GEOG8731 GIS Modelling GE

2016

4.5

1 x 2-hour lecture weekly
1 x 2-hour workshop weekly
1 x 4-hour project work weekly

1 of GEOG3020, GEOG3731, GEOG7015, GEOG7711, WARM8731 has been successfully completed

It is assumed students will have skills and knowledge in Geographical Information Systems such as can be found in GEOG8700 Geographical Information Systems GE or GEOG8008 Geographical Information Systems.

GIS is a decision support tool. GIS models are useful instruments for the analysis of complex systems, predicting outcomes, prioritising management strategies and quantifying risks. GIS modelling involves solving spatially-explicit problems related to complex systems in the natural or built environments. The topic first introduces students to conventional modelling techniques such as generalised linear models, generalised additive models, classification trees, regression trees, artificial neural networks and multivariate adaptive regression splines. However, the topic then applies these approaches to spatially-explicit case studies, solving human, ecological and environmental problems. This requires an understanding of scale issues, spatial autocorrelation and non-stationarity. Spatially-explicit techniques including geographically weighted regression are also explored.

The aim of this topic is to provide students with a good understanding of spatially-explicit static modelling approaches, together with associated scale, spatial autocorrelation and non-stationarity issues, as they apply to human, ecological and environmental problems.

At the completion of the topic, students are expected to be able to:

1. Describe a range of linear and non-linear aspatial regression modelling and data mining methods
2. Identify the risks and describe solutions to spatial autocorrelation and scale issues
3. Demonstrate an application of geographically weighted regression and describe its strengths and limitations
4. Demonstrate three predictive spatial modelling techniques and test their predictive power using independent data and comment on their respective strengths and limitations