An open-source model for creating synthetic demographic data for applied spatial models
Topics:
Keywords: demographic, forecast, synthetic data, microsimulation, model
Abstract Type: Paper Abstract
Authors:
Nik Lomax, University of Leeds
,
,
,
,
,
,
,
,
,
Abstract
High resolution demographic data are an essential input for a huge range of applied spatial models. Information about the size and composition of populations and households, both now and forecast into the future, is needed to make informed decisions about the allocation of resources and development of policies which are well designed and appropriately targeted. Having access to individual level data is important for some models, where the heterogeneity of each ‘agent’ is needed to assess emergent phenomena and behaviour. However, such high-resolution data are rarely available, meaning that modellers either must make do with less detailed inputs, or create synthetic inputs which meet the necessary criteria. This paper provides an overview of how the synthetic data output from an open-source demographic model called SPENSER (Synthetic Population Estimation and Scenario Projection) have been used as input to a wide range of applied spatial models. SPENSER creates synthetic projections of individuals and households at a fine spatial resolution for the whole of the United Kingdom. To date SPENSER data have been used as input to a disease spread model, to transport Agent Based Models, to an urban development model, and have been augmented with energy efficiency data to create a new input to energy demand models. Because it is open source, SPENSER has been adapted to produce high resolution projections for Canada. This paper makes the case for having a dedicated model which produces demographic scenarios and projection outputs, freeing up other modelling teams to deal with their domain specific problem.
An open-source model for creating synthetic demographic data for applied spatial models
Category
Paper Abstract