Multi-model synthesis of future extreme temperature indices over Zambia

Abstract

Although temperature is one of the most important climate variables to be considered in adapting systems to climate change, its study over Zambia has until recently been largely ignored. A dearth of the literature on future temperature extremes is especially apparent. For the first time, future extreme temperature variability is analysed in Zambia for the period 1961–2100. The nonparametric Mann–Kendall test statistic is used at 5% significant level to compute trends. Sen’s slope estimator is used to give the magnitude of the observed trends. A two-tailed Kolmogorov–Smirnov statistical test is used to calculate the significance of the observed changes. What stands out from the findings is that warm days and nights become warmer while cool days and nights are projected to become non-existent by 2100. Additionally, an increase in radiative forcing and time is seen to intensify extreme temperature events. Notably, much of southern Zambia is projected to experience ~ 30% increment in TX90p while the northern half will experience ~ 40% under RCP4.5 by the middle of the century (2021–2050). If the business-as-usual trajectory (RCP8.5) is followed, an intensification of TX90p is observed with ~ 40% in the southern half of the country and ~ 50% in the northern half. These projections are observed to almost double towards the end of the century (2071–2100). The projected increase in warm nights (TN90p) is more over Luapula and Northwestern province while the lowest is observed over the Livingstone/Magoye region. These results provide a foundation for strategic planning purposes to avert losses related to extreme temperature.