Chair – Dr Seraphim Alvanides
Secretary – James Cheshire
Conference Coordinator – Dr Yi Gong
Prizes and Awards Coordinator – Dr Joana Barros
Postgraduate Representative – Tom Burgoine

The Transactions in GIS website can be found here:

http://www.wiley.com/bw/journal.asp?ref=1361-1682

The abstract of the dissertation appears below:

Marine ecosystems contribute significantly to global biodiversity, yet are under threat from both natural and human induced disturbances. Designing conservation strategies requires, as a first step, mapping these ecosystems. However, in situ mapping can be difficult due to the highly dynamic nature of the marine environment. Because wave exposure is known to play a major role in shaping marine ecosystems, it offers a potentially useful proxy variable for mapping them. However, in situ measurement of wave exposure is logistically difficult over large areas. Additionally, numerical wave models are limited to deep water, and are complex and time consuming to run. A Geographic Information System (GIS) based cartographic wave exposure model that measures the relative exposure of a site on the basis of its fetch length provides a simpler and faster alternative.

The academic literature describes a wide variety of cartographic exposure models, all designed for different purposes and study areas. Despite the fact that these models can be computationally intensive, their sensitivity to parameter setup has yet to be tested. Bringing these together within a common modelling framework where results can be standardised makes it possible to do such testing, and eventually to determine an appropriate model setup for a given study area and purpose. To that end, this study presents a generic modelling framework which integrates and extends existing relative wave exposure modelling tools within the ArcGIS programming environment (GREMO). It then uses this framework to assess the sensitivity of relative wave exposure models to key parameter settings, and outlines a conceptual framework for developing methods for parameter optimisation. The sensitivity analysis demonstrated that relative wave exposure model processing times and wave exposure estimates are influenced by how key model parameters are set in complex ways, and that this varies based on the spatial arrangement of waveblocking obstacles within a study area. This justifies further work to investigate sensitivity to unravel these complex interactions, and to develop tools to help determine optimal, or at least feasible, parameter settings for a given study area. Along with this, the conceptual framework for parameter optimisation outlined a range of tools which could be further developed to assist with this. For example, it may be that conditions across a given study area are not homogeneous, thus requiring division into sub-areas where different parameter settings can be applied. As a new integrated modelling environment for relative wave exposure, GREMO facilitates much needed sensitivity testings of these models as well as promoting the use of a mixed models / model comparison approach. Further development of GREMO that would make it even more useful include: (1) how to reduce its computational demand; (2) how to estimate settings for additional input parameters; (3) implementation of the conceptual framework for parameter optimisation; (4) further model verification, particularly of its spatial aspects; and (5) model validation.

We look forward to receiving your entries for next year!

Date: 28th Oct at 1700 GMT

Title: A dynamic social network model for disease transmission.

Speaker: Ling Bian (Buffalo)

Chair: Kirk Harland (Leeds)

The seminars are open to all. For details of how to join the e-seminar using the Marratech™ video conferencing environment, and further seminars in the same series, see: http://www.wun.ac.uk/ggisa/seminars.html

Dave Unwin (d.unwin@wun.ac.uk) WUN Global GISc Academy Coordinator
Steve Carver (s.j.carver@leeds.ac.uk) University of Leeds