Saleem et al. /Int.J.Econ.Environ.Geol.Vol. 11(1) 95-101, 2020 95 c Using Remote Sensing for Identifying Suitable Areas for Flood Shelter: A Case Study of Thatta, Sindh Pakistan Umer Saleem 1* , Takeshi Mizunoya 2 , Yabar Helmut 2 , Ammara Ajmal 3 1 Graduate School of Life and Environmental Science, 2 Faculty School of Life and Environmental Science, 3 Graduate School of Human Science, University of Tsukuba, Ibaraki, Japan 3 *Email: sparks.usar@gmail.com Received: 07 February, 2020 Accepted: 22 April, 2020 Abstract: The most recurring type of disaster in the world these days is flood because of the spread and extent of its effect on people, among all-natural disasters of the world. Human activities have paved the way for many of these flood behavior to change as they used to be in the past. Pakistan experienced one of the most devastating natural disasters in its history all across the country in 2010, but Thatta district in southern part got severely affected during this flood. For the research, a simple yet efficient methodology Normalized Difference Vegetation Index (NDVI) by using remote sensing images for identifying flood hazard areas was utilized. Geographic Information Systems (GIS) helps in finding shelter areas with a minimum effect of floods. It is essential to realize the importance of mapped results in consideration of manual flood management in future. The method used in this study is robust enough to explain the flood hazard for suggesting suitable shelter sites in case of flooding events. This would help disaster management bodies and other related agencies to formulate the development plans while keeping the hazard areas, which are unsuitable for development due to flood risk in the future. Keywords: Floods, remote sensing, NDVI analysis, disaster management. Introduction The most important for flood disaster management is a monitoring system that needs to be done in real-time, especially for the early response, designing and planning for both short and long-term mitigation strategies (Wang, 2004). Widespread flood analysis is significant, considering the socioeconomic status and environmental consequences (Markantonis et al., 2013). Remote sensing data and repetitive combination with Geographic Information Systems (GIS) technique have made its implication possible for flood mapping and monitoring in real-time. The availability of sensors with the properties to provide multiple ranges of the electromagnetic spectrum has enabled us to answer cost-effectively about the flooding range and its estimation (Smith 1997; Sanyal and Lu 2004). In the flood-prone river, Kosi in north Bihar, India, the utilization of remote sensing and GIS technique showed an integrated approach for flood risk mapping by utilizing land-use, land cover, topography, geomorphology, and population density (Bapalu and Sinha 2005; Sinha 2008). Others have focused on more theoretical approach using Multiple Criteria Decision Analysis (MCDA), analytical hierarchy process (AHP) model, Frequency Ratio (FR) model, remote sensing (RS) Normalized Difference Vegetation Index (NDVI) model and GIS techniques which are instrumental, reliable and provide accurate information as well analysis for flood-prone zones. The efficiency of the MCDA approach is its application when threre is lack of data for some areas, therefore, local planners use it for flood mitigation. The analytical hierarchy process (AHP) model is suitable for China in making flood diversion plans (Zou et al., 2013). However, the limitation of the AHP model was a correlation to its subjugation on the provision of information by experts (Chen et al., 2011). Additionally, FR is an understandable way to assess flood risk analysis and mapping for future management (Liao and Carin, 2009). Using remote sensing (RS) and Normalized Difference Vegetation Index NDVI to detect the smog and temperature values reflect the impact of urban development areas (Jahan et al., 2019). In terms of functioning, all models produce similar and comparable results for flood mapping. Pakistan is one of the most disaster-prone countries in the world (Ahmed, 2013). It is among few countries blessed with very diverse topography, consisting of both highlands like northern parts and lowland areas in the southern region. Pakistan experienced the worst natural disaster in terms of the number of people affected in 2010. During the monsoon season, heavy rainfall caused flooding in the northern region of Pakistan. This rainfall, when reached Khyber Pakhtunkhwa region of Pakistan, the Indus river, started to breach from the embankments and canals along the river course, which resulted in massive destruction in most parts. Flooding initiated in mid- July 2010 and continued until early September. Its effect was all over affecting more than 20 million people (Table 1). Flooding in the Indus river caused a massive loss of more than 2113 lives and an economic loss of US $ 9,500, 000 (EM-DAT, 2019). According to United Nations estimates in 2010 flooding, the humanitarian loss alone was the largest among the three worst natural disasters in the past decade including the Asian tsunami, earthquakes in Kashmir and Haiti combined. The frequency of floods has Open Access ISSN: 2223-957X Int. J. Econ. Environ. Geol. Vol. 11 (1) 95-101, 2020 Journal home page: www.econ-environ-geol.org Copyright © SEGMITE