Comparative Assessment of Climate Engineering Scenarios in the Presence of Parametric Uncertainty.

Climate engineering (CE) measures are increasingly discussed when dealing with the adverse impacts of climate change. While much research has focused on individual methods, few studies attempt to compare and rank the effectiveness of these measures. Furthermore, model uncertainties are seldom acknowledged and lesser still, estimated when CE scenarios are assessed. In this work, we quantify the variance in outcomes due to poorly constrained model parameters under several idealized CE scenarios. The four scenarios considered are (1) warming under the high emission scenario Representative Concentration Pathway 8.5 without CE applied and the same emission scenario with (2) afforestation,(3) solar radiation management, and (4) artificial ocean alkalinization. By considering the parametric uncertainty in model outputs, we demonstrate the problems with comparing these scenarios using a single parameter setting. Using statistical emulation, we estimate the probability distributions of several model outcomes. Based on such distributions, we suggest an approach to ranking the effectiveness of the scenarios considered according to their probability of avoiding climate thresholds. Plain Language Summary: Various intervention techniques have been proposed to manipulate the climate system at a large scale to combat the adverse effects of climate change. While many studies have focused on specific techniques, relatively little has been done to compare and rank the effectiveness of these methods within a single model. Furthermore, the uncertainties arising from the use of climate models are seldom acknowledged or assessed. In this work, we analyze the uncertainty in the model's simulated outcomes caused by poorly constrained model settings under four idealized future scenarios, with and without climate interventions. Our results highlight the importance of taking into account the model's uncertainty when analyzing or communicating the simulated outcomes. Moreover, we suggest an approach to ranking the effectiveness of the interventions considered based on goals of societal importance. Key Points: We demonstrate how uncertainty analysis can become an essential part in comparative studies on climate engineeringModel outcomes are probability distributions and should be treated as such when communicating the effects and impacts of climate engineeringThe probability of avoiding dangerous threshold is calculated for various climate engineering scenarios [ABSTRACT FROM AUTHOR]