Human Systems Modelling

School of Geography, University of Leeds

Principal Investigators

Stan Openshaw
Phil Rees

Dates

1st November 1994 - 31st October 1996

Grant

EPSRC

Summary

The project involved the conversion of existing geographical modelling software, written to work on sequential processors, to run on a high performance, parallel processor computer.

Amongst model programs successfully ported to this new environment were code for spatial interaction models (SIMs) and code for multi-regional population projection for European Community regions (ECPOP). In the latter case the aim was to be able to run the model several thousands of times using input parameters sampled randomly from normal distributions around scenario trajectories. The ECPOP model was converted to run a projection on each processor and probabilistic forecasts for NUTS Level 1 regions in 13 European Union countries.

Main results

• The production of high performance versions of useful geographical programs.
• The ability to run these programs in ways we couldn't before. This enhanses our knowledge of, for example, European populations in the future.

For further information see...

Turton, I. and Rees, P.H. (1998) 'Investigation of the effects of input uncertainty on population forecasting'. In 'Proceedings of the Third International Conference on Geocomputation, University of Bristol, 17-19 September 1998' CD ROM. ISBN 0-9533477-0-2.