Seasonal (Pre)Conditioning of Daily Weather by Low-Frequency Climate Variability

We investigate the influence of low frequency climate variability on the seasonal probability density function (PDF) of daily weather. First, the best-fit theoretical PDF is selected to represent daily temperature variance for summer and winter at locations across North America. These climatological PDFs are interesting and useful in and of themselves as they represent local climate conditions. Then the theoretical PDFs are fitted to each season's local daily temperatures. Interannual, decadal and longer time scale evolution of their parameters is examined in relation to natural modes of climate variability and long-term trends over the last six and a half decades. Using multivariate statistical analysis techniques, we will quantify the low-frequency variability in the shape of the seasonal PDF's of daily temperature and how their shape deviates from climatology over time and space as conditioned on the background climate state. We will specifically examine the probabilities of extremes represented by the warm and cold tails of the seasonal PDFs and how these probabilities vary in relation to the regional and global climate state. These results could be used to condition the probability of outcomes being forecast by short- and medium-range and sub-seasonal prediction models. To the extent that low frequency climate variability can be quantified to pre-condition weather outcomes, seasonal probability forecasts of daily weather PDFs can also be based on this work and will be examined.