Range Management under Uncertainty
把牧场管理建模为随机控制问题，管理者定期选择放牧率并决定是否采取改善措施，用有限马尔可夫链分析最优策略，并检验风险中性和风险规避情形下的敏感性。
Abstract The problem of rangeland management is modeled as a stochastic control problem. The manager periodically chooses stocking rates and decides whether to use a treatment which improves the rangeland. His decisions have an uncertain impact on the rangeland. The theory of finite Markov chains is used to analyze the solution to the control problem. The stochastic specification of the model is varied to test the sensitivity of the optimal policies. These are quite robust except at certain levels of the range condition. Cases of a risk‐neutral and risk‐averse manager are studied.