Spatial-Temporal Modelling

Definition of Spatial-Temporal Modelling

Spatial-temporal modelling relates to problems where we want to analyse and predict how something varies over space and/or time. Such problems can exist at widely different spatial and temporal scales, ranging from detailed medical imaging data to Geographic Information System (GIS) data, and from ‘real-time’ medical and operational time frames to decades long ecosystem monitoring time frames.

Who works in Spatial-Temporal Modelling

Students: Daniel Kennedy, Marcela Cespedes, Susanna Cramb, Cathy Hargrave, Ben Fitzpatrick, Julie Vercelloni, Earl Duncan
Post-docs: Sam Clifford, Paul Wu,
Lecturers: Nicole White
Professors: Kerrie Mengersen
Past BRAG Members: Dr Su Kang, Dr Clair Alston

What Expertise do we have?

  • Identify the right scale to model spatial-temporal data
  • Manipulating a range of spatial, temporal and longitudinal data types using a variety of statistical, mathematical and custom software
  • Identify, quantify and predict variations from one observation in space-time to another either directly or as part of a hierarchy
  • Identify the important variables that affect outcomes of interest (variable selection)
  • Identify unusual spatial-temporal trends
  • Identify, quantify and predict variations between baseline performance and performance when the system is subject to time-varying inputs or disturbances
  • Identify the nature of spatial dependence in data and model it accordingly
  • Visualise spatial-temporal data and results of analysis
  • PyMCMC is open source python software that has many of the major Bayesian algorithms coded in a highly optimized form. It contains a standard library, consisting of linear models, generalised linear models, mixture models and mixed models. The user can also structure their own models within python using the algorithms provided.
  • Software platforms: R, WinBUGS, Matlab, PyMCMC

Case studies: Concrete examples of what we have done

Environment:

  • Dr Clifford- hierarchical modelling of school air quality in Brisbane
  • Dr Wu – modelling environmental windows and the effect of dredging on marine species
  • Mr Fitzpatrick – modelling spatial distribution of soil carbon in agricultural land
  • Ms Vercelloni – semi-parametric hierarchical modelling of large-scale coral cover
  • Dr Kang- in collaboration with the Australian Institute of Marine Science, model-based adaptive monitoring methods are being developed for spatio-temporal coral reef data. Modelling allows updating of existing information using past information, so new decisions incorporate lessons learnt from past observations.
  • Dr Alston – Analysis of Landsat images in relation to fireant habitation to assist in planning the surveillance strategy for upcoming seasons. The habitat model was constructed using a combination of CART, univariate and multivariate mixture models. This consulting was done in collaboration with Biosecurity Queensland (Department of Agriculture, Fisheries and Forestry)

Medical:

  • Cathy Hargrave – tracking changes in medical images over time and identifying critical points to optimise treatment
  • Earl Duncan – modelling medical service utilisation in different geographic areas over time
  • Jannah Baker – modelling hospital admissions in different postcodes and joint risk factors for type-2 diabetes

Epidemiology:

  • Susanna Cramb – cancer incidence rates and their space-time distributions
  • Dr. Su Kang- mapping disease presence and its spread over time

Business and Operations:

  • Dr. Paul Wu – dynamic network modelling of airports to provide real-time operational decision support

Links to other capabilities:

  • Decision support systems, applications, visualisations, computationSpatio-Temporal models,

Contact

Sam Clifford: samuel.clifford@qut.edu.au
Paul Wu: p.wu@qut.edu.au

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s