A. Busalacchi
Search for other papers by A. Busalacchi in
Current site
Google Scholar
T. N. Palmer
Search for other papers by T. N. Palmer in
Current site
Google Scholar
Full access

This special section of the Journal of Climate contains papers based on presentations made at the First International Climate Variability and Predictability (CLIVAR) Science Conference “Understanding and Predicting our Climate System” held in Baltimore, Maryland, in June 2004.

CLIVAR is a project of the World Climate Research Programme. The origins of the project lie in the earlier Tropical Ocean Global Atmosphere (TOGA) and World Ocean Circulation Experiment (WOCE) projects. While TOGA focused on the effects of tropical air–sea interactions, notably associated with an El Niño event, WOCE provided a global perspective for the temporal variability of the World Oceans, from top to bottom. CLIVAR has a wider purpose: to observe, simulate, and predict the earth’s climate system, with a focus on ocean–atmosphere interactions, enabling better understanding of climate variability, predictability, and change, to the benefit of society and the environment in which we live.

Over 640 scientists from 56 different countries attended the CLIVAR conference, which addressed a number of key themes central to CLIVAR efforts: seasonal-to-interannual climate prediction, the monsoons, decadal prediction, anthropogenic climate change (paleoclimate), and applications of CLIVAR science. The papers published in this section represent but a fraction of the many hundreds of presentations made in Baltimore.

Vera et al. discuss both the North American Monsoon Experiment and the Monsoon Experiment of South America. These two experiments share common objectives: to understand the key components of the American monsoon systems, to determine the role of these systems in the global water cycle, to improve observational datasets, and to improve simulation and prediction of the American monsoons. Recent progress and future challenges are outlined.

The theme of monitoring and predicting the earth’s climate is the basis of Trenberth et al.’s paper, in which the notion of a “climate information system” is developed to observe and track changes in the climate system as they occur, and also to enable an understanding of these changes and an analysis of their origin. Trenberth et al. emphasize the need for benchmark observations to anchor space-based observations and trends, and note the need for institutionalized climate reanalysis procedures.

Lehodey et al., in a paper addressing the relationships between climate variability and fisheries, provide an example of the wider relevance of CLIVAR science. They present examples of these relationships at different time scales and for species covering various marine ecosystems ranging from equatorial to subarctic regions. They describe some of the known mechanisms linking climate variability and exploited fish populations, as well as some leading hypotheses and implications for the modeling and management of fish population dynamics. The paper concludes with recommendations for collaborative work between climatologists, oceanographers, and fisheries scientists to resolve some of the outstanding problems in the development of sustainable fisheries.

On the longest time scales, Cane et al. discuss progress in paleoclimate modeling. Here model results are outlined showing how reduction in El Niño activity, as observed in the early to mid-Holocene, can be seen as a response to altered orbital configuration at that time. Cane et al. note that an urgent challenge for paleoclimate modeling is to explain and simulate abrupt changes observed during glacial epochs. Finally, they note that climate variability over the last millennium is associated more with radiance changes due to volcanic eruptions and solar output than to orbital changes.

Hegerl et al. discuss the problem of detecting and attributing climate change, going beyond the mean temperature signals. For example, they note that while sea level pressure trends are detectable from the observations, these are significantly stronger in reality than can be simulated in climate models. This raises questions about the extent to which climate models have adequately captured all aspects of the dynamics of anthropogenic climate change. Hegerl et al. also discuss the detection of the penetration of the anthropogenic temperature signal into the ocean interior.

Goswami et al. discuss the dynamics and predictability of the Asian–Australian and American monsoon systems, respectively. Predicting monsoon variability remains a challenge, particularly for the Asian monsoon, not least because externally forced interannual variability competes with, and is comparable in magnitude to, unpredictable “internally forced” variability associated with chaotic intraseasonal oscillations. Goswami et al. conclude that to extract the predictable part of the monsoon signal on interannual time scales, systematic model biases must be reduced, and the space–time structure of summer intraseasonal oscillations must be simulated accurately.

The theme of Atlantic climate variability and predictability is discussed in detail by Hurrell et al., who focus on tropical Atlantic variability, the North Atlantic Oscillation, and the Atlantic thermohaline circulation. Improved understanding of these modes of variability is essential for assessing the likely range of future climate fluctuations and their predictability. Hurrell et al. discuss the need to build and sustain observing systems to meet the scientific challenges of the CLIVAR program in the Atlantic.

Chang et al. discuss the dynamical processes in each of the three ocean basins, with emphasis on those processes that are most relevant to the coupling of the atmosphere and oceans. In the Pacific basin, these processes describe not only El Niño but also variability in subtropical cells and in the ventilation of the thermocline. In the Atlantic Chang et al. discuss the potential impact of the thermohaline circulation on tropical Atlantic variability; in the Indian Ocean, the authors analyze the east–west mode of variability known as the Indian Ocean dipole and the impact of this mode on the climate of local land areas.

These papers collectively illustrate some of the progress made in the CLIVAR project since its inception in 1998. However, many challenges remain. One of the key challenges concerns the integration of the regional and global aspects of CLIVAR. On the one hand, observations are made locally, and the value of climate forecasts is determined by their influence on the lives of individuals. On the other hand, the tools we use to assimilate these observations and to make these climate predictions are increasingly global in domain. Regional analysis of global model datasets presents opportunities to integrate CLIVAR science on global and regional scales, for example, in assessing how the key modes of ocean–atmosphere variability, as outlined above, are influenced by anthropogenic forcing.

Indeed, the global warming problem is one area where CLIVAR is playing a key role in the science community, notably in masterminding the production of multimodel ensemble integrations for the Intergovernmental Panel on Climate Change assessments. However, another key success story for CLIVAR is bringing the sort of seasonal-to-interannual prediction studies first pioneered under TOGA into routine operations. One of the key challenges in this area is the coupling of quantitative application models (e.g., in health, hydrology, agronomy, etc.) directly to the climate forecast models. Through these types of activity, CLIVAR will be seen to have risen to its challenge, not only to advance the science of climate variability and predictability but also to ensure that this science benefits society in general.

However, despite the need to develop such application studies, it is crucially important that CLIVAR continues to emphasize the ongoing need for basic science. Climate models still suffer from substantial long-term biases, arising from uncertainty in the representation of computationally unresolved scales. This in turn leads to substantial uncertainty in forecasts of the magnitude of global warming, for example. Three conditions are necessary to improve the realism of climate models and thereby reduce the uncertainty in climate forecasts: enhanced high-performance computational facilities to allow more of the laws of physics to be applied to climate models; detailed observations of the climate system in order to formulate and test theoretical hypotheses, initialize forecasts, and assess the realism of climate model simulations; and finally scientific studies to analyze the dynamic and thermodynamic mechanisms that underlie the turbulent fluid envelope whose well being is essential for our survival. With respect to observations, particular mention can be made of progress in the study of global warming and short-term climate prediction, underpinned by CLIVAR’s successes in implementing global ocean observations.

The papers in this section stand as testimony to CLIVAR at its midterm—with its best years still to come.