScientistTech pages: http://www.newscientisttech.com/article/dn11773-virtual-earths-let-researchers-mash-up-data-.html)

User-friendly technologies

GeoRSS

http://www.georss.org

Very simple way to geolocate information. Also consider Atom (a better standard in many ways but currently less well-supported by software). Supports time-stamping and hence animations. GeoRSS can be visualized very simply in Worldkit (http://www.worldkit.org).

This is a simple GeoRSS file containing the location of some ocean observations.

KML (Keyhole Markup Language)

http://earth.google.com/kml/

The format of Google Earth but also being adopted by other geographic information systems such as NASA World Wind. Good balance between understandability and richness of data representation. Follow the KML tutorial for more information. KMZ is simply zipped-up KML, sometimes with images and icons in the zip file.

KML / KMZ examples

• UK Albedo (simple GroundOverlay) - Mel Sandells
• Eruption of Cleveland Volcano - thanks to Bill Brody (Arctic Region Supercomputing Centre), Peter Webley (Alaska Volcano Observatory/ARSC), John Bailey (AVO/ARSC)

KML Mashup 1: Ocean science

This mashup shows the results of an ocean model together with the data that was assimilated into it, and is a first step towards a more sophisticated system that will allow direct comparison of model and observational data in Google Earth for diagnostic purposes.

• Global sea surface temperature from 1/4 degree DRAKKAR model (NEMO code), generated through Godiva2 website.
• Nearly 12,000 ocean observation sites (Hadobs ENSEMBLES data), converted by Alastair Gemmell.
• DAMOCLES data (Arctic sea-ice and climate change): http://www.damocles-eu.org/. Note that this links to an external site and so its contents may change.

KML Mashup 2: Hurricane Katrina

This mashup shows the path of Hurricane Katrina (August 2005), revealed through the atmospheric vorticity field. By incorporating the sea surface temperature data it is possible to see the waters of the Gulf of Mexico cooling as the hurricane passes.

• Tropical storm tracks for 15-08-2005 to 09-09-2005, as identified by TRACK software (Kevin Hodges and Lizzie Froude, ESSC). Converted to KML using this Python script.
• Vorticity field for same time period from ECMWF analysis, generated through Godiva2 website.
• Sea surface temperature data for the same period from the UK Met Office's global FOAM model, generated through Godiva2 website.

KML generation software

• This Python script and associated icons converts plain-text output from the TRACK software into KML. The zip file contains some example TRACK output files.

Slightly more complex

Web Map Service (WMS)

Standard means for serving map imagery on the Web. Each map is defined by a long Web address (e.g. this). ReSC has developed a fast WMS implementation primarily for NetCDF data called ncWMS.

Scalable Vector Graphics (SVG)

http://www.w3.org/Graphics/SVG/

Describes images such as X-Y data plots in XML and can be interpreted by many modern browsers such as Firefox. SVG can produce interactive plots, allowing for pan and zoom and revealing of data values on mouse-over.

The next level

The above technologies are user-friendly means for sharing and exploring data. Moving to more sophisticated technologies commonly requires technical help.

Richer markup languages

Geography Markup Language (GML) is rich and extensible and can be "profiled" to suit particular community needs. Climate Science Modelling Language (CSML) is a GML "application schema" for atmospheric and oceanographic data.

Web Services

The following (in common with WMS) are specifications of the Open Geospatial Consortium:

• Web Coverage Service (WCS): Serves "coverage" data (commonly gridded data)
• Web Feature Service (WFS): Serves geographic features (e.g. roads, rivers, weather fronts, ship tracks)

Key players in this field are the NERC Data Grid, who are developing a secure, standards-based data infrastructure for NERC data.

-- JonBlower - 21 Feb 2007