Abstract
There is growing interest in improving our understanding and modeling capabilities of subseasonal variability in the Earths climate system, partly due to the evidence that there is unexploited predictability at time scales between 2 weeks and 2 months. This predictability arises from large-scale recurrent patterns of subseasonal variability that reside in the Tropics (e.g, MJO) as well as in the mid-latitudes (e.g., PNA, AO). NOAA has a strong interest in providing forecast capability at these time scales in order to accommodate their seamless suite of forecast products and fill in a present gap in capability/products for lead-times between 2 weeks and 1 month. Based on the elevated level of interest in this time scale, the strong indications that it can lead to as yet untapped predictability, and the fact that modeling/forecast systems have shown recent improvements in their ability to represent the above patterns of variability (e.g., NOAA/NCEP), we are carrying out a set of analysis activities that will contribute to the understanding of a number of outstanding questions regarding subseasonal predictability and that will also facilitate NOAAs subseasonal predictive capability. Specifically, we are examining the dynamic predictability and present-day forecasting skill of subseasonal variability - namely in terms of the MJO, SST, and extra-tropical modes of variability, in a large suite of model hindcasts from the NOAA/NCEP operational coupled forecasting system (CFS). The presentation will mainly focus on our results to date concerning the predictability and forecast skill analysis of the MJO in a large subset of the CFS hindcast data set. In this case, the data set consists of three 5-member 90-day forecasts centered on the 1st, 11th, and 21st of each month from 1982-2003. We hope these efforts will result in an improved understanding of the dynamical predictability of the MJO, an assessment of present-day MJO forecast skill within NOAAs current operational forecasting system, and facilitate NOAAs efforts at incorporating and interpreting subseasonal predictions into their seamless suite of forecasting products.