Application of Weighted Multi-Model Ensemble Means, a Method to Manage Uncertainties between Climate Models

The performance of General Circulation Models (GCMs) should be assessed from region to region based on their ability to simulate temperature and precipitation for the baseline period in each gridded cell. In addition, the GCMs are associated with different uncertainties related to the climate system, numerical method, parameterization, calibration processes and so on. An ensemble mean model that includes GCMs with more similarity to the climate of the region increases the accuracy of the results related to climate change impact studies. On the other hand, improvement of the simple ensemble mean is inevitable and there is a research gap to consider different GCMs in one comprehensive ensemble model which will be able to manage the uncertainties between GCMs. This study aims to assess a weighted multi-model ensemble means applying to manage the uncertainty between five GCMs from AR4, and seven GCMs from AR5 under their corresponding emission scenarios. The climatic variables were downscaled and generated on a daily basis by LARS-WG. The Δ-changes between observed and future period of minimum temperature, maximum temperature, and precipitation were represented on Boxplots for months May to November. The ranges of uncertainties were different based on months, GCMs and emission scenarios. Overall, the weighted multi-model ensemble means method was able to model the uncertainty of climate models through the contribution of LARS-WG with great performance for the downscaling and generation of climatic variables for the future period. This method can be used for impact studies of climate change regarding Agricultural and hydrological models.

Keywords: GCMs, Uncertainty, LARS-WG, Temperature, Precipitation