Using Bayesian Network to predict the watershed land use type of Çankırı Acıçay-Tatlıçay

Abstract

In recent years, experts have identified that climate change and global warming affects stream flow regime. These changes cause floods and erosion in creeks, streams, rivers etc. Especially in semi-arid watersheds, the structure of the land usage type is an important factor in preventing possible disasters. The aim of this study is to determine watershed land usage type by using hydro-morphological structure of stream and some physical water quality parameters. To do so, hydro-morphological observations and some physical water quality parameters were collected from 513 different points in Acıçay and Tatlıçay watershed. For this purpose, four different Bayesian network scenarios were considered to see the changes in the type of the land use. In this scenario, the prediction probability of the watershed land usage type was determined with different parameters. In conclusion, coniferous forest was predicted with the highest probability rate of %97.

Keywords

• bayes ağları
• arazi kullanma türü
• rosgen dere sınıfı
• çankırı
• çankırı
• bayesian network
• land use type
• rosgen stream classification
• çankırı