improving landslide prediction results using shannon entropy theory

a review of damages caused by landslide shows the need for studying effective parameters in landslide occurrence and prediction. this study aims to improve landslide prediction results for tutkabon region in gilan province. to this end, shannon entropy theory was employed for modeling and considering data uncertainty. slope, height, geomorphologic conditions, earth’s curve, closeness to river, and closeness to fault have been considered as the parameters affecting landslide. using shannon entropy theory, the weight of each parameter along with the uncertainty effect on the results was calculated and a landslide risk map for study area prepared. finally, the comparison of the situation of landslide points in the under-study region with the modeled risk map was used for evaluation of the results. the under curve area of prediction rate curve was calculated at 0.69 considering shannon entropy, and 0.54 disregarding shannon entropy.