Spatio-temporal modelling, network design, health effects, and dimensional analysis

世界人口迅速增长，最新预测在目前条件下，到2100年将继续增长到150亿，这导致人们越来越担心环境风险水平的提高。预计将出现更多的事件，例如高温事件和化石燃料（例如煤炭）消耗量的增加。这一黑暗的未来预示着更加需要有效地监测空间和时间上的环境过程沿着在空间和未来无法监测的情况下预测这些过程的准确方法，以支持通过有效的政策和监管来管理和减轻这些风险。统计学家在这个未来中发挥着重要作用，特别是在量化大量的不确定性方面。拟议的研究将继续申请人关于时空过程监测网络的最佳设计的长期研究方案，并利用所得数据为空间预测和未来过程建立这些过程的模型。我们继续研究算法，以克服与选择最佳位置相关的计算挑战，同时确保它们是在无偏的情况下挑选的。但是，我们也将提供扩展的理论来解决现有的监测网络出现的问题，没有被选择在一个公正的方式。特别是，由于现有的网络台站有时设在优先选定的地理位置，我们将继续寻求方法，以减少以这种方式定位台站所引起的统计估计误差。该研究还将探索一种有前途的客观方法来确定它们的位置，确保它们在空间上分布良好。 研究的其余部分涉及使用来自这些网络的数据，可能会对优先选址的影响进行调整，以创建良好的时空模型。我们将继续目前的研究，并寻求温度场的模型，以扩展其地理适用范围-以前的工作集中在北美的太平洋西北地区，但现在我们将进入英国，与英国气象局合作监测温度，因为在气候变化的时代，高温是一种健康危害。更具体地说，我们将对考虑到人类行为的极端温度下的暴露进行建模，即通过对汽车舱或家庭等微环境中的暴露进行建模。但气候变化影响的不仅仅是人类。关于海洋哺乳动物的其他研究将继续进行，因为某些区域的海洋哺乳动物数量正在减少，原因尚不清楚。新的研究对它们的路径进行建模，作为确定它们的能量需求以及如何满足这些需求的基础，考虑到变暖对这些生物食物供应的影响。

The burgeoning World population, newly predicted under present conditions to continue growing to as much as 15billion by the year 2100, is leading to increasing concerns about elevated levels of environment risk. More incidents e.g. high temperature episodes and the increased consumption of fossil fuels e.g. coal are expected. This dark future portends an increased need to effectively monitor environmental processes over space and time along with accurate methods for predicting these processes where they cannot be monitored in space and in the future, to support the management and mitigation of these risks through effective policy and regulation. Statisticians have a major role to play in this future, particularly in quantifying the uncertainties that abound. The proposed research will continue the applicant's long-term research program on the optimum design of spatio - temporal process monitoring networks and the use of the resulting data for modelling these processes for spatial prediction and future processes. We continue work on algorithms for overcoming the computational challenges associated with selecting optimal locations while ensuring that they are picked in an unbiased. But we will also provide extended theory for addressing the problems that arise with existing monitoring networks that have not bee selected in an unbiased way. In particular, since existing network stations are sometimes sited at preferentially selected geographical locations, we will continue seeking methods to reduce errors in statistical estimates induced by locating stations in that way. The research will also explore a promising objective method for siting them, one that insures they are well distributed over space. The remainder of the research concerns the use of data from these networks, possibly with adjustments for the effects of preferential siting, to create good spatio - temporal models. We will continue current research and seek models for temperature fields that extends their geographical domains of applicability - previous work focused the Pacific Northwest region of North America but now we will be moving into the United Kingdom to work with the UK Met Office on monitoring temperatures, since in the age of climate change, high temperatures are a health hazard. More specifically we will model exposures to extreme temperatures that take account of human behaviour, i.e. by modelling such exposures in micro-environments like car cabins or homes. But climate change is affecting not just the human species. Other research on marine mammals will continue since their numbers in some regions are being depleted for reasons not yet understood. New research on modelling their paths as a basis for determining their energy needs and how well these will be met, given the impact of warming on the food supply for these creatures.