CONTACT US: Contact info
|
Citation |
Sakieh Y. Int. J. Disaster Risk Reduct. 2017; 25: 125-136. |
|
Copyright |
(Copyright © 2017, Elsevier Publishing) |
|
DOI |
|
PMID |
unavailable |
|
Abstract |
This study adopts an integrated spatial-statistical approach to explicitly analyze the relationships between landscape structure (composition and configuration) of human settlements and their exposure to different levels of flood hazard. In this regard, the flood hazard surface of the research location was firstly produced using the MCE (Multi Criteria Evaluation) method. Then, 2013 urban land-use map of the study area was overlaid with flood hazard layer and average hazard values were extracted from hierarchical buffer rings (500m, 1000m, 1500m and 2000m) around urban and rural centers (dependent variables). In addition, landscape indices of the corresponding centers were also quantified to serve as independent variables. Having the data-set generated, scatterplots between different metrics and average flood hazard values across multiple buffer rings were drawn to study whether there are distance-dependent relationships between morphology of human settlements and their exposure to different levels of flood hazard. The regression models were then developed for each buffer area and those landscape metrics with better explanatory power were selected based on the step-wise approach. According to the results, majority of the metrics clearly demonstrated distance-dependent behaviors and developed regression models also confirmed this matter. Policy implications derived from this study indicated more connected and aggregated patterns of human settlements in rural areas can significantly increase resistance of such environments against natural hazards. Landscape ecology approach is a potential framework for studying urban sustainability at regional scales and an innovative paradigm for environmental hazard management and informed decision making. Language: en |
|
Keywords |
Flood hazard; Iran; Landscape ecology; Landscape metrics; Linear regression |