Flood disaster monitoring with ALOS/PALSAR observation

We investigate an effective algorithm to identify flood-inundated area with ALOS/PALSAR observations. Radar cross section from water surface is generally smaller than that from ground surface without water, which derives larger differences of radar cross section in flood-inundated area because there is no water in this area before disaster. Variances of differences in 3×3 pixels are also enough to identify flood-inundated area, and more effective algorithm to identify flood-inundated area will be obtained by taking radar cross section differences and variances of differences into account.