Wildfires are severe natural disasters and environmental problems. This paper describes a methodology for constructing a dynamic decision model which aims to support the human expert's decision making in fire containment. The main idea is based on Markov decision processes (MDP) which can handle the uncertainty and dynamic features of decision making problems. In order to apply MDP to large scale real world problems, we use a factored manner to describe and formulate the dynamic decision model. This helps the problem expression and solution, and the policy generated by the model is explicit and specific actions of wildfire suppression measures which is easily understood by human decision makers. Furthermore, we simulate two scenarios and verify the decision effects solved by the model. The results reveal that our model has prominent performance on wildfire containment and deals with multi-objective decision problems easily and intrinsically.