A Novel Initialization Technique for Decadal Climate Predictions

Model initialization is a matter of transferring the observed information available at start forecast to model. An optimal generally recognized be able improve climate predictions up few years ahead. However, systematic errors in models make process challenging. When transferred model time, discrepancy between and mean causes drift prediction toward model-biased attractor. Although such drifts can accounted for with posteriori bias correction techniques, evolving along might affect variability that we aim predicting, disentangling small magnitude signal from initial removed represents challenge. In this study, present an innovative technique aims reducing by performing quantile matching state time distribution. The adjusted belongs attractor amplitude scaled one. Multi-annual integrated 5 run EC-Earth3 Global Coupled have been initialized novel methodology, their skill has compared non-initialized historical simulations CMIP6 same decadal system but based on full-field initialization. We perform assessment surface temperature, heat content ocean upper layers, sea level pressure, barotropic circulation. added value shown North Atlantic subpolar region over Pacific temperature as well years. Improvements are also found predictive Meridional Overturning Circulation stream function Labrador Sea throughout when full field method.