Fellows Research Meetings

Marcelo Esteban Munoz (PhD student, University Hamburg, Germany

Using Spatial Microsimulation for the Estimation of Heat Demand

The presented work aims to: (1) compare different available R-libraries for the reweighting of survey data to Geo-referenced area; and (2) demonstrate the application of a synthetic population for the estimation of heat demand. The first section of this paper: (a) discusses the algorithms used to reweight a survey to small area benchmarks; and (b) presents a random-selection algorithm for the generation of a synthetic population. The second section of the paper presents an application of this synthetic population for the estimation of heat demand. In order to estimate heat demand we create a synthetic building stock representing the individual building characteristics and the households residing them.

For our analysis we benchmark the German micro-census (2010) to German "Gemeinden". These areas correspond to the European statistical units NUTS-3. Aggregated statistics for these areas are available from the last German census (2011). The micro-census used in this analysis is a freely available micro data set, it contains 3% of the original micro census survey (1% of the total population). This survey contains 23374 records of German residents and the buildings they live on.

The generation of a synthetic building stock based on demographic survey allows us to estimate heat demand at a low aggregation level without the need to process local data sets like the digital cadastre. This simplification on the computation of heat demand results in three advantages: (1) increases its transferability; (2) improves the computational time; (3) reduces the required amount of data for the computation of "heat cadastre" needed for the proper planing and dimensioning of heat supply infrastructure.

Last reviewed: 27 August, 2015

Past Events

International Conference on Robust Statistics, July 2017