Allocation of emergency response centres in response to pluvial flooding-prone demand points using integrated multiple layer perceptron and maximum coverage location problem models
- Publication Type:
- Journal Article
- International Journal of Disaster Risk Reduction, 2019, 38
- Issue Date:
|1-s2.0-S2212420918309348-main1.pdf||Published Version||6.62 MB|
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
© 2019 The increases in the frequency and intensity of rainfall events due to global climate change and the development of additional pavement, roads and water storage sites due to population growth have enhanced the probability of pluvial flooding (PF) in urban areas. The estimation of urban pluvial flood vulnerability and prompt emergency responses are crucial steps towards urban planning and risk mitigation. However, uncertainties exist in the optimal allocation of emergency response centres (ERCs). This study assessed the current situation of ERCs in terms of PF-prone demand points. In this study, fire and police stations, hospitals and military camps were defined as ERCs, and residential buildings, where people spend most of their time, were considered demand points. Our study area was Damansara City in Peninsular Malaysia, which is frequently affected by PF. We combined an optimised PF probability model with ideal location allocation methods on a geographic information system platform to construct the proposed model for achieving accurate ERC spatial planning. Firstly, PF-prone urban areas were identified using a recent machine learning multiple layer perceptron (MLP) model. Then, a Taguchi method was used to calibrate the MLP variables, namely, seed, momentum, learning rate, hidden layer attribute and class. Fourteen important PF contributing parameters were weighted on the basis of historical flood events. The predicted PF-prone areas were validated by comparing the predictions with the data from meteorological stations and observed inventory events. In addition, the current locations of ERCs were utilised in the location allocation model to assess the ideal time for providing essential services to elements at risk. Minimum impedance and maximum coverage location problem models were implemented to assess the current allocated location of ERCs and multiple scenarios. The coverage of existing ERCs was calculated, and their suitable and optimal locations were projected for vulnerable inhabitants with either redundant or late emergency response. Results showed that areas near downtown have high PF probability but are not covered by ERCs within the standard 5 km radius. The proposed model can conserve time, reduce cost and save human lives by ensuring that vulnerable people in the Damansara River basin are covered by the nearest ERC.
Please use this identifier to cite or link to this item: