Home
Geophysical Institute

GFI/BCCR Seminar: Current seasonal prediction skill of Norwegian Climate Prediction Model

Main content

Yiguo Wang (NERSC)

Current seasonal prediction skill of Norwegian Climate Prediction Model

Abstract

The Norwegian Climate Prediction Model (NorCPM) combines the Norwegian Earth System Model (NorESM) with an advanced data assimilation method - the Ensemble Kalman Filter (EnKF) - for the purpose of seasonal-to-decadal climate predictions. In this talk, I will start by giving an introduction of data assimilation and in particular of the Ensemble Kalman Filter. Practical implementation of data assimilation in NorCPM will also be discussed. We test the skill of the system for seasonal time scales (up to 1 year). The skill of NorCPM is estimated from the period of 1985-2010 with 4 hindcast starts per year (and 9 members) and compared to the North American Multi-Model Ensemble (NMME). Over the upper ocean, NorCPM achieves highly competitive skill compared to NMME both in correlation and RMSE. NorCPM performs the best in multiple regions including in the NA and the Nordic Seas. The prediction skill in the atmosphere (Z500, precipitation and SLP) is found to be related to ENSO and its teleconnection patterns. Seasonal variability of theprediction is discussed.