eResearch

You are here

Prototype: Geospatial Data Sharing

Datasets

The WildNet database contains 3.5 million records of wildlife sightings and listings of around 20,000 species such as plants, mammals, birds, reptiles, amphibians, freshwater fish, marine cartilaginous fish and butterflies in Queensland. Species are classified by a taxonomy including multiple levels, kingdom names, class names, family names, scientific names, and common names. WildNet maintains a large store of ecological data, which depends heavily on many other services, and is itself a service to many other applications. The following datasets are included in this case study:

  • Snake sightings data in southeast Queensland region, provided by the Queensland EPA;
  • Bird sightings data along Queensland coastline, provided by the Queensland EPA;
  • Bird taxonomy data, extracted from Australia Museum via BioMaps
  • Weather data, downloaded from Australia Bureau of Meteorology.

Features

The prototype has two main functionalities:

  • Selective catalogue search: The catalogue service supports two kinds of search: keyword-based search and location-based search. For keyword-based search, given a keyword, only relevant data sources are returned from the catalogue search. Figure 1 shows the keyword input dialogue. When the input keyword is "snake", only the snake source is returned, and when the keyword is "bird", only bird data sources are returned, as shown in Figure 2 and Figure 3 respectively. For location-based search, we can specify an area that we are interested in, and only the data sources whose data coverages overlap with the polygon selection are returned, as illustrated in Figure 4.
  • Data sharing based on WFS: Figure 5 to Figure 7 show the data sharing functionality of the prototype. Given a common name, for example, we can get more information about birds with the name, e.g., the taxonomy information, the climate profile, while such information is usually distributed across different data sources.

Screenshots


Figure 1: Keyword input


Figure 2: Catalogue search result for keyword "snake"


Figure 3: Catalogue search result for keyword "bird"


Figure 4: The catalogue search result when a polygon selection is created


Figure 5: The dialogue for bird name input


Figure 6: The distribution of "Australian White Ibis"


Figure 7: The climate profile for "Australian White Ibis"

Implementation

The implementation is based on Open Geospatial Consortium (OGC) standards, as data involved in our case study has a geographical or spatial nature. The implementation architecture is shown below:


Implementation Architecture

In the implementation, we use two machines; A and B. Machine A stores snake sightings data, bird taxonomy data, and weather data; machine B stores bird sightings data. A new data source always publishes its services to the catalogue server (step (1)). During a search (step (2)), the catalogue server is first contacted for related data sources (step (3)); then the search request is dispatched to these data sources (step (4)); and finally data are accessed (step (5)).