Population disaggregation to capture short trips - Vishakhapatnam, India

Abstract Spatially disaggregate models are required to account for non-motorized trips (walk and bicycle) as inter-zonal trips. However, models are developed at census defined ward level with inter-zonal distance being relatively more than the average walking and bicycling distance in cities. As a result, a large proportion of non-motorized transport (NMT) trips are counted as intra-zonal trips and therefore NMT trips are not included in mode choice and trip assignment steps. One of the major limitations in developing spatially disaggregate models is availability of data at finer scales. In Visakhapatnam 3062 enumeration blocks (EB) were demarcated for conducting census survey in 2011 that contain demographic data. Of these, spatial information is available for only 49% of the total EBs. This study aims to disaggregate city population at spatially finer level than wards so as to account for maximum NMT trips as inter-zonal trips. In the study, the spatially finer zones are termed as non-motorized traffic analysis zones (NMT-TAZ). For the city area with missing data, 634 NMT-TAZ boundaries are defined based on the Google Earth imageries (2010), road profile (2011), ward boundaries (2011) and natural boundaries. To disaggregate population, four methods are applied and tested on the available dataset – equal weighing technique, zonal classification based weighing technique, global regression and regional regression. Based on the residual statistics, it is concluded that regional regression method outperforms the other three methods at both NMT-TAZ and ward level. Regional regression provides more accurate estimates in high population density wards as compared to low population density wards. The model shows that local and collector road density, number of residential units and available built up area in NMT-TAZ have significant and positive impact on population count. The population at NMT-TAZ level for the city area with missing data is estimated using regional regression resulting in under prediction of total population of the region by 10% that is adjusted with respect to the error obtained at ward level. The use of small size NMT-TAZ helps in accounting for 83% of the total walk trips as inter-zonal trips as compared to only 17% when ward boundaries are used as TAZ. The method is useful in estimating population at spatially finer scale in case of limited data availability.

[1]  Catherine Linard,et al.  Disaggregating Census Data for Population Mapping Using Random Forests with Remotely-Sensed and Ancillary Data , 2015, PloS one.

[2]  Suely da Penha Sanches,et al.  Incorporating Nonmotorized Modes in a Mode Choice Model , 2002 .

[3]  Chris Tofallis,et al.  Erratum: A better measure of relative prediction accuracy for model selection and model estimation , 2015, J. Oper. Res. Soc..

[4]  C. Lo.  Population Estimation Using Geographically Weighted Regression , 2008 .

[5]  B. Bhatta.  Analysis of Urban Growth and Sprawl from Remote Sensing Data , 2010 .

[6]  Ronald W Eash,et al.  Destination and Mode Choice Models for Nonmotorized Travel , 1999 .

[7]  C. Willmott,et al.  Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance , 2005 .

[8]  Anthony D. May,et al.  Modelling requirements for local transport plans: An assessment of English experience , 2006 .

[9]  José Manuel Viegas,et al.  A traffic analysis zone definition: a new methodology and algorithm , 2009 .

[10]  Mitchel Langford,et al.  An Evaluation of Small Area Population Estimation Techniques Using Open Access Ancillary Data , 2013 .

[11]  E. Aktas,et al.  An Integrated Transportation Decision Support System for Transportation Policy Decisions: The Case of Turkey , 2007 .

[12]  Xiaomin Qiu,et al.  Population Estimation Methods in GIS and Remote Sensing: A Review , 2005 .

[13]  Cynthia A. Brewer,et al.  Dasymetric Mapping and Areal Interpolation: Implementation and Evaluation , 2001 .

[14]  Bor-Wen Tsai,et al.  Multi-layer multi-class dasymetric mapping to estimate population distribution. , 2010, The Science of the total environment.

[15]  Zhiyong Hu,et al.  Modeling urban growth in Atlanta using logistic regression , 2007, Comput. Environ. Urban Syst..

[16]  West Bengal,et al.  Primary census abstract , 1976 .

[17]  Michael Batty,et al.  Cities and complexity - understanding cities with cellular automata, agent-based models, and fractals , 2007 .

[18]  Eva Ericsson,et al.  Framing the role of Decision Support in the case of Stockholm Congestion Charging Trial , 2009 .

[19]  Zhengdong Huang,et al.  A Doubly Weighted Approach to Urban Data Disaggregation in GIS: A Case Study of Wuhan, China , 2007, Trans. GIS.

[20]  Andrea E. Gaughan,et al.  Dasymetric modeling: A hybrid approach using land cover and tax parcel data for mapping population in Alachua County, Florida , 2016 .

[21]  B. Blijie.  The impact of accessibility on residential choice - empirical results of a discrete choice model , 2005 .

[22]  Rafael Lozano,et al.  Modeling causes of death: an integrated approach using CODEm , 2012, Population Health Metrics.

[23]  A. El-geneidy,et al.  Validating walkability indices: How do different households respond to the walkability of their neighborhood? , 2011 .

[24]  Geetam Tiwari,et al.  How the present would have looked like? Impact of non-motorized transport and public transport infrastructure on travel behavior, energy consumption and CO2 emissions – Delhi, Pune and Patna , 2016 .

[25]  W. Limp,et al.  Remodeling census population with spatial information from LandSat TM imagery , 1997 .

[26]  B. Bajat,et al.  Spatial modelling of population concentration using geographically weighted regression method , 2011 .

[27]  Geetam Tiwari,et al.  Impact of public transport and non-motorized transport infrastructure on travel mode shares, energy, emissions and safety: Case of Indian cities , 2016 .

[28]  Andrew R. Maroko,et al.  Using Geographic Information Science to Estimate Vulnerable Urban Populations for Flood Hazard and Risk Assessment in New York City , 2009 .

[29]  Role of travel demand models in appraisal and policy-making , 1998 .

[30]  Daniel A. Rodriguez,et al.  The relationship between non-motorized mode choice and the local physical environment , 2004 .

[31]  F. J. Gallego,et al.  Disaggregating population density of the European Union with CORINE land cover , 2011, Int. J. Geogr. Inf. Sci..

[32]  Catherine Linard,et al.  Mapping populations at risk: improving spatial demographic data for infectious disease modeling and metric derivation , 2012, Population Health Metrics.

[33]  Angela Hull,et al.  Developing a set of decision-support tools for sustainable urban transport in the UK , 2008 .

[34]  Frank Canters,et al.  Incorporating spatial non-stationarity to improve dasymetric mapping of population , 2015 .

[35]  Paul A. Zandbergen,et al.  Comparison of Dasymetric Mapping Techniques for Small-Area Population Estimates , 2010 .

[36]  Mitchel Langford,et al.  Obtaining population estimates in non-census reporting zones: An evaluation of the 3-class dasymetric method , 2006, Comput. Environ. Urban Syst..

[37]  K. Kockelman,et al.  Models for Anticipating Non-motorized Travel Choices, and Role of the Built Environment , 2014 .

[38]  Carlo G. Prato,et al.  Walking, cycling and the urban form: a Heckman selection model of active travel mode and distance by young adolescents , 2016 .