This was already observed in the past, where discharge is considerably larger in wet years than in dry years and the model simulations are well in line with this observation (see Fig. 8). Under such conditions any projections with climate models have to be interpreted with caution – only small variations (increases/decreases) in precipitation projections cause large differences in the impact on discharge. This was also confirmed by the sensitivity tests (see Table 5 and Fig.
10, bottom) – where a decrease of precipitation by −10% caused a decrease in discharge selleck inhibitor by almost −850 m3/s, or −32%. Note that this high sensitivity of discharge to precipitation contrasts the conclusions of Beck and Bernauer (2011) that climate has relatively small effects on water availability in the Zambezi basin, which may be related to their approach of calibration to long-term average conditions. Our simulations under climate change scenarios show a range of −14% to +10% for mean annual Zambezi discharge at Tete in the near
future (2021–2050 as compared to Baseline simulation 1961–1990). These results (and the large uncertainty) have to be interpreted within the context of the results of previous studies. Harrison and Whittington (2002) focussed on the upper selleck Zambezi River at Victoria Falls. For the 2080s their three climate scenarios show a warming of about +5 °C and a reduction in rainfall between −2% and −18%, which results in a reduction in runoff by −10% to −36%. In a preliminary analysis the World Bank (2010) used GCM data (A1B emission scenario) for the whole
Zambezi region. For 2030 they estimate a change in runoff between −13% and −34% (depending on the sub-region). Beilfuss (2012) summarized existing climate change assessments for the Zambezi and concludes that by 2050 runoff is likely to decrease by −26% to −40% if the reduction in rainfall lies between −10% and −15%. This corresponds well to our climate sensitivity tests where Quinapyramine for a reduction of −10% in rainfall the simulation shows a reduction of −32% in discharge. However, apart from these dramatic projections with reduction in flows we also have to acknowledge that rainfall may actually increase in the future, highlighting the uncertainty in the climate model scenarios. In addition to climate change, also future development of large-scale irrigation is expected to have a considerable impact on Zambezi discharge. For the high-level irrigation development the simulations show a decrease of mean annual Zambezi discharge at Tete by −460 m3/s (−18%). This is similar in magnitude as the reduction caused by evaporation from existing reservoirs (437 m3/s). Overall, the impact of the existing reservoirs is much larger than just reducing mean annual discharge, because in addition they also affect the discharge conditions.