Three Essays on Risks of Natural Disasters in the U.S. Forest Sector.

详细信息   

• 作者：Chen ; Xuan.
• 学历：Doctor
• 年：2013
• 毕业院校：The University of North Carolina
• Department：Economics.
• ISBN：9781303798511
• CBH：3586089
• Country：USA
• 语种：English
• FileSize：2893195
• Pages：147

文摘

The U.S. forest sector faces various kinds of natural disaster risks. Three of them are evaluated in this paper,including wildfire risks,southern pine beetle and lightning fire ignitions. In the first essay,wildfire risks are assessed. Wildfire is one of the leading causes of damage and economic losses. While individual wildfire behavior is well studied,new literature on broad-scale e.g. county-level) wildfire risks is emerging. My paper studies wildfire risks,conditioning on crucial informational variables,across both spatial units and time periods. Several statistical models are used to quantify the risks. I develop several maximum likelihood estimation methods to account for spatio-temporal autocorrelation in conditional risks. A group index insurance scheme is proposed and its associated actuarially fair premium rates are estimated and presented. Implications for wildfire management policies are also discussed. In the second essay,the risks associated with southern pine beetle SPB) are evaluated,which are among the leading factors killing timber in the southern U.S. In light of the spatio-temporal autocorrelation in SPB outbreaks Mawby and Gold,1984),I apply a spatio-temporal block bootstrapping method. This approach is motivated by overlapping block bootstrapping methods developed in the autoregressive time series scenario Kunsch,1989) and the dependent spatial data scenario Hall,1985). Compared with conventional methods that built spatio-temporal dependence of SPB risks into structural models Gumpertz,Pye et al.,2000； Gan,2004； Zhu,Huang,and Wu,2005),a block bootstrapping method has several advantages. The bootstrapping method doesnt require a weight matrix which has a specific,analytic structure that needs to be predetermined and carefully defined to guarantee identifiability. Also the block bootstrapping method can circumvent complications and strong assumptions associated with maximum likelihood estimation. It is found that both vegetation structures and climate factors play a significant role in SPB outbreaks. Further,a county-level group index insurance plan is proposed and associated actuarially fair premium rates are estimated. Given the fact that no compensation scheme for SPB epidemics currently exists,this insurance product will help lessen forest owners losses if such an insurance market emerges in the future. In the third essay,natural wildfire ignitions caused by lightning are assessed. Unlike human related wildfires for which timber owners may be able to trace the responsible persons to claim losses,forestland owners essentially have no means to recover their losses against lightning-induced wildfire. In light of the fact that there are very few risk management instruments available to compensate timber losses,our paper studies risks of lightning induced wildfire. Both linear models and count models are used to quantify the risks. Block bootstrapping methods are also applied. Some relevant observable variables,such as environmental and climatic factors,are found to be statistically significant factors related to lightning wildfire risks. A group index insurance scheme is designed and its associated actuarially fair premium rates are calculated. Implications for forest management policies against natural wildfire risks are also discussed.