This research project is directed at improved understanding and prediction of atmospheric variability with time scales longer than a week. This 'low-frequency variability' is generated by external forcing (as in the annual cycle), dynamical processes internal to the atmosphere such as blocking, and by interactions between the atmosphere and the underlying media (primarily the oceanic mixed layer and the land-surface hydrology). The research involves the diagnosis of observational data sets using statistical techniques such as compositing, digital filtering, principal component analysis, singular value decomposition, and cluster analysis. Results are interpreted in light of the current theoretical understanding of weather and climate systems. The research will address (1) detection and interpretation of interannual and interdecadal climate variability in Northern Hemisphere temperature and pressure fields, (2) nonlinear aspects of low-frequency variability including the question of whether the Northern Hemisphere winter circulation exhibits distinct 'circulation regimes' manifested in truly bimodal frequency distributions, (3) relationships between patterns of sea-surface temperature anomalies observed in different seasons and in different parts of the world ocean, (4) some poorly understood aspects of the climatological mean annual cycle and (5) systematic errors in extended-range numerical weather prediction. This research is important because it seeks to improve our understanding of weather and climate processes and hence has the potential for improving our ability to predict weather and climate.
