Publication: Estimating population size using spatial analysis methods
Issued Date
2007-01-01
Resource Type
ISSN
18632351
Other identifier(s)
2-s2.0-84873476837
Rights
Mahidol University
Rights Holder(s)
SCOPUS
Bibliographic Citation
Lecture Notes in Geoinformation and Cartography. No.199039 (2007), 271-287
Suggested Citation
A. Pinto, V. Brown, K. W. Chan, I. F. Chavez, S. Chupraphawan, R. F. Grais, P. C. Lai, S. H. Mak, J. E. Rigby, P. Singhasivanon Estimating population size using spatial analysis methods. Lecture Notes in Geoinformation and Cartography. No.199039 (2007), 271-287. doi:10.1007/978-3-540-71318-0_20 Retrieved from: https://repository.li.mahidol.ac.th/handle/20.500.14594/24427
Research Projects
Organizational Units
Authors
Journal Issue
Thesis
Title
Estimating population size using spatial analysis methods
Abstract
© Springer-Verlag Berlin Heidelberg 2007. In population size is required within the first 24-72 hours to plan relief-related activities and target interventions. The estimation method should be easy to use by fieldworkers from various backgrounds, and minimize intrusion for the displaced population. Two methods have already been on trial: an adaptation of the Quadrat technique, and a newer T-Square technique. Here, we report the results of a field trial to test these alongside a newly adapted spatial interpolation approach. We compared the results with a population census of nine hamlets within the Tanowsri sub-district, Ratchaburi Province, Thailand. We mapped the study area to define the population for inclusion, as applications of this method would occur in closed settings. Before implementation, we simulated the spatial interpolation using geo-referenced positions of households in three hamlets. This procedure enabled us to establish some operational parameters to estimate population size, including the number of random points needed for the field test, the radius of the sample region and the dimension of the grid intersection on the interpolated surface. Each method was tested over the same area. The interpolation method seems to produce accurate results at 30m-grid spacing (at 104% of the census) with a worst-case estimate at 124%. These results are comparable to those of the Quadrat (92% to 108%) and T-Square (80%). The methods are proven feasible to apply, with a high acceptability among local workers we trained. The interpolation method seemed the easiest to conduct. The results were tested statistically where possible, though this was an experimental setting, and further trial is recommended