Project

General

Profile

Actions

Geographic Components » History » Revision 9

« Previous | Revision 9/30 (diff) | Next »
Pablo Sastre Olmos, 08/03/2007 04:09 PM


/\ Under construction... Platform components for taxonomists dealing with geographic visualisation, geocoding and predictive modelling tools


Geographic Components

Introduction

The general aim of this activity is to provide the resources and applications able to publish, visualise, and analyze the distributional information associated with taxonomic information. Taxonomists require an easy and freely available application allowing to display and/or publishes the distribution information directly from simple data sources. However, as present distributional data are far from accurate we urgently also need tools able to:

  • i) examine the degree of completeness of this information,

  • ii) discriminate well surveyed localities from those do not have reliable inventories, and

  • iii) locate the localities in which is necessary to carry on additional surveys in order to recover the environmental and spatial variation of the area. The activity is collaboratively carried out by all the partners.

Objectives

To provide a generic and open source software solution for the Internet Platform for Cybertaxonomy and to use this as the base for specific tools to:

  • provide output for printed and on-line taxonomic publications

  • visualize distributional information

  • statistically analyse distributional information with regard to completeness of surveys

  • identify gaps to prioritise surveys in order to obtain an unbiased set of data for environmental analysis.

Deliverables 5.35 Predictive distribution modelling report and 5.38 Gap analysis in local inventories report

Taxonomists have to continue doing what they have done during the last three hundred years: to describe the variety of life organisms and their location. Although this colossal task is important by itself its relevance is higher now due to current need of reliable biodiversity data. In the attached report we review the available scientific information on the possibilities and usefulness of the compiled species distribution data for basic and applied purposes, two of the deliverables of EDIT Work Package 5.4 “Geographical platform components” (deliverables 5.35 Predictive distribution modelling report and 5.38 -Gap analysis in local inventories report-).

The main conclusions raised by this report are that (i) our current species distribution information is biased and insufficient for most taxonomic groups, and (ii) modelling methods can not provide reliable and useful distribution predictions if they are based in these biased of data. Therefore, we identify as a key priority for bioinformatics the development of tools to: i) examine the degree of completeness of distributional information, ii) discriminate well surveyed localities from those that do not have reliable inventories, and iii) identify sets of areas where to carry out additional surveys, in order to increase the level of coverage of the environmental and spatial variation of a given region. We encourage that these tools are made freely available and easy to use to universalize their application. A list of the available software is attached at the end of the report.

Our purpose is to use this report as a kick-off for a debate between the people interested in the utility of current taxonomic and distributional data. Such debate will be carried out in a forthcoming e-conference (EDIT deliverables 5.32 and 5.33), where the participation of taxonomists, conservationists and bioinformaticians are welcome. EDIT aims to provide resources for taxonomists and the development of these tools would be an opportunity to increase the correct use of the biological information, promoting also the participation of taxonomists in the use of their (our) own data. The e-conference is an opportunity to contrast opinions and identify key issues needed for the development of effective bioinformatic tools, such as the ones we suggest.

To download the report please click here: http://wp5.e-taxonomy.eu/blog/files_edit_wp5/2007-07-26_D5.35_&_D5.38.doc

GIS data downloads

Download of Geographic Information Systems (GIS) data layers:

http://edit.csic.es/web/page1/page1.html

Description of the GIS layers of surface units for the analysis of spatial completeness (comments are very welcome):

http://edit.csic.es/web/docs/EDIT_GIS_layers_Surface_units.htm

Web application - Map viewer prototype

After some time evaluating the available open-source software and the possibilities it could offer to the EDIT Geoplatform we decided to start working with Mapbuilder, a JavaScript library that provides a client-side solution for dynamically generating web pages from XML (such as OpenGIS Consortium documents) as well as the OGC Requests (GetMap, GetFeatureInfo, GetFeature...) necessary to view and query the geo-data.

Mapbuilder version used is 1.0.1. It works with most modern browsers (Firefox 1.0+, Internet Explorer 6.0+, Mozilla 1.3+, Navigator 6+) but we are not sure about other web browsers. You must have javascript enabled! please check it in your web browser.

The geo-data is stored in PostGIS, a database with a consolidated spatial extension able to make spatial queries (intersect, point-in-polygon, calculate distances, centroids), reproject data, etc. and "usual" queries, including statistical functions. On the next steps we will take profit of both possibilities to statistically analyze geo-referenced data in order to locate well surveyed localities...

The link between data (PostGIS) and web-application (Mapbuilder) is done through GeoServer. It takes the requests and sends a response: a beautiful image (after applying styles).

The URL of the web application is: http://edit.csic.es:8080/edit_geo/prototype/edit.html

This web application is not definitive at all. In fact, it lacks of two main issues:

  • Complete interactivity: user doesn't insert data and the analysis (point-in-polygon) to get biological information is not done "on-the-flight". We will have to work with programming (PHP probably) to send the parameters to a spatial SQL function (Contains) to be executed in PostGIS.

  • Complete interoperatibility:

  • 1) legend images are static. We will have to manage to interactively generate a legend according to the data the user inserts. MapBuilder doesn't provide the possibility to interactively construct legends. MapServer can be a good solution.

  • 2) legend images are not OGC compliant. It means that they cannot be viewed through any other OGC compliant web-application. For example, you can try to view our geoserver data through the Intergraph WMS Viewer (http://www.wmsviewer.com/main.asp): If in “Edit Servers” you insert http://edit.csic.es:8080/geoserver/wms you can check all the WMS layers we are serving. Using the adequate style for each layer, you can see the data as if you were in our application, but you cannot see the legends (left side of the page, next to Layers). If instead of our server you insert, for examle http://devgeo.cciw.ca/cgi-bin/mapserv/windatlas? and check any of the layers, you will can see a Legend. You can check also to get the legend inserting, as an URL to the browser, the following: http://devgeo.cciw.ca/cgi-bin/mapserv/windatlas?SERVICE=WMS&VERSION=1.1.1&REQUEST=GetLegendGraphic&LAYER=roughnesslength&FORMAT=image%2Fgif

Demo: Spatial completeness of biodiversity data

The objective of the demo is to make an idea of one of the functions of the future application: the analysis of spatial completeness.

The URL of the web application where you can see the demo is: http://edit.csic.es:8080/edit_geo/prototype/edit.html

In this demo, it’s supposed that the user has already:

  • 1.- selected the extent for the analysis (Iberian Peninsula is used as example)

  • 2.- submitted his file of point sample data (Jorge M. Lobo's data on Iberian Scarabaeidae are used as example)

  • 3.- selected a taxonomic level from those included in his data file (genus is used as example)

  • 4.- selected a GIS layer of surface units (UTM squares of 2500 sq.km. are used as example)

  • 5.- choosed or clicked on “perform anaysis of spatial completeness”

Then, three different maps are displayed:

  • Map of sampling effort (number of records in each square)

  • Map of taxonomic richness (number of genera in each square)

  • Map of inventory uncertainty. Inventory uncertainty in each surface unit is based not only on the number of taxa (S) and the number of records (N), but also on the relative frequency of the taxa (FrSp = Fr1, ...., FrS). In this example, inventory uncertainty (IU) is measured as the probability of missing some of the taxa:

IU = 1 - ∏Sp (1-(1-FrSp)N)

The map of inventory uncertainty indicates the “red” surface units where is necessary to carry on additional surveys in order to recover the spatial variation of the area, or where data on absences should be recorded.

A review of the available scientific information on the possibilities and usefulness of the compiled species distribution data for basic and applied purposes is available for download at http://wp5.e-taxonomy.eu/blog/files_edit_wp5/2007-07-26_D5.35_&_D5.38.doc.

Free GIS software links

Links to main Geographic Information Systems (GIS) free software:

Updated by Pablo Sastre Olmos over 16 years ago · 9 revisions