Climate Analysis and Extended Range Seasonal Prediction

摘要：Gray， William M. Colorado State University该项目涉及分析和进一步发展与大西洋盆地季节性飓风活动、非洲萨赫勒地区季节性降雨、ENSO变化、亚洲选择性区域季风降雨和中纬度地区季节性可预测性潜力的扩展范围定量气候预报方案有关的基本科学概念。与有关潜在机制的基础科学一起，这项工作寻求9-14个月预测前置时间的准确性，足以允许制定和实施有意义的政策和缓解行动。潜在气候前兆信号的频谱将被详尽地探索。将以后发模式研究潜在的可预测性，主要根据过去35至45年（主要是1950年以来）的全球数据集，但也有一些扩展到更早的数据。同样，将广泛探讨使用最小绝对偏差和普通最小二乘程序组合预测信号的统计回归方法。虽然已经获得了令人惊讶的熟练的季节预报测试结果，但这些结果表明，通过对气候变量的额外组合进行进一步分析，可以提高预报技巧。这项研究同样强调诊断工作，以便更好地了解产生远距离前兆信号的潜在大气和海洋过程。这项研究很重要，因为它试图提高我们预测气候变化对经济和人类安全的重要性的能力。

Abstract ATM-9417563 Gray, William M. Colorado State University Title: Climate Analysis and Extended Range Seasonal Prediction This project involves the analysis and further development of basic scientific concepts related to extended range quantitative climatic forecast schemes for Atlantic basin seasonal hurricane activity, African Sahel seasonal rainfall, ENSO variations, selective regional Asian monsoon rainfall, and the potential for seasonal predictability in the middle latitudes. Along with the basic science concerning underlying mechanisms, this work seeks accuracy at 9-14 month forecast lead times, sufficient to allow the development and implementation of meaningful policy and mitigating actions. The spectrum of potential climate precursor signals will be exhaustively explored. Potential predictability will be studied in the hindcast mode, principally from global data sets for the last 35 to 45 years (primarily since 1950), but with some extensions to much earlier data. Similarly, statistical regression methodology for combining predictor signals using both Least Absolute Deviations and Ordinary Least Squares procedures will be extensively explored. Although surprisingly skillful seasonal hindcast test results have been obtained, these results suggest that greater forecast skill is possible through further analyses with additional combinations of climate variables. Equal emphasis in this research is directed toward diagnostic work for greater physical understanding of the underlying atmospheric and oceanic processes responsible for long range precursor signals. This research is important because it seeks to improve our ability to predict climate variations of major economic and human safety importance.