Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Climate actions aiming to mitigate the climate crisis, strengthen human resilience and build climate justice require understanding the current impacts of climate change. To attribute the impacts of climate change, traditional impact attribution methods make use of climate models to represent the climate in our real world and a hypothetical world where climate change would not happen. Yet, operating these models requires substantial time and human resources. As the need for attributing observed impacts of the ongoing climate change grows along with climate-related disasters and socioeconomic inequalities in the face of global warming consequences, developing faster and simpler impact attribution frameworks becomes essential. In this context, ISIMIP3a was developed by the Inter-Sectoral Impact Model Intercomparison Project as an observed-based protocol proposing to attribute the impacts of climate change by dispensing with the climate models. To initiate an evaluation of the method regarding traditional GCM-based impact attribution, we offer a comprehensive analysis of the climate data at use in ISIMIP3a. Climate change patterns, trends and extremes are investigated at various spatial scales and contrasted with GCM-based climate data provided by ISIMIP3b. Climate reanalysis and simulations were analyzed at a 0.5° annual resolution for the historical period 1985-2014. Climate simulations were forced by six Global Climate Models (GCM) from the ISIMIP3b subset of the Coupled Model Intercomparison Project Phase 6 (CMIP6) ensemble. Climate reanalyses were retrieved from three reanalyses provided by ISMIP3a. Impact simulations were analyzed at the same 0.5°annual resolution than climate data and were forced by WaterGap. An exploratory data analysis over ISIMIP3a and ISIMIP3b factual climates showed stronger hot and cold global extremes in the ISIMIP3b factual despite similar global mean temperatures. A trend analysis over the ISIMIP3a counterfactual climate detected remaining long-term trends for temperature and relative humidity in extended regions across the world. These trends are suspected to arise from an insufficient detrending of observations by the ATTRICI method. Long-term precipitation trends were found over the Congo River basin and are suspected to arise from inhomogeneities over the 1978-1979 transition period in climate reanalyses datasets. Extreme impact analysis over the Congo River revealed a weakening of low discharge extremes in ISIMIP3a counterfactual by the detrending method ATTRICI. This resulted in the attribution to climate change of a reduction in 100-year return low discharge of 41% over 1985-2014 in ISIMIP3a. Yet, the 100-year return low discharge presented a negligible change attributable to anthropogenic climate change in ISIMIP3b. Despite the treatment of climate extremes to be improved in the detrending method, ISIMIP3a represent an innovative and promising approach for developing rapid impact attribution without the use of climate models.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

A rapid rise in lake levels was observed in Lake Baringo, Kenya, over the past decade, which culminated in 2020 with record-breaking levels and co-occurring floods that impacted thousands of people, infrastructure and biodiversity. Media sources and local communities argue that these high lake levels were aggravated by climate change. Consequently, in 2022, affected communities filed a climate change lawsuit against the Kenyan government, in which they claim to be victims of climate change-related floods and accuse the government of inadequate actions to mitigate climate change and prevent its adverse impacts. However, the link between Lake Baringo’s high lake levels, related floods and anthropogenic climate change had not yet been exhaustively scientifically investigated. By combining probabilistic extreme event attribution (EEA) methods with a water balance model (WBM) of the lake, this thesis aimed to explore the role of anthropogenic climate change in the rising water levels of Lake Baringo in order to (i) shed light on changing hydrometeorological extremes in the basin, (ii) contribute to the scientific evidentiary base of Kenya’s new climate change lawsuit and (iii) to unveil some limitations of carrying out EEA studies in the region. The first part of the study evaluated the extent to which the WBM, adapted to Lake Baringo, could realistically reconstruct the observed historical fluctuations and recent rapid rise in lake levels when forced with observational precipitation datasets. The WBM failed to accurately simulate the observed rapid rise in lake levels of the past decade and underestimated the record-breaking lake levels of 2020. Thus, due to the uncertainties in the precipitation products and the WBM simulation of the lake levels, confidence in the WBM’s use within an EEA framework, using lake level variations as impact-variable, is limited. As a second part of this thesis, it was assessed whether a precipitation-based attribution could provide a more reliable and definitive statement than the lake level-based attribution. However, significant disagreement in the sign of anthropogenic trends among the different observational precipitation datasets prevented a more confident attribution statement. In conclusion, based on the synthesis of the observed trend and climate model simulations using event definitions based on both lake level variations and seasonal averaged precipitation, no definitive statement could be made regarding the influence of anthropogenic climate change on the probability and magnitude of the record-breaking lake levels observed in Lake Baringo in 2020. Nevertheless, this thesis contributed to a better understanding of the drivers behind the high-impact event and the limitations behind carrying out extreme event and impact attribution studies in the East African region.