Flood mapping using analytical hierarchy process and fuzzy logic.

Abstract

Kampala city has experienced floods on a number of occasions after a downpour, causing damage to people’s homes, water pollution, disruption of traffic and economic activities. This study aims at flood susceptibility mapping using analytical hierarchy process, fuzzy criteria and a combination of two algorithms AHP and fuzzy in Kampala city Uganda. Flood maps are generated based on six flood influencing factors (rainfall intensity, LULC, soil moisture index, soil type and drainage density). Thematic maps for each of the six flood conditioning factors were generated, and the generated thematic maps were used to develop flood maps using AHP, fuzzy criteria and a combination of both algorithms AHP and fuzzy. Using AHP the weight derived for the factors were Rainfall 59%, Slope 7.514%, Drainage density 14.291%, Soil type 5.243%, LULC 11.002%, Soil moisture 2.95%. The generated flood susceptible maps were assessed based on the area under the curve as shown by the receiver operating curve, AHP had an accuracy of 56.16%, fuzzy criteria had an accuracy of 81.42% and the combination of both algorithms had an accuracy of 71.98%.