Search

This study promotes the critical use of air pollution modelling results for health and agriculture impacts, with the primary goal of providing more reliable estimates to decision makers. To date, the accuracy of air quality (AQ) models and the effects of model-to-model result variability (which we will refer to as model uncertainty) on impact assessment studies have been often ignored, thus undermining the robustness of the information used in the decision making process and the confidence in the results obtained. A suite of twelve PM 2.5 and ozone concentration fields produced by regional-scale chemistry transport Air Quality (AQ) models during the third phase of the Air Quality Model Evaluation International Initiative (AQMEII) has been used to calculate the impact of air pollution on premature deaths and crop yields. An innovative technique is applied to bias-adjust the models to available observations. The model results for ozone and PM 2.5 are combined in a multi-model (MM) ensemble, which is used to estimate the damage and economic cost to human health and crop yields, as well as the associated uncertainties. The MM ensemble quantifies directly the uncertainty introduced by AQ models into the air pollution impact assessment chain, while the indirect use of experimental information through a bias adjustment, reduces the uncertainty in the ozone and PM 2.5 fields and subsequently the uncertainty of the final impact assessment and cost valuation. The analysis over the European countries analysed in this study shows a mean number of premature deaths due to exposure to PM 2.5 and ozone of approximately 370,000 (inter-quantile range between 260,000 and 415,000) and a relative yield loss of approximately 7% to 9% (depending on the exposure metrics used, for wheat and maize together). Furthermore, the results indicate that a reduction in the uncertainty of the modelled ozone by 61% and by 80% (depending on the aggregation metric used) and by 46% for PM 2.5 , produces a reduction in the uncertainty in premature mortality and crop loss of >60%, and of an equivalent percentage in the final uncertainty of cost valuation, providing decision makers with more accurate estimations for more targeted interventions. [ABSTRACT FROM AUTHOR]

• We propose a new multi-model climate projection combining approach. • The spatiotemporal dynamics of climate extremes in Korea was investigated. • The proposed approach improved the reliability of climate extreme projections. • The frequency of wet events in South Korea will increase, whereas that of drought events will decrease. The severity and frequency of climate extremes will change in the future owing to global warming. This can severely impact the natural environment. Therefore, it is common practice to project climate extremes with a global climate model (GCM) in order to quantify and manage the associated risks. Several studies have demonstrated that a multi-model ensemble approach increases the reliability of predictions by exploiting the strengths and discounting the weaknesses of each climate simulator. However, the available multi-model averaging approaches exhibit significant drawbacks as they are not capable of extracting different climate extreme characteristics from the climate models. This study proposes a new approach that combines multiple models for projecting climate extremes by accounting for different extreme indices in the climate model performance weighting scheme. The capability of this method was evaluated with respect to reliability ensemble averaging (REA) and Taylor diagram-based GCM skill approaches for reproducing wet and dry precipitation events. The proposed multi-model averaging approach outperformed the available approaches in reducing the root mean square error (RMSE) by 5% and 54% in the wet and dry precipitation conditions, respectively. Therefore, it can be concluded that incorporating the different precipitation extremes in a multi-model combination approach could enhance the assessment of climate change impacts on the climate extremes. The climate change impacts on the extreme events, based on the proposed multi-model ensembles, is thus assessed using the standardized precipitation indexes of 3 month, 6 month, and 12 month durations. In general, the results exhibited that the frequency of wet events increases, whereas that of drought events decreases. [ABSTRACT FROM AUTHOR]