I am studying precipitation patterns in the central Sahel. These seem to depend strongly on two temperature gradients, which influence the relative strength of the northern and southern Hadley circulation: to the north, the temperature gradient between the eastern Mediterranean and the southern Sahara (large gradient contributes to strong wind, keeping the southern Hadley cell and thus the rain at bay), and to the south, the temperature gradient between the coast of Nigeria/Cameroon (from where a lot of moisture is transported inland between rows of mountains) and the southern Sahara.

The problem is that more rainfall in the central Sahel (i take this as 12-15N, 0-20E) contributes to a reduction in average Tmax in the entire area, including the southern Sahel. There thus seems to be a complex feedback loop: high Sahara temperatures relative to the Mediterranean contribute to strong northerly winds, contributing to low rainfall, thus sustaining high Sahara temperatures. At the same time, high Sahara temperatures relative to the Atlantic coast contribute to strong southerly winds, which bring in more moisture, and should reduce Sahel temperatures.

When I detrend P against the trend in temperature gradients, I get the graph for JA (data for P and Tmax from CRU) as attached. This is a very strong correlation (r=0.85) with both multidecadal and interannual variability, but I am not sure what to do with the positive feedback loop between the northern temperature gradient and Tmax in the southern Sahara, and the negative feedback loop between the southern temperature gradient and Tmax in the southern Sahara.

If I only take the temperature gradient between the Atlantic Coast and the Mediterranean Coast, r stays around 0.7.

Note that I calculated correlations for the period 1950-2018, as measurements during World War 2 were very limited in the Mediterranean. The graph is drawn for the period 1920-2018, as a reasonable amount of rainfall data is available from around 1920 onwards (even though CRU data goes back to 1900).

Kind regards,

Timmo Gaasbeek