It is well known that in the Northern Hemisphere (NH) summer season, the ENSO influence is much weaker than in the winter season, hence seasonal predictions by coupled global climate models manifest significantly lower skill in the NH summer than in the NH winter season, especially over land. Research over the past decade has strongly suggested that proper land state initialization (especially soil moisture) is important to summer season precipitation predictions. To examine the impact of two different land models and different sources of initial land states on summer season predictions in this study, we use the NCEP global coupled climate model for seasonal prediction, known as the Coupled Forecast System (CFS). Specifically, CFS experiments are designed and executed with two different choices of land models and different sources of initial land states to demonstrate the extent to which upgrades to the land model and land data assimilation component of the CFS can improve CFS summer season predictions over the continental U.S. (CONUS).

The CFS is configured to execute either the modern Noah Land Surface Model (Noah LSM) or its older ancestor known as the Oregon State University Land Surface Model(OSU LSM). Additionally, the initial land states for the CFS are obtained from one of the three sources: 1) the NCEP/DOE Global Reanalysis 2 (GR2/OSU), which applies the OSU LSM as its land component, 2) the NCEP-NASA Global Land Data Asssimilation System utilizing the Noah LSM (GLDAS/Noah), and 3) the climatology of the latter (GLDAS/Noah/climo). CFS summer season prediction experiments have been carried out for a 25-year period (1980-2004) with 10 ensemble members per year and whose initial starting dates are from April 19 to May 3. We examine the impact of different land surface models and choice of different initial land states on seasonal precipitation, 2-meter temperature, 200 mb and 500 mb heights, SSTs, and on land surface fields including surface latent heat and sensible heat fluxes, among others.

Results from the experiments indicate that different initializations of the land surface model component have a large impact on the global climate model's seasonal predictions. Specifically, achieving competitive CFS performance requires the execution of a companion global data assimilation system with the very same land model as utilized in the land-component upgrade of the global climate model. Non-optimal initialization of the land surface component clearly degrades the CFS summer season performance (e.g. executing Noah LSM in CFS but land states initialized from GR2/OSU). This strongly indicates that it is naive to merely upgrade the land component of a global climate model for seasonal forecasting without simultaneously upgrading the land component of the companion global data assimilation system.