Application Of Time Series Remote Sensing Images In Impervious Surface Monitoring To Reduce The Risk Of Urban Flood Disasters

The capital region of Ghana,Accra,is the most urbanized in infrastructure,population and capacity.It has experienced perennial flooding in most parts over the years which has caused a lot of damage to lives and property.This problem seems to get worse and more people are affected year on year.This dissertation identifies four correlated fundamental problems that culminate into the annual flooding disasters.Rapid urban population,rapid urbanization,high-intensity short-duration rainfall and a restrictive land tenure system,all contribute to the annual disaster.By using remote sensing and other geospatial tools,this research attempts solutions to this flooding issue within the short,medium and long terms.This dissertation identifies four correlated fundamental problems that culminate into annual flooding in Accra.The main research work of this dissertation includes:1.Chapter 3 of the dissertation answers the question: How accurately can Landsat images be used in extracting impervious surfaces in a semi-arid topography or terrain? Geospatial Technologies and tools have made it quite easy than before,in studying the various phenomenal changes within the environment.Recent urban researchers have combined varied remote sensing data,techniques and software,and have performed various techniques to detect,extract,quantify and map this urban phenomenon.Generally,highresolution satellite images used by researchers and developers of extraction methods do not have a global cover.High-resolution images are very costly and may not worth the purpose of urban environmental monitoring.Therefore,it is difficult and almost impracticable to apply a particular extraction method on many heterogeneous terrains.Hence most methods cannot be replicated everywhere.Besides,monitoring the growth of urban built-up with in-situ methods is complicated.Landsat data is utilized to extract the impervious areas of the Greater Accra Region with a new spectral index,impervious surface extraction index(ISEI).ISEI solves the mixed pixel issue by utilizing the spectral richness of SWIR bands in Landsat sensors.Results are compared with five other spectral index algorithms.2.Chapter 4 of the dissertation answers the questions:(1)How has the urban land change patterns of Accra been affected by land-use practices?(2)How can geospatial tools be used to augment efforts at curbing uncontrolled urbanization in Ghana? This section is in two parts;(1)other extraction indexes were combined with ISEI to monitor major land-use of urban greater Accra region.The products were subjected to spatial metrics and spatial statistical analysis to determine the spatial distribution and pattern of mainly imperviousness and other land-use.(2)Uncontrolled urbanization is fundamentally linked to land-use administration which is intrinsically dependent on the type of land tenure system.Ghana Spatial Data Infrastructure Commission is a geospatial technology based model put forward to synthesize the traditional land ownership regime and the existing modern land administration practices.Gha SDIC model retains the traditional authority of land ownership and the regulatory powers of government local authorities.3.Chapter 5 of the dissertation answers the questions: How can geospatial techniques be employed in the delivery of both proactive and reactive disaster plan implementations in Ghana? This section is in two parts;(1)flood risk map is produced to indicate the flood risk areas within Accra.Peak runoffs are calculated and the flooded areas computed for 10 flood years.This determined the actual correlation between impervious surface areas and flooded areas for the greater Accra region.(2)High disaster risk scenarios are attributed to lack of proper spatial planning within uncontrolled urbanized communities.Local government responses in Ghana are always reactive,that is,sending relief items to rescued and displaced victims.An Integrated Disaster Risk Reduction Framework(IDRRF)model is put forward to bring unto a single platform and coordinate all disaster related institutions,agencies and departments which is decentralized down to the community levels.IDRRF model embraces some aspects of Sendai framework recommendations.Based on Gha SDIC,it incorporates existing reactive processes with proposed proactive short term,medium term and long term measures to deal with the perineal urban flooding disaster of Accra.4.Chapter 6 forms the concluding section for the dissertation It summarizes problems and the various proposed solutions to the research questions put forward and the set out objectives of the whole dissertation.Recommendations on how to implement proposed models are outlines.Future direction of research has also been expressed in this chapter.The initiatives of this dissertation include:1.The author put forward a method to extract impervious surfaces from Landsat image of a mixed urban environment of semi-arid terrain within a coastal savanna topography.This method has been published in an SCI indexed Journal ? Applied Ecology and Environmental Research?.2.The author modified spatial statistics and coupled with some spatial statistics to characterize the land-use spatial distribution of Accra.Furthermore,the author put forward a land administration model,Gha SDIC that synthesizes traditional land ownership wit h government land acquisition processes.It retains both the traditional authority of land and the regulatory powers of government agencies and departments.The model idea was first presented at the Doctorial Forum: Experiencing China-Smart Wuhan University.This method has been published in an SCI indexed Journal ?Sustainability?.3.The author put forward an integrated disaster risk reduction framework model,IDRRF is based on Gha SDIC and also brings onto a single platform all disaster related departments,institutions and agencies.It comprises a coordinated arrangement of both proactive and reactive disaster risk reduction measures in the short,medium and long term periods.The framework model was first presented at The Second International Workshop on Artificial Intelligence and Big Data Analysis,Macau.It has been submitted to an SCI indexed Journal?International Journal of Digital Earth?.