Assessment of flood hotspot at a village level using GIS-based spatial statistical techniques

The spatial mapping of flood menace extents is crucial for the effective and competent enactment of risk-lessening strategy. We focused on geographical pattern and variation in flood-affected villages in Bongaon sadar sub-division, West Bengal, India, during the period between 1996 and 2016. To appraise the indigenous smoothing and dissimilarity of flood-affected/non-affected villages, GIS-based Voronoi statistics were used. Inverse distance weighting (IDW) is used to interpolate and predict the pattern of flood-affected/non-affected zones across the sub-division. Moran's I index statistics was considered to appraise spatial auto-correlation among the flood affected and non-affected villages. Getis-OrdGi*(d) statistics was employed to recognize the flood hotspot and cold spot areas within the study site. The higher magnitude of Moran's I was calculated as 1999-2001, 2004, 2011, 2013, 2015, and 2016. The high Z score was recorded in 1996-1999, 2001-2003, 2011, 2013, and 2014 indicated a spatial clustering of flood-affected villages. The predictive map derived through IDW showed that 7.76% (64.59 km(2)) area comes under very high threat zones of flood, followed by 16.27% as high risk, 24.49% as medium risk, 23.97% as low risk, and 27.51% as very low risk. This study determines the solicitation of GIS-based prophecy for the impost of revelation mapping, so as to define the latitudinal extent and frequency of areas where most affected villages are located and potential risk areas.