No abstract

No abstract

No abstract

Object oriented image classification (OOIC) and neural network aided Markov Chain (MC) modeling tools were used to map and predict land use and land cover (LULC) changes. A case study in the Kiskatinaw River Watershed (KRW) of Canada was presented. With an overall classification accuracy of 90.45%, the multi-temporal Landsat satellite images of KRW were analyzed for 11 selected LULC types. It was found that KRW experienced a significant wetland depletion along with a change in forest cover types from 1984 to 2010. The vulnerability of LULC change in different parts of KRW was predicted through MC modeling based on the obtained transition probability, and the results indicated slight LULC changes from 2010 with a wetland depletion of 67.89 km 2. In summary, the proposed methods generated valuable results for informed LULC management and hold the potential to be applied to other watersheds.

No abstract