EPILOGUE
It has become increasingly important to provide attribution for low frequency variations and change in Earth's climate. Whether this is for improved scientific understanding, predictability assessment, or to better inform societal planning and decision making, CDC is dedicating increasing resources to climate change research. At CDC, we seek to offer dynamical explanations for observed low frequency variations and change, thereby drawing strongly upon our expertise on seasonal to interannual variability, especially regarding air-sea interactions and teleconnective influences.
A key challenge that we will pursue in CDC is to understand and anticipate the regional characteristics of climate change. While there is now little question that the climate has changed in a globally and annually averaged sense, it is unclear what the local manifestations of this are, nor do we appreciate their seasonal dependencies. Beyond its relevancy to long term planning, this problem is of high relevance to seasonal climate predictions. The fact is that the leading source of US winter temperature skill in the 1990's is due to the so-called optimal climate normals (OCN) tool, which we understand to be essentially a trend prediction. It is necessary that a physical explanation for such trends be given, and that they be clearly distinguished from low frequency climatic variations. Most apparent of these trends is the US wintertime surface warming, but other seasons show a more complicated pattern for temperature and rainfall change. We believe that progress can be made by improving our understanding of the regional responses to the slow, systematic changes in tropical oceans such as illustrated in Fig. 5.1, and we expect that much is to be gained from our existing knowledge of the interannual impacts of tropical forcing.
The change in the oceans itself is a problem that will focus future CDC decadal climate research. The mean change in ocean temperatures is a question that will require increased analysis of coupled ocean-atmosphere models. We expect to partner with GFDL, NCAR and other interested scientists to diagnose and understand the variability in coupled model simulations, both natural and forced. We are especially interested in the sensitivity of ENSO to climate change, both its statistical properties and its interannual global impacts. A related challenge is to understand whether the year-to-year predictability of climate will change appreciably under the influence of human-induced mean change. Will ENSO as an oceanic phenomena become more predictable? Is it possible also that new regions will begin to have useful ENSO-related climate predictability? Likewise, we would like to understand whether the seasonal cycle of predictability will change due to an altered mean climate. These questions, among others, cut across time scales, and the greatest payoff in solving them may in fact be to advance key problems on shorter time scales, such as interannual prediction.
Contributed by: M. Alexander, A. Capotondi, H. Diaz, M. Hoerling, H. Huang, X. Quan, D. Sun, and K. Weickmann.