Artificial Intelligence in Geospatial Analysis for Flood Vulnerability Assessment: A Case of Dire Dawa Watershed, Awash Basin, Ethiopia

Table 1

Detailed data and sources used in the study.

S/N	Raw data	Derived criteria	Sources	Spatial/Temporal resolution	Purposes
1	Precipitation (mm)	Rainfall	National Meteorology Agency	Dire Dawa and its watershed for 2015–2020 periods	To derive rainfall erosivity factor (R) and areal rainfall distribution (kriging)
2	Landsat 8 OLI/TIRS/C2L2	LULC		Dire Dawa and its watershed for 2006, 2008 and 2021	To generate land use/land cover for the watershed (supervised classification)
3	Landsat 8 OLI/TIRS/C2L2 (B4 and B5)	SAVI	https://earthexplorer.usgs.gov/	Dire Dawa and its watershed	To generate information about soil influences on causing flooding
4	Landsat 8 OLI/TIRS/C2L2 (B3 and B5)	NDVI		https://earthexplorer.usgs.gov/	To quantify the vegetation along the flood-prone areas
5	Soil types	K	Ministry of Water, Irrigation, and Energy (MoWIE)		To derive soil erodibility factor (K)
6		TWI			To identify hydrological flow path
7		SPI			To measure the erosive power of the flooding
8	DEM (12.5 × 12.5 m)	E	https://search.asf.Alaska.edu/	Dire Dawa and its watershed	To generate the differences between the consecutive points in the watershed
9		S			To derive the surface differences
10	River networks	SD	Ministry of Water, Irrigation, and Energy (MoWIE)		To derive the Euclidean distance between the settlements and main river course
11	Landforms	G	geoportal.rcmrd.org/layers/servir%3Aafrica_landforms		To generate the physical/geographical of the city and its watershed
13	Road networks	RD	Digitized from Google Earth and processed in GIS environment		To derive the Euclidean distance between the settlements and flood-prone areas
14	Census data	PD	Central Statistical Agency (CSA)	2006, 2008, 2021	To identify the settlement with dense population in the city and its watershed