The water cycle of the mid‐Holocene West African monsoon: The role of vegetation and dust emission changes

Abstract

During the mid-Holocene (6 kyr BP), West Africa experienced a much stronger and geographically extensive monsoon than in the present day. Changes in orbital forcing, vegetation and dust emissions from the Sahara have been identified as key factors driving this intensification. Here, we analyse how the timing, origin and convergence of moisture fluxes contributing to the monsoonal precipitation change under a range of scenarios: orbital forcing only; orbital and vegetation forcings (Green Sahara); orbital, vegetation and dust forcings (Green Sahara-reduced dust). We further compare our results to a range of reconstructions of mid-Holocene precipitation from palaeoclimate archives. In our simulations, the greening of the Sahara leads to a cyclonic water vapour flux anomaly over North Africa with an anomalous westerly flow bringing large amounts of moisture into the Sahel from the Atlantic Ocean. Changes in atmospheric dust under a vegetated Sahara shift the anomalous moisture advection pattern northwards, increasing both moisture convergence and precipitation recycling over the northern Sahel and Sahara and the associated precipitation during the boreal summer. During this season, under both the Green Sahara and Green Sahara-reduced dust scenarios, local recycling in the Saharan domain exceeds that of the Sahel. This points to local recycling as an important factor modulating vegetationprecipitation feedbacks and the impact of Saharan dust emissions. Our results also show that temperature and evapotranspiration over the Sahara in the mid-Holocene are close to Sahelian pre-industrial values. This suggests that pollen-based paleoclimate reconstructions of precipitation during the Green Sahara period are likely not biased by possible large evapotranspiration changes in the region.

Modulation of Mid‐Holocene African Rainfall by Dust Aerosol Direct and Indirect Effects

“…Using a model that only resolves direct aerosol effects, Pausata et al (2016) accounted for changes in MH vegetation cover and the associated dust reduction and found that a reduced dust load enhanced local rainfall by as much as 2.5 mm/day over the Sahara and shifted the West African Monsoon (WAM) northward. With the same model, Messori et al (2018) showed that increased moisture recycling was a key driver of the rainfall response to reduced atmospheric dust load. However, the potential role of indirect aerosol effects, which have at least equal global radiative forcing (Myhre et al, 2013), on AHP precipitation, have never been explored.…”

“…To date, most model simulations, including those from Paleoclimate Modelling Intercomparison Project Phase 3, specify preindustrial (PI) dust loading (Braconnot et al, 2011;Taylor et al, 2012). With the same model, Messori et al (2018) showed that increased moisture recycling was a key driver of the rainfall response to reduced atmospheric dust load. With the same model, Messori et al (2018) showed that increased moisture recycling was a key driver of the rainfall response to reduced atmospheric dust load.…”

Technical Note: Characterising and comparing different palaeoclimates with dynamical systems theory

“…Kutzbach et al, 1996;Kutzbach and Liu, 1997;Claussen and Gayler, 1997). More recent modelling efforts have confirmed this, and have further highlighted the potential role of an incorrect representation of atmospheric aerosols in favouring the dry bias (Pausata et al, 2016;Gaetani et al, 2017;Messori et al, 2019). Such hypothesis has triggered a lively discussion in the literature (cf.…”

Development and testing of scenarios for implementing Holocene LULC in Earth System Model Experiments

“…and could be implemented as an additional sensitivity experiment when available. Changes in atmospheric dust loading are not included in the past1000 simulation but are important during the earlier part of the Holocene (Pausata et al, 2016;Tierney et al, 2017;Messori et al, 2019). Although continuous reconstructions of dust loading through the Holocene are not available, it would be possible to use estimates for particular time-slices (Egerer et al, 2018) to test the sensitivity to this forcing.…”