Autor(es): Popal, Abuzar
Data: 2016
Identificador Persistente: http://hdl.handle.net/10362/17850
Origem: Repositório Institucional da UNL
Assunto(s): Disaster management; Damage Assessment; Natural disaster change detection; Multi-criteria analysis; ArcGIS; FME; Python
Dissertation submitted in partial fulfilment of the requirements for the degree of Master of Science in Geospatial Technologies
The past decade has witnessed many natural disasters hitting highly populated areas causing billions of dollars in damage as well as many human casualties. During natural disasters, when attaining ground measurements are limited, remote sensing and geographical information systems (GIS) are useful tools for in-depth analysis of the affected area. This report will introduce a new semi-automatic workflow in which the road network will be used to break up the area into “blocks” and then zonal statistics will be applied to detect change based on the created blocks rather than the conventional methods of change detection; pixel by pixel and object oriented. This hybrid approach will take advantage of the simplicity and ease of applying pixel change detection methods on fixed objects or “blocks” to assess for damage. The change detection analysis results can then be used to map and quantify damage caused by natural disasters using pre and post Landsat imagery of the affected area. Multi-Criteria Analysis is performed on the damage map, proximity to roads, proximity to waterbodies and building size to find the most suitable locations for temporary housing sites. The image differencing of NDWI mean produced the highest overall accuracy of 71.70% among eleven bands/indices and the multi-criteria analysis successfully selected fourteen temporary housing center sites from a possible 114. When time is of essence with limited resources and GIS expertise on the field, local authorities can greatly benefit from a rapid generalized analysis that will provide a “bird-eye view” of the affected area to efficiently and effectively allocate emergency efforts within a short time frame.
