future changes in rainfall

Endangered black rhinos using a
water hole in Kenya; from KWS website.

In a partnership with UNESCO and financial backing from Japan, Kenya very recently discovered large aquifers in its arid Northern Turkana region, a place plagued by droughts and suffering from severe water scarcity. The discovery, hailed as a "scientific triumph" that will improve access to drinking water as well as improve agriculture for "generations of people", illustrates the desperate need for water in this arid region, and its reliance on consistent sources of drinking and irrigation water. A UNESCO statement highlighted the importance of such a discovery in the face of Kenya's water vulnerability, emphasizing that Kenya's water scarcity will likely increase precipitation with climate change.  

Children gathering water in Kenya;
such water sources will likely decrease
in frequency in the future.

Such examples highlight how critical water is in East Africa in general and Kenya specifically, particularly given future changes in water availability with climate change. These changes in precipitation patterns will likely have strong ecological and economic effects. Such effects include accelerated water cycling, increased drought, and the cascading effects on both wild plants and animals, as well as crops and livestock. These changes are of paramount importance not only because they will have strong effects on the ecology, farming, and livestock of the region, but because dryland tropical economies are strongly driven by rainfall. Subsequent posts to this blog will explore specific aspects of how and why future changes in precipitation will (or will not) affect economics and livelihoods in the the arid tropics, using Kenya as an example, but first it is essential to understand the predictions for the region.

Photo from ILRI website

Climate change will affect two important aspects of future precipitation: total amount of rainfall and variability in rainfall (or, when rainfall falls during the year, including time between rainfall events and rainfall per event). While we have a poor understanding of how total rainfall will change, most models predict an increase in the variability of rainfall over the tropics. An increase in variability means that rainfall events will likely occur less frequently, but more rainfall will fall per event. 

Total annual rainfall in the tropics will likely decrease in dry areas and increase in wet areas (see the figure below, with lower precipitation over the African Sahara but higher precipitation over the South American wet tropics), meaning that dry areas will get drier and wet areas will get wetter; trends since 1980 and from as early as 1930 show that this has already begun to happen. This pattern (dry-gets-drier and wet-gets-wetter) likely arises because of changes in temperature; tropical regions that get warmer should get wetter, and those that cool should get drier. Tropical regions differ substantially in how much they will warm: regions may differ by 1.5 degrees C. These differences in amount of warming will lead to large differences among tropical regions in the amount of rainfall that falls. It is particularly challenging to predict changes in total precipitation over the ocean, but it is relatively easy to predict changes over land.

Predictions of changes in precipitation from ©2007 IPCC WG1 AR-4. 

The vast majority of climate change models predict a substantial increase in precipitation variability (again, increased time between rainfall events but more rainfall per event). Increased variability in precipitation is largely a product of increased evaporation, due to an accelerated water cycle. Variability will increase even in regions that get drier; in regions that get wetter, variability in precipitation will increase even more. 

Though we know that the variability of precipitation will increase, the frequency of precipitation extremes (events that are are very extreme and very rare, in the highest 99.9%), is difficult to predict for tropical regions, with predictions varying widely across models. Discrepancies among model predictions likely arise because these models are very sensitive to how they incorporate air circulation. More specifically, these models are sensitive to how they incorporate vertical air motion. We have a poor sense of how air moves vertically in nature (because it requires detailed and expensive-to-obtain satellite data), and thus, different models of climate dynamics treat vertical air motion differently.

These models try to predict global-scale changes in precipitation patterns, and do a fairly good job, but predictions for specific regions are more difficult to make.

Kenya generally has two rainy seasons, with long rains occurring in March to June and short rains from September to December. Even without climate change, Kenya's rainfall is quite variable, both in time and space. Years differ greatly in the duration and timing of the rainy seasons and the amount of rain that falls. Rainfall also varies dramatically over small regions; for example, annual rainfall can vary by as much as 600-1600 mm annually within one rainfall basin.

Different models of climate change predict either increases or decreases in total annual rainfall in Kenya, but most models predict increased variability in precipitation. While a recent IPCC report predicts increases in rainfall in East Africa generally and Kenya specifically, other reports incorporating more data suggest that rainfall will actually decrease over much of East Africa, including Kenya, likely due to warming over the Indian Ocean. Drying has already been observed in Ethoipia, with drying usually occuring during the long rains season, rather than the short rain season.

Climate models predict that variability in precipitation will increase in Kenya, meaning that rainfall events will become more intense and less frequent in Kenya. This has already been happening over much of Kenya; see the figure below, where total rainfall has not changed over time, but the rainfall per storm has increased, and rain events have become less frequent. 

Graph from Franz et al. 2010, showing no significant change in total rainfall (top row), but higher variability (increased rain per storm, second row, and longer time between storms, third row). 

While changes in mean annual precipitation and precipitation variability may be mitigated by discoveries of water or other scientific advances, Kenya cannot count on such advances in its preparation for the future. In a region characterized by water scarcity, whose economies and ecology are driven by erratic variation in rainfall, it is important to understand precipitation's effects on the ecology, economics, and livelihoods in the region. Subsequent posts will illustrate specific examples of the putative effects of changes in future precipitation on economic and ecological phenomena.