Future projections of temperature and precipitation extremes for the city of Beira, Mozambique

Due to its geographical location, geomorphology and socio-economic context, Beira (Mozambique) is one of the cities most threatened by climate change worldwide. The large-scale destruction wreaked by Cyclone Idai in March 2019 has turned the spotlight on the vulnerability of the city to extreme weather events. Correspondingly, Politecnico di Torino has launched an interdisciplinary master thesis laboratory to support Comunità di Sant’Egidio in developing a vision for the future of Beira, based on sustainability and resilience to climate change. In this context, the objective of the present work is to depict the future of the extreme climatic conditions that are likely to affect the area in the next decades. To pursue this aim, a multi-model ensemble of climate simulations from CMIP6 (sixth phase of the Coupled Model Intercomparison Project) has been employed to analyse future projections under two alternative scenarios. Based on daily temperature and precipitation data, indices of moderate climate extremes are computed and analysed. The performance of models in simulating the indices in the study area is firstly assessed by comparing them with indices derived from the ERA5 reanalysis dataset. Subsequently, trend analysis and the temporal comparison of probability density functions are employed to assess the projected changes of the indices. The results obtained for indices based on maximum and minimum daily temperature are consistent with global warming. An increase in the intensity and frequency of temperature extremes is projected along the 21st century for both future scenarios. On the other side, projections concerning precipitation indices are more uncertain, given that models often disagree on the sign of change. Nonetheless, the results suggest the increase in the drought conditions and of the intensity and variability of heavy rainfall under the high emission scenario. Besides, extreme precipitation events are expected to last longer, thus potentially increasing the risk of flooding. Furthermore, a methodology has been proposed to extrapolate from the observed indices information about the critical precipitation events that have devastated the area in the past (e.g. cyclone Idai). This empirical approach has led, despite a certain degree of uncertainty, to correlate the values of internationally recognised indices with high-return period precipitation events and to predict an increase of their frequency according to future projections. The output of this work is intended to be useful for data-based decision making regarding adaptation strategies and for future studies on the climate of the area.