Fitting new pieces into the African drought puzzle
Oct 1, 2008 - by Staff
Oct 1, 2008 - by Staff
About the Research
Andrea Sealy, Ph.D., is a scientist at the Caribbean Institute for Meteorology and Hydrology. She spent two years as a postdoctoral fellow in NCAR's Advanced Study Program. (Photo courtesy Andrea Sealy.)
If you want a textbook case of a climate problem with disastrous human consequences, the ongoing drought in the African Sahel is it. Nearly everyone in this region makes a living by farming or herding animals, so they depend on rainfall for their very survival. Since 1970, every year but four has been below normal in rainfall. Societal problems such as poverty and poor planning by some governments have greatly compounded people's suffering, but the drought is at the root of the increased deaths by starvation over those long years.
If we had better understanding of what disrupts rainfall across this African region, we might be able to help the local people by forecasting the disruptions. Unfortunately, we're a long way from that point. Rainfall in the Sahel is so erratic that atmospheric scientists didn't even think there was a drought during about the first decade, citing natural variability instead. Since then, they have pinpointed some contributors, but the relative weight of these factors, and particularly the way they interact, are still unclear.
Andrea Sealy has been working for the last two years to find some answers to the many questions surrounding the Sahelian drought. Sealy received a Ph.D. from Howard University in 2006. Her thesis research applied computer modeling to studies of soil moisture and precipitation in Africa. As a postdoctoral fellow at NCAR, she expanded to model interactions between dust and dynamic vegetation and how they affect precipitation—a scenario that comes a few steps closer to the complexity of the real world. Her investigative tool is the NCAR Community Atmosphere Model.
One reason to study this problem with computer models is that there are no long-term dust observations from the African continent. But African dust wafts all the way across the Atlantic on the prevailing easterlies. Sealy used a data set on dust in Barbados, which is not only the most easterly Caribbean island but her native country. "People often refer to the Barbados observations if they want to get a sense of the seasonal cycle and how the dust has varied over time," she says. She did some model runs that included the radiative effects of dust and some that did not.
Another variable in the experiment was the treatment of vegetation: some runs used the model's default, which specifies fixed states for vegetation covering the land surface. Others used a dynamic vegetation treatment. "Dynamic vegetation is more of a natural way of looking at things," Sealy explains.
Plants change seasonally and also in response to climate changes; they spread with increased precipitation and die off during droughts. Species mature and are replaced over time. There are biogeochemical connections as well, such as vegetation's role in the carbon cycle.
This graph reveals erratic swings in seasonal precipitation since 1900 for the five-month season from June through October in the African Sahel region. Note how many years show below-normal rainfall since the late 1960s. [ENLARGE] (Image courtesy Andrea Sealy and NOAA.)
The treatment of sea-surface temperatures was a third variable; Sealy used observed temperatures in some runs, while in others the SSTs were interactive, responding to changes in other parts of the model. SSTs are a major contributor to the drought, according to several scientists.
The engine powering African monsoon rains is the temperature difference between the ocean surface surrounding the continent and the land surface within. A warmer ocean means less temperature difference; as a result, moisture stays over the oceans, and Africa stays dry and dusty. So SSTs have an effect on dust, but dust also has an effect on SSTs, reducing incoming solar radiation over the oceans and keeping them cooler.
Sealy found that the dynamic vegetation treatment did affect both precipitation and dust in the model runs, reproducing reality better in such things as the timing of peak dust concentrations in Barbados. When she compared observed SSTs versus interactive ones, the effects on dust, precipitation, and dynamic vegetation also varied, in complex ways that are not yet clear.
She also compared the difference in dust during the dry period of the 1970s and 1980s in the Sahel and the most recent wetter period, in the 1950s and 1960s. With dynamic vegetation, "you see a bigger difference in dust from the wet period to the dry period," she explains. The reason for this is not completely understood, she says. "It could be that during the dry period you're getting more desert," and the greater expanse of exposed soil can become airborne more easily. "However, there are a number of other things that could be in play there, like changes in albedo and radiative forcing."
Sealy describes her work modestly as "another little piece in the age-old puzzle of what affects Sahel precipitation." Her postdoc adviser at NCAR, Natalie Mahowald (now at Cornell University), is less understated. "Her study was the first looking at these interactions in an integrative model, but it's not going to be the last. It's a very important study.”