The Sahel is a particularly sensitive region with respect to the effects of global warming. The Sahel has been plagued by decadal-ish periods of drought, such as in the 1980’s. It’s generally understood that the variations in Sahel rainfall are mostly attributable to ocean variability (Giannini et al. 2003), but there are still many questions about the importance of other natural and anthropogenic factors. These questions are hard to answer with current climate models, because ocean temperature biases can lead to large differences in the distribution of rainfall. In spite of the lack of agreement between models, this paper was able to identify a robust result across a group of models from the CMIP3 data archive.
First, let’s familiarize ourselves with the region. Although there are no strict boundaries of the Sahel, it resides between the “hyper-arid” Sahara to the north and the relatively wet equatorial region to the south. In May-Jun, before the main monsoon season, the southern hemisphere Atlantic trade winds change direction and bring wet air into West Africa from the South. This is often referred to as a “monsoon jump” (Hagos and Cook 2007).
The Harmattan also push dry, dusty air down from the north. The Harmattan winds are sometimes referred to as “trade winds“, but are also associated with a area of low pressure in northwest African known as the “Saharan heat low“.
Unfortunately, future projections of rainfall over the Sahel are mixed, and the sign of the change is very uncertain.
“… summer rainfall is predicted either to decrease or increase by up to 20% depending which model is used.”
So, it is refreshing that this paper found a consistent result among model projections.
“Here we show that a robust Sahel response to GHG forcing is a seasonal redistribution of rainfall.”
Using the CMIP3 model archive they show that models are inconsistent in their response during the peak of the monsoon. This isn’t too surprising given how erratic convective parameterizations can be, aside from the erratic nature of convective thunderstorms themselves. What they do find is that the monsoon jump appears to occur later on average.
“Across the models, rainfall anomalies are predominantly negative at the beginning of the rainy season (May and June), but positive at its end (October), indicating a delay of the main rainy season.”
This result is shown as the blue line in Figure 1b (below), which indicates robust negative rainfall anomalies prior to the monsoon season, and positive anomalies at the end of the season.
There is still a lot of model-to-model differences in the length of this delay of the rainy season, but the overall result is robust.
One interesting complication of this is that some crops may experience a shorter growing season and reduced yields because their production is limited by the length of day. Another interesting consequence is that evaporation generally increases (red line in Fig. 1b). The net result is that precipitation minus evaporation is negative, suggesting the soil might become drier on average.
The authors also find evidence for a shorter rainy season, but the model-to-model variability makes this result less robust.
The authors conclude that the delay of Sahel rainfall is likely a “regional manifestation of the global response” to rising CO2. They mention a few ways that this localized feature might be influenced remotely. The most interesting explanation has to do with Arctic sea ice loss. Many studies have shown that the latitudinal position of the intense rainfall associated with the ITCZ is influenced by temperatures at higher latitudes. Colder temperatures in the Northern Atlantic should result in a southward shift in the ITCZ. We expect delayed summertime warming of Northern Atlantic surface temperature because less sea ice leaves more open water, which has a higher heat capacity and causes the surface temperature to lag changes of incident solar radiation. This should also cause the ITCZ to stay in the southern hemisphere a bit longer, which would delay the start of the monsoon. However, it is worth noting that similar mechanisms could be at work in the Southern Hemisphere, and teasing out how these mechanisms balance out is tricky.
The authors also note that changes to the land surface evaporation in the Sahel could delay the onset of precipitation. However, we should be able to identify the start of the monsoon by an increase in northward moisture flux, so I doubt decreased evaporation could have that big of an impact.