Welcome to CGD

expand collapse

CGD Climate Highlights: Past Climate Variability

What do we know about past climate variability?

What does the Holocene tell us about Sahel rainfall and its variability? How can we use what we find to help us better understand future climate change?

The monsoons are of great social and economic importance to large populations. Monsoons are seasonally varying climate features (the origin of the word monsoon comes from the Arabic word for season) and are located in regions of the major continents at low latitudes. They are characterized by a wet season, when the low level winds generally blow from the ocean onto land, and a dry season during which the wind direction reverses. These seasonal variations are strongly tied to seasonal changes in the amount of sunlight or insolation (contracted from incoming solar radiation). Major monsoon regions include India, Africa, southeast Asia and Australia. Local agriculture provides the food for many of the people inhabiting these regions. The strength of the monsoon, as measured by the amount of precipitation that occurs during the wet season, and the timing of precipitation during the wet season are extremely important in determining the success (or failure) of crop production.

In addition to the variations that occur in monsoons within a season and from one year to the next, observational records and paleo-proxy records show that monsoons vary on time scales of centuries, and millennia. One region where good proxy data documenting monsoon precipitation exists is over northern Africa, including the Sahel. Independent reconstructions based upon records of pollen and lake status indicate that much of northern Africa was wetter than it is presently. These differences were great enough so that regions that are currently desert supported vegetation in the past, resulting in the greening of the Sahel.

Roughly 6000 years ago during the mid-Holocene, the earth's orbital configuration and tilt relative to the sun were different than they are now. These differences resulted in considerable changes to how the solar insolation was distributed as a function of latitude during the year. This can be seen in fig. 1. The strongest positive anomalies occur in the NH during summer and early fall and in the SH during fall and early winter. The strongest negative anomalies are during late winter in the NH and late summer in the SH. These changes act to amplify the NH seasonal cycle of solar insolation and diminish the SH seasonal cycle.

Simulations of past climates provide a way to measure climate model performance over a much wider range of forcing and response than is possible using only the modern period. Many different numerical models have had some success in simulating changes in both northern African monsoon strength and location during the mid-Holocene making this a robust feature of modern climate models. The model simulation and proxy record both show enhanced precipitation southward of 20o N but the model simulation has no increase northward whereas the proxy record shows enhanced precipitation extending to the Mediterranean coast (see Fig. 2). One reason is that the models do not include an important positive feedback between the northward extension of the monsoon and northward extension of vegetation.

There are other questions being addressed using simulations of the northern African summer monsoon using the climate models at NCAR. For example, how do changes in the sea-surface temperature predicted by the coupled climate system model affect the monsoon? This question is important because sea-surface temperature changes have been implicated in a decrease in the summer time Sahel rainfall during the periods between roughly 1950-1980 and 1980 to present. Also, what are the mechanisms responsible for the stronger and more northward positioned monsoon in climate simulations of the Holocene. It is well understood that the increase in solar insolation results in an increase in the land ocean temperature difference that drives the monsoon. But the processes and the evolution of the processes during the monsoon season are not well understood.

The performance of climate models is commonly evaluated over the recent past - roughly the last 150 years or so when the changes in the forcing and corresponding changes in climate are relatively small compared to the changes predicted in the coming decades and centuries. Much larger changes in climate forcing occurred somewhat further back in the earth's history. Understanding past climate can help us answer questions such as: How well do computer models used to predict future climate change perform? Where are their strengths and weaknesses? How much confidence should we have in the answers they provide? These are important questions to the scientist developing the models, to the scientist using the models to predict future climate and to the political leaders who use the predictions to aid in formulating policy.