Browsing by Author "Khadra, Mohamed Shawky Mohamed Ali"

Now showing 1 - 1 of 1
  • Thumbnail Image
    ItemOpen Access
    Remote Sensing For An Improved Geospatial Flash Flood Susceptibility Modeling Over An Arid Environment
    (2020-03-31) Khadra, Mohamed Shawky Mohamed Ali; El-Sheimy, Naser; Hassan, Quazi K.; Zhang, Yun; Kattan, Lina; Wang, Xin; Charif, Omar
    Flash floods are the foremost cause of irretrievable environmental damage in the arid Arabian Peninsula. The better understanding of the geomorphologic, topographic, climatic, and hydrologic characteristics of a selected watershed, and determining their geospatial relationships with respect to the flood extent are the core steps for mitigating and minimizing negative impacts of flooding. Therefore, the overall aim of the current study was to employ different remote sensing datasets in predicting prone areas to flash floods in the 'wilayats' (i.e., cities) of El Hamra, Bahla, and Nizwa, Ad Dakhiliyah Governate, the Sultanate of Oman as an example of the arid areas. Precipitation is a crucial variable for studying various climate-related research such as flash flood monitoring and prediction. The performance of five global satellite precipitation estimates (GSPEs) was evaluated using the available sub-daily and daily ground rainfall records. Generally, the five sub-daily and daily GSPEs showed good performance compared to the in-situ measurements. Moreover, statistical error models were employed to quantify the uncertainties in the daily GSPEs. Accurate digital terrain model (DTM) and channel network/orders with fine spatial details are mandatory for flood extent modeling. The DTM and its derived channel network have been employed successfully in many studies to extract various geomorphometric, topographic and hydrologic attributes. Therefore, a new pixel-based method was developed to quantify the horizontal accuracy of channel networks/orders-based three global DEMs using those extracted from LiDAR datasets as references. The vertical accuracy of global DEMs were also evaluated utilizing reference LiDAR elevation data. PALSAR DTM (12.5 m) and its derived channel network/orders were found to be the optimal candidates to derive geospatial layers required for flood susceptibility modeling. Flood susceptibility models were developed to define the likelihood of future flash flooding in the study area. The spatial relationships between flood triggering factors and flood inventory map were quantified. The integrated bivariate and multivariate statistical methods-based flood susceptibility models provided precise maps to predict future flood-prone areas under a rainfall intensity close to that which prevailed during the past flood event, at both high- and low- lands. The developed flood susceptibility models can contribute to mitigating the negative impacts of flash floods.