The early to mid-Holocene in North Africa is a fascinating example of a fundamental natural environmental change. During this time much of the present-day Sahara was transformed by an expansion of vegetation and lakes. This was caused by a huge increase in precipitation ultimately driven by a change in Earth's orbit. However, simulations with a range of climate models have systematically failed to reproduce adequate precipitation response for the mid-Holocene. 

In this talk I will cover new simulations with the Hadley Centre models aimed at understanding the 'Green' Sahara.  I'll begin with the potential role of mineral dust, which we quantify using three different dust model versions of the HadGEM2 Earth System model. Then I'll switch to a set of ensembles with HadAM3 which we have used to optimise convection parameters, leading to a version of the Hadley Centre that does satisfy the mid-Holocene precipitation reconstructions over North Africa for the first time with this model. This model still doesn't 'Green', and I'll discuss work to improve the vegetation moisture stress in the TRIFFID dynamic vegetation model. I'll briefly discuss how a comparison of two ensembles, the mid-Holocene and a 'future-like', in which CO2 is doubled, may provide some insight into the links between skilful simulation of the 'Green' Sahara and future climate.