top of page

African Lakes and Climate

Research in this area has been funded by:

The Leverhulme Trust TILAKA project

The Belmont forum and NERC PACMEDY project

The International Quaternary Association - PotASH project 

Main colleagues: Dr Sallie Burrough, Dr Charlie Williams, Prof. Emily Black, Prof. Simon Dadson

For further details of the PotASH project (Palaeolakes of the Arid Southern Hemisphere) please see the PotASH website.

Makgadikgadi saltpan (fieldwork 2016, photo Joy Singarayer).

Rock art from Tsodilo Hills (photo courtesy Sallie Burrough).


Robust future projections of changes in rainfall in semi-arid regions of northern and southern Africa are critical in the assessment of vulnerability to climate-related hazards. These areas are thought to be particularly susceptible to climate change given the high proportions of the population employed in rainfed agriculture. State-of-the-art climate model simulations of the next century suggest that North and East Africa are projected to experience an increase in rainfall, while southern regions may become drier. Interannual variability and the frequency of extreme events such as heavy rainfall occurrences or droughts are broadly expected to increase. Uncertainty in these projections is large and complicated by the importance of evaporation and soil moisture for predicting impacts on water availability. 

We know that there have been times in prehistory when there were also very different rainfall regimes to today in tropical and sub-tropical Africa. One particular well-documented event was the African Humid Period (AHP; roughly 14,000 to 6,000 years ago) when, due to enhanced monsoon rainfall, much of the Sahara was transformed into a region with extensive wetlands, savannahs, and large lakes, occupied by people and a variety of iconic animals, including hippos and crocodiles. But there is evidence of many such events across the whole continent occurring at various times. For example,  Lake Makgadikgadi in Botswana was at times a 'mega' lake (see right image), but is now a mostly dry saltpan.


What can the past dynamics of these highly changeable systems tell us about the dominant causal mechanisms and how sensitive they will be to future climate change? 

Sallie Burrough with the geological coring team at lake Makgadikgadi, 2016



We perform model-data comparisons between existing model outputs (e.g. PMIP3/4 simulations; TraCE-21ka transient simulation with CCSM3; HadCM3 late Quaternary simulations by Singarayer and Valdes, 2010), as well as new model simulations, with empirical data compiled in these projects in order to understand mechanisms driving the patterns of change.  

Members of these projects have been involved in developing a meta-database of published southern hemisphere lake records, and existing models for relevant lake regions as part of the Global Lakes Database. Using the database we can identify key transitions and the degree of homogeneity in hydro-climate changes. These may then be used alongside climate model experiments to understand the system. 

We also use simple hydrological models, driven by climate model output to more directly compare with paleo-lake records.


The PotASH project sought to provide a formal network of dryland palaeolake researchers working on similar questions across the Southern Hemisphere.

Change in winter surface air temperature (degrees C) due to early Holocene extensive lakes and wetlands, as modelled with HadGEM3.

bottom of page