The earth's climate system includes processes that cause variability at different temporal and spatial scales. Some processes are local and act in the short or immediate term (a few days) and cause the variability of "weather". Other processes are affected by the interaction of the atmosphere with the oceans and the land surface. Some of these processes result in variations of climate at the scale of months to seasons (e.g., those processes affected by El Niño). Still other phenomena depend on natural and anthropogenic factors that affect the chemical composition of the atmosphere and cause variability of the climate at the scale of several decades to centuries. The latter includes the variability of climate that is commonly referred to as "climate change".

All of these processes act simultaneously and result in the observed earth's total climate variability. The magnitude of climate variability at these temporal scales is different and the relative magnitude also varies among regions of the world. An example of the relative magnitude of climate variability at different time scales is shown in Figure 1. The figure was constructed by partitioning the total variability observed in annual rainfall in the Sahel for the period 1900-2006. Panel (a) shows the rainfall variability at the longterm (linear trend in the last 100 years, which is a crude representation of the man-made climate change signal), the scale that is usually called "climate change". The second panel (b) shows the variations of rainfall measured at the decadal scale (after removing the linear trend), and reveals decades when rainfall tended to be above average (e.g., the 1950’s and the early 1960’s) and decades when rainfall tended to be below average (e.g., the 1970’s and 1980’s). Finally, panel (c) shows the variability of rainfall in the year-toyear time scale that remains after removing the linear and the decadal trends. The figure shows the relative magnitude of the rainfall variability at these three temporal scales as measured by the percent of the total variance explained by each temporal scale. The proportion of total variance explained by the short-term (interannual) variability is 3 times greater than the corresponding to the long-term variability (“climate change”), and 2 times greater than that of the decadal variability.

The slow and persistent forcing of increasing GHGs, is producing noticeable changes in the mean climate upon which shorter-term variability is superimposed. Increasing GHGs may also change the magnitude of the short-term variability, for example by enhancing the strength of the hydrological cycle. Changes in the mean state, the variability, or both, will alter the statistical distribution of climate and weather, and will likely result in more frequent extreme events that can have devastating socioeconomic and environmental impacts. Consequently, an effective manner for assisting societies to be better prepared and adapt to possible climate change scenarios is by assisting them to cope better with current climate variability.

Thus, a possible approach to introduce the issue of "adaptation to climate change" into the policy and development agendas is to consider the longer-term variations ("climate change") as part of the continuum of the total climate variability, from seasons to decades to centuries, and generate information at the temporal scale that is relevant and applicable for the particular time frames or planning horizons of the different decisions. This approach allows considering "climate change" as a problem of the present (as opposed to a problem of the future) and aims to inform the decision-making, planning and policymaking processes, in order to reduce current and potential future vulnerabilities to climate variability and change.

One of the key premises of this approach to engage in adaptation to climate change is that improving year-to-year planning activities and decisions lead to societies that are better adapted to longer term climate change. However, there are situations in the different socioeconomic sectors where important issues require fundamentally different approaches and activities. Thus, several important decisions need information and climate projections at temporal scales of 10-30 years (e.g., transportation infrastructure projects, water reservoir design, long-term business plans, problems related to sea level changes or to water supply from glaciers, etc.). Therefore, focus is also needed on climate risk management work for adaptation to "near-term" climate change, i.e., 10-30 years. This "decadal climate variability" is still posing important scientific challenges and the climate science community is investing huge efforts in exploring ways to improve the ability to predict it. In the meantime much can be gained by interpreting and characterizing the decadal trends in the observed historic records and on methods for producing seasonal forecasts under a changing climatic baseline (as opposed to the "static" baseline).

An advantage of this approach is that it provides immediate assistance to the public and private sector: while it helps stakeholders to confront possible future climate scenarios, it identifies immediate actions needed to manage the climate variability that is currently affecting societies. Furthermore, the impacts of the taken actions and interventions are also evident and verifiable in the short term making them more attractive to policy makers and decision makers. Research organizations such as the IRI (International Research Institute for Climate and Society) are focusing on this approach and labeling it "Climate Risk Management".

Figure 1: Partition of the total observed rainfall variability in the Sahel. Rainfall is expressed as anomalies (i.e., deviations from the mean annual rainfall of 1900-2006). (a) long-term variability (linear trend), (b) decadal variability (after removing the linear trend), (c) inter-annual variability (after removing the linear and decadal trends).