Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
Changing climatic conditions and increase of extreme events induced by climate change have impacts on non- adapted infrastructures, leading to destruction, damage costs and indirect impacts. To adapt infrastructures to those new conditions, there is a need to identify vulnerabilities and risks due to climate change. To that end, a Climate Risk and Vulnerability Assessment tool was designed based on 'Non-paper Guidelines for Project Managers: Making vulnerable investments climate resilient' proposed by the European commission. It relies on three steps: vulnerability assessment, risk assessment and climate mitigation. The tool was applied in a water treatment plant project located in Gorongosa, Mozambique. The climate projections used for the application of the tool are part of NEX-GDDP-CMIP6 dataset. Produced by NASA, this set of modelled data results from the bias-correction with quantile mapping and downscaling of the CMIP6 dataset. Comparing them to historical observed data revealed that average temperatures did match in trend, mean and extremes. However, the trends of the historic modelled maximum temperature were higher than the observed ones. Moreover, modelled precipitation did not match the observed distribution for high and extreme values. Therefore, the bias correction method quantile mapping was performed on those two modelled climate variables to diminish the biases. It performed well on maximum temperature data; the bias correction output complied much more with observation data than raw data. However, for precipitation data, extreme precipitation events biases corrected were much lower than the observed data and were all similar between models. To better capture the behavior of precipitation, it is recommended to use parametric distributions for mapping instead of non-parametric ones, as it was done in this study. Because of unverified compliance of the biases to the stationarity assumption and inconclusive results of bias-corrected precipitation, the raw NEX-GDDP-CMIP6 data with its biases were used to apply the tool at Gorongosa. The application of the tool permitted to learn about the climate evolution at this location. According to NEX-GDDP-CMIP6 and CORDEX (Coordinated Regional Climate Downscaling Experiment) results, a general increase of the temperature is projected in Gorongosa for future years. The precipitation projections agree on a decrease in average precipitation, but an increase for extreme events at Gorongosa. The tool also permitted to highlight the vulnerability of pumps and of the switchgear of the water treatment plant in Gorongosa to extreme air temperature increase. The sensitivities of other project's elements were also pointed up thanks to the tool, such as the sensitivity of rapid sand filter, pumps, and transformers to increasing air temperature. Some technical solutions exist and should therefore be applied to reduce the sensitivity and contribute to decrease the vulnerability of the project. The tool could be refined in its categorization approach and by imposing different thresholds, depending on the climate variable studied.