Flood Susceptibility Assessment Using Frequency Ratio Model: A Case Study of District Ghotki and District Kashmore, Sindh, Pakistan

Abstract

In recent years, flash floods in Ghotki and Kashmore districts in Pakistan have seriously affected both people and their ways of earning a living. Addressing challenges related to flooding means utilizing a methodology that considers both the hydrology, of water, the environment, the soil, the economy and social impacts. Flood susceptibility mapping helps inform how to control and plan floods. A bivariate probability analysis employing the frequency ratio (FR) methodology was conducted during this investigation to develop flood vulnerability assessments for Ghotki and Kashmore. A map was produced using the 130 past flood locations in the two districts. To establish the models, the data from these localities were randomly divided into 70% for model development and 30% for assessment. Among the parameters incorporated in the analysis were aspect, slope, elevation, rainfall, type of soil, use of land, proximity to roadways and rivers and NDVI and NDSI figures. How each factor affects flooding was assessed by checking its relationship with previous floods. From the analysis, scientists found that approximately 18% of the study area was classified as extremely flood susceptible, 30.9% as highly flood susceptible, 20.7% as moderately flood susceptible, 20.6% as minimal flood susceptibility and 9.8% as negligible flood susceptibility. Using the metrics from the validation set, the Foul Reader showed an accurate prediction rate of 75%. Moreover, the resulting susceptibility maps were compared to the real floods of 2010 and 2022, showing that the model reliably predicts flood-prone areas. As a result, the FR model is demonstrated to support the activities of governmental organizations, administrators and policy-makers in preventing and managing floods in the region.