A Comprehensive Assessment and Modeling of Land Use Changes in a Flood-Prone Watershed, Northeast of Iran

Abstract

The Gorgan-rud river basin in the northeast of Iran has undergone extensive land use changes causing floods during the past decades. This study aims to investigate land use change and its correspondence to flood events during 1987–2017 in the study area. Six pixel- and object-based classifications algorithms were used for land use change modeling. The objectedoriented support vector machine with a scale length of 50 m was selected as the superior algorithm. The results indicated that croplands and residential/industrial areas have continuously expanded during 1987–2017. However, the area of forests, rangelands, and barren lands has decreased. It is predicted that by 2035, the area of croplands will expand by another 20,397 ha, and the forest area will shrink by about 16,966 ha. The spatiotemporal pattern of flood occurrences and damages correspond well with the pattern of land use changes in the study area.