The dramatic melting pattern of African mountain glaciers

The seasonal snow melt of mountain glaciers, like those from Mount Kilimanjaro and Mount Kenya, is a key supply of water to out flowing rivers of mountain ranges during warm and dry periods. These glaciers act as a water tower, and several rivers have now become seasonal.

One case study that has gotten large-scale attention is the gradual, yet drastic retreat of glaciers on Mount Kilimanjaro. The thinning and lateral retreat of these ice-caps has caused the 85% loss since 1912. Some studies have assigned this to be a cause of climate change. The rate of glacial decline has been increasing since 1970 (start of industrial revolution and pronounced human impact on the climate system) and can be explained by increased dryness, reduced moisture convergence over East Africa and thus lack of snowfall on the mountain to replenish the seasonal melting (e.g. Ward, 2012). In fact, most of the ice is shown to sublimate to the atmosphere directly. These climatic drivers of glacial retreat are expected to increase further with expected climate change scenarios in the future.

However, the very start of the retreat predates the start of anthropogenic impact on the climate by several decades. Opinions of scientists have thus been divided on the exact cause of Mount Kilimanjaro glacial retreat. This discussion is only ongoing as neither side can be proved completely wrong due to the lack of high altitude data in East Africa. It is making it hard to find a baseline against which to analyse changes of temperature and dryness and relate those to historical trends with forcing factors.

The recent examination of ice cores from the glacier itself shines a new light on the issue. A 30mm thick layer of dust (indicating prolonged dry period in regional climate) ≈4,200 years ago has not been accompanied by any decline in the glacial thickness, as would have been expected if dryness and decreased precipitation really would be the dominant forcing for Kilimanjaro glacial retreat. The core further shows that the recent melting pattern is unique within the whole 49m core, supporting that recent earth-system changes (by human impact!) must be pushing the ice-cap shrinking.

The pattern seen in this case study is not a singular occurrence across Africa: the melting of Mount Kenya, as well as in the Rwezori mountains, illustrate the common pattern of “glacier mass loss, shrinkage, and retreat at high elevations (>5,000 m above sea level) in lower latitudes (30° N to 30° S), particularly in the thermally homogeneous tropics”(Thompson et al, 2009). Coming back to the discussion on the main drivers of this change, this uniform pattern suggests a common underling driver upon which the local influences (e.g. land use change) may be superimposed to accelerate it.

The impact of climate change induced sea-level rise and storm surges upon coastal groundwater reserves

The African continent lies between the Atlantic and Pacific/Indian ocean – and thus this post discusses how changes in the level of these oceans may affect coastal freshwater resources. It is now generally observed (mean global rate of rise 1993-2009 was 3.3 ± 0.4 mm/year) and predicted that under any of the future emission scenarios we will experience a significant rise in global temperatures, causing the melting of polar and glacial ice and warming the world’s oceans – all contributing to the rise in global sea levels. The magnitude of this sea-level rise is still uncertain, as are the regional differences to be expected (Nicholls and Cazenave, 2010). To give a rough estimate, IPCC AR4 projections show that future ice dynamics discharge could produce about a rise of 80cm by 2100. On top of this, local changes such as land subsidence (often exacerbated by human activities in densely populated areas) will further allow the impact of salt water upon the coastal water.

As I have shown in previous blogs, a large fraction of Africa’s low-income population is strictly reliant on groundwater for its freshwater supply. These aquifers can be relatively thin in the low-lands and form adjacent to the coastline. The Ghyben-Herzberg function explains the existence of a freshwater lense “floating” on top of seawater, separated by the saltwater-freshwater interface (for 1 feet of freshwater above the interface, 40 feet of saltwater lie below).  This interface is vital in determining the quantity of freshwater storage in the coastal aquifer, and the mixing of the layers effects the quality. Many studies have evaluated the pressures existing on these vulnerable freshwater reserves, leading to sea water intrusion (the encroachment of the interface) and excessive chloride contamination of the freshwater.

While the unsustainable abstraction for human use has been evaluated to be a greater threat in many coastal aquifers than climate change (e.g. Ferguson and Gleeson, 2012), I want to highlight the ways climate change may exacerbate this.

The diagram below shows how a rise in sea-level affects a coastal aquifers’ characteristic by reducing the freshwater lens (Figure 1). An area identified as being under significant pressures from this mechanism is Southern Africa. A modelling study was conducted by Ranjan et al (2006) that estimated the loss of freshwater by the encroachment of the interface – also taking into account predicted changes in aquifer recharge. The rate of change of groundwater loss (% per year until 2100) in South African coastal aquifers was named to be 0.027 (A2 scenario) – 0.022 (B2 scenario).

Figure 1: click for Source

Besides this reduction in reserves, climate change also is predicted to increase cyclonic activity and storm events. Storm surges cause coastal flooding, salinisation of surface waters and ecosystems but also will result in instantaneous events of saltwater intrusion to coastal aquifers from above into the freshwater lens (Anderson, 2002). Groundwater chemistry is thus adversely affected by increases in chloride concentrations, adding to the stress on coastal water reserves.

In a comparative impact study of the climate change effects described above (sea-level rise and storm surges), Dasgupta et al (2009) from the World Bank Development Reasearch group, find what I had expected while examining the impacts on Africa. The vulnerability to these threats is found to be concentrated to populations and large cities at the lower end of the international income distribution (for Africa e.g. Djibouti, Tanzania, Mozambique). Once again, this raises the point that the most socio-economically vulnerable are those to expect the largest negative impacts of anthropogenic climate change, while having contributed least to it. A very recent publication by Barbier (2015) effectively highlights the threat climate change poses to African low-elevation coastal zone populations by increasingly pushing these into the ‘poverty-environmental trap’ and calls for a rethinking of policy and development strategies.

Surface crusting improves water availability despite reductions in annual rainfall in semiarid Africa

Today I want to focus on a specific paper by Favreau et al (2009) that explores how land-cover change and climate change affect water resources in conjunction in semi-arid Africa. 
I found this to be extremely interesting, as findings here suggest that a change in the land use and resulting changes in the soil cover, can eliminate the pressures on water resources expected under decreased monsoonal rainfall and higher temperatures.
The area under study (southwest Niger) underwent a large scale land-cover change from natural savannah to millet crops - which dramatically altered the local water balance. Surface crusting of the soil meant that there was an increased occurrence of gullys and ponds in the landscape, fed by in increase in Hortonian overland flow on slopes. These ponds are points of focussed groundwater recharge to the underlying aquifer. A physically based, distributed hydrological model quantified the change: while the adverse effects of climate change had reduced runoff by 2 fold, the change in land cover had increased it by 3 fold, resulting in a net increase. A different modelling approach on the wider regional area analysed groundwater table fluctuations and related the evolution of land use to an increase in recharge from 2mm/yr to 25mm/yr (+/- 7mm). 

This "Sahelian paradox" highlights that issues of water in Africa need to be examined on the ground - and that a predicted decrease in rainfall (here 23% reduction in mean annual rainfall, 1970 – 1998 versus 1950– 1969) does not equal to a reduction in water availability. In fact, human alterations of the natural systems (climate and land cover) here increase the water resources regionally.

Groundwater as beneficiary of anthropogenic climate change?

The latest posts have outlined that the greatest challenge to be faced by Africa is not necessarily the decrease in the total mean annual rainfall under climate change, but rather the increased variability in rainfall events and thus the distribution within the year. Having established that the general trend of precipitation is towards fewer, more intense rainfall events, I want to dedicate this post to explaining how under certain conditions this can improve the water resources available in Africa.
The dependence on groundwater across Africa varies but serves as the only reliable perennial source of water for many rural communities. The overall dependence is estimated to be as high as 50%.

Groundwater recharge is strongly determined by the effective precipitation in a given year – however, as evidence from East Africa suggests, it is a non-linear relationship, in which recharge is more strongly related to extreme rainfall events than to the totals (Owor et al. 2009 and Taylor et. al., 2013). While these recent studies examine a past data records, there are important lessons to be learnt for future adaptation to an intensified hydrological cycle.

I want to provide an overview of the findings first:

1. The Upper Nile basin is located within the tropics of Africa, a region where climate change effects are expected to be especially pronounced. A data record of 1999-2008, pairing groundwater to rainfall data, is used to show that the magnitude of water level fluctuation is better correlated to the number of heavy rainfall events (>10mm/day) than to the total sum of rainfall. (Owor et. al. 2009)
2. A 55-year long record from a different basin in East Africa, in central Tanzania, provides further coherent evidence for this phenomenon. The observed recharge events interrupt long periods of groundwater level decline, and are highly correlated with extreme rainfall events occurring during the summer monsoon. In fact, the highest recharge events occur during ENSO years which brings intense rainfall to Tanzania (and thus result from regionally dominant modes of tropical climate variability) as shown in Figure 1. The evidence further shows that 80% of cumulative recharge over the whole observed timespan is derived from the highest 13 recharge events. (Taylor et. al., 2013)

Figure 1. Showing El Nino and La Nina events alongside the main recharge events (Source)

My thoughts regarding future climate change:

The intensification of the hydrological cycle through climate change (last 2 posts) will provide an increase in the occurrence of extreme events, which are those that episodic recharge events depend on in these two areas. Groundwater reserves are already seen to become the freshwater source of choice in the future due to their resilience in quality and quantity relative to surface waters (for a great review see Kundzewicz and Doell, 2009). 

The findings in the studies above suggest that ground water resource replenishment can in fact benefit from this new turn in the global climate. While there is significant uncertainty in the prediction of ENSO events through GCMs, the probability of positive IOD modes is suggested to increase under anthropogenic global warming, bringing higher monsoonal rainfall to East Africa. The enhanced groundwater reserves have the potential to act as a buffer against the increased variability in surface water resources, and could be the solution to water security in part of Africa that experience a similar recharge response to extreme events. When looking at the geographical distribution, using the brilliant maps of MacDonald et al (2012) for example, we can see that most parts of Africa have the potential to develop groundwater, with much being at low depths from the surface. This means that in these low-lands even poorest farmers will be able to use this relatively stable water supply with limited technology (and economic input!) to support their livelihoods by elevating the insecurity of using rain-fed agriculture  in future climate change scenarios.

We cannot however consider this as a solution that fits all – it is strongly dependent on geological properties of the soil surface and subsurface, and the way other effects of the expected climate change (e.g. sea level rise) influence in situ.

African Voices on Climate Change: outlining climate change impacts upon Ethiopia

The video shows Fassil Kebede of the Ethiopian Environment and Forest Research Institute, explaining some ways in which climate change is already impacting his country. Few of the points touched on are closely linked with water. One of them is the increased occurence of a pair of extremes: drought and flooding events. Another problem, faced in Ethiopian low-lands especially, is the reduced growing-season (by the intensification of seasonal precipitation to erratic, less predictable rains).

Clausius-Clapeyron giving hope

We’ve seen that the Earth is getting warmer – and extremely so on the African continent. The evapotranspiration robs moisture. To assess impacts on the continents 'blue water' formation, however, we must take into account a physical relation that affects a different part of the hydrological cycle.

The Clausius-Clapeyron relation describes the relationship temperature has with the atmosphere’s potential to hold moisture. With rising temperatures, the atmosphere can hold more moisture (see Figure 1), leading to longer moisture retention before a rainfall event (longer drought events or dry seasons) and more intense rainfall once it does occur. This intensification of the hydrological cycle is more extreme in warmer regions (such as those semi-arid/arid regions of Africa) due to the Clausius-Clapeyron relation progressing in an exponential manner. Reviewing a wide range of climate modelling studies, a 2-3 degree rise in global temperatures will intensify the water cycle by 16-24%.

Figure 1. Clausius-Clapeyron relation, showing that the saturated water vapor pressure increased exponentially with increasing temperatures. (source)

The expected in-situ effect of this change on precipitation is different across Africa, especially as changes in global circulation further complicates prediction. Additionally, modelling of rainfall is severely limited in certainty by the lack of continuous observational data across the continent (Niang et al., 2014).

The intensification of rainfall can have beneficial consequences for the formation of 'blue water', by helping to maintain the ephemeral surplus in moisture despite a rise in evapotranspiration by higher temperatures. I want to relate back to my post 2 (which assumed no change in precipitation): the expected rainfall pattern intensification will likely result in the increase of moisture supply in a shorter period of the wet season - where the threshold (P-ET) is surpassed - and blue water resources can be formed. This is obviously a theoretical perspective and in reality it is the subtleties of the relative changes in ET and P at a location that will determine climate change's impact on 'blue water' formation.  

Thoughts on the implications on the ground:

While the potential for 'blue water' formation by this intensification of rainfall is clearly possible, the benefit of this to local populations and environments depends on the potential of water storage and retention to last through the extended dry seasons (that are also an effect of the intensification). The adverse effects of extended droughts and higher time-variable availability of water resources may often not be relieved by the 'hope' for this source of excess moisture. Even given that yearly sum 'blue water' does not change, the amplified variability of supply poses threats especially to subsistence households, farmers that live off rain-fed agriculture and highly adapted ecosystems. 

From causing disastrous floods (guardian.com) to dramatically improving groundwater storage (Taylor et al., 2013) everything is possible, depending on catchment characteristics, its geophysical properties and land-use. 

Wide spread evidence for the heating

In my last post I hoped to show the mechanism with which a rise in temperatures by athropogenic climate change impacts the potential for blue water formation. In this post I want to provide a brief summary of the up-to-date evidence concerning African temperatures.

Looking at the continent as a whole, we have already seen an increase in temperatures in recent decades that is significantly higher than what was expected under natural variability. Due to the likely anthropogenic impact on the climate, it is important to consult models that integrate 'Representative Concentration Pathways’ (emission scenarios) when predicting future temperatures.

Africa is a key warming region - expected to have a faster rise in temperatures over the 21st century than the global mean projection. The figure below (Niang et al, 2014) shows the temperature increases in different African regions , expected under the RCP8.5 pathway(exceeding 4 C over most land areas!).

 

Using the CIMP5 global climate model ensemble, Diffenbaugh and Giorgi (2012) examined climate-change "hot spots",  and identified the Sahel and tropical West Africa as such. Other studies focusing on particular semi-arid regions of Africa (e.g. Ethiopia by Conwayand Schipper, 2011) highlight the likelyhood of large temperature increases during all seasons.

How does this link back to water?

Generally, higher temperatures will raise the proportion of freshwater lost to evapotranspiration, and reduce the physical water resource availability from annual precipitation. The evidence I provided shows that all of Africa will be experiencing a water loss to evapotranspiration under the future scenarios. Relating back to my diagrams in the last post - this is especially dramatic if the renewable freshwater formation depends on the seasonal rain pulse (as in the semi-arid regions). And indeed, the regions with the projected greatest warming year-round ("Hot spots") are those that have a precipitation scheme similar to what was represented in my schematic diagrams. Therefore we expect these (already highly adapted, vulnerable) environments, and populations, to encounter the greatest pressures on blue water resources (alone from temperatures and not yet taking precipitation projections into account!).