Hemispheric-Scale Changes in Air-Temperature Probability Distributions

摘要分析气温的概率分布可以对最近的气候变化提供相当深入的了解。 首席研究员过去的工作表明，对时变温度的分析是传统气候变化检测方法的有益延伸，特别是那些只关注平均气温的方法。 这项研究将使用半球尺度的每日数据来估计气温概率分布的时变斜率和其他参数，这些参数可以以创新的方式评估最近的气候变化。 除了随时间变化的温度，可变性，偏度和峰度，以及空气温度的持久性的抵抗措施，将来自历史的空气温度数据在北方半球。 随时间变化的温度，方差，形状的措施，和持久性是明显的，只有在日常的气温记录，这是最近成为在大的空间尺度。参数之间的函数关系将被用来评估分布和持久性的形状变化如何响应变化的平均气候。 广泛的映射与聚类分析相结合，将被用来确定凝聚力的空间模式，可以进一步分析。 将评估气候变化在整个气温分布中同样发生以及形状和持久性参数不随平均值变化而变化的假设。 这项研究将表明，每日气温概率分布内估计的变异性可以远远不同，比那些单独从平均气温得出的启发。评估气温概率分布变化的性质对于全面了解近期气候变化至关重要。 在确定气候变化的生态、经济和社会方面时，气温的平均值、变率和持续性的同时变化具有广泛的影响。 该项目的成果将有助于国家和国际机构更好地了解过去，并为未来的气候变化作出规划。这项研究的更广泛的影响还将包括（i）在专业会议和科学期刊上广泛传播结果，（ii）支持研究生研究助理，以及（iii）加强研究生和本科生教学。 除了在专业会议和期刊上传播研究结果外，还将在项目网站上提供研究人员调查结果的数据和图表。

Abstract Analyzing air-temperature probability distributions provides considerable insight into recent climatic change. Past work by the principal investigator has shown that the analysis of time-varying percentiles is a useful extension of traditional methods for detecting climatic change, particularly those that focus solely on mean air temperature. This research will use hemispheric-scale daily data to estimate time-varying percentiles and other parameters of air-temperature probability distributions that can evaluate recent climatic change in innovative ways. In addition to time-varying percentiles, resistant measures of variability, skewness, and kurtosis, as well as the persistence of air temperature, will be derived from historical air-temperature data over the Northern Hemisphere. Time-dependent changes in percentiles, variance, shape measures, and persistence are apparent only in daily air-temperature records, which have recently become available over large spatial scales. Functional relationships between parameters will be used to evaluate how changes in the shape of the distribution and persistence respond to a changing mean climate. Extensive mapping combined with cluster analysis will be used to identify cohesive spatial patterns that can be analyzed further. Null hypotheses that climatic change occurs equally throughout the air-temperature distribution and that shape and persistence parameters will not change with changes in the mean will be evaluated. This research will demonstrate that variability estimated within daily air-temperature probability distributions can be far different and more illuminating than those drawn from mean air temperature alone. Evaluating the nature of changes in air-temperature probability distributions is critical to gaining a complete understanding recent climatic change. When determining the ecological, economic, and social dimensions of climatic change, the concurrent changes in mean, variability, and persistence of air temperature have wide-ranging impacts. Results from this project will help national and international agencies to better understand the past and to plan for future climatic changes. The broader impacts of this research also will include (i) wide dissemination of results at professional meetings and in scientific journals, (ii) the support of graduate research assistants, and (iii) the enhancement of graduate and undergraduate teaching. In addition to dissemination of results at professional meetings and journals, data and graphics from the researcher's findings will be made available from the project's website.