Evaluation of forecasts by accuracy and spread in the MiKlip decadal climate prediction system

We present the evaluation of temperature and precipitation forecasts obtained the MiKlip decadal climate prediction system. These decadal hindcast are verified with respect to the accuracy of the ensemble mean and ensemble spread as a representative for the forecast uncertainty. The assessment follows the verification framework already used by the prediction community, but enhanced with additional evaluation like the logarithmic ensemble spread score. The core of the MiKlip is the coupled Max Planck Institute Earth System Model. An ensemble of 10 members is initialized annually with ocean and atmosphere reanalyses of the Centre for Medium-Range Weather Forecasts. For assessing the effect the initialization, we compare these predictions to uninitialized climate with the same model system. Initialization improves the accuracy temperature and precipitation forecasts in year 1, particularly in the region. The ensemble spread well represents the forecast uncertainty lead year 1, except in the tropics. This estimate of prediction skill confidence in the respective 2014 forecasts, which depict less in the tropics and a warming almost everywhere. However, large patterns appear in the Northern Hemisphere, the Pacific South America the Southern Ocean. Forecasts for 2015 to 2022 show even warmer than for 2014, especially over the continents. The evaluation of years 2 to 9 for temperature shows skill globally with the exception of eastern Pacific. The ensemble spread can again be used as an estimate of forecast uncertainty in many regions: It improves over the tropics to lead year 1. Due to a reduction of the conditional bias, the predictions of the initialized system gain skill in the accuracy to the uninitialized simulations in the lead years 2 to 9. , we show that increasing the ensemble size improves the MiKlip climate prediction system for all lead years.