Impacts of Land Process on the Onset and Evolution of Asian Summer Monsoon in the NCEP Climate Forecast System

Impacts of land models and initial land conditions (ICs) on the Asian summer monsoon, especially its onset, were investigated using the NCEP Climate Forecast System (CFS). Two land models, the Oregon State University (OSU) land model and the NCEP, OSU, Air Force, and Hydrologic Research Laboratory (Noah) land model, were used to get parallel experiments. The experiments also used land ICs from the NCEP/Department of Energy (DOE) Global Reanalysis 2 (GR2) and the Global Land Data Assimilation System (GLDAS). Previous studies have demonstrated that, a systematic weak bias appears in the modeled monsoon, and this bias may be related to a cold bias over the Asian land mass. Results of the current study show that replacement of the OSU land model by the Noah land model improved the model's cold bias and produced improved monsoon precipitation and circulation patterns. The CFS predicted monsoon with greater proficiency in El Nino years, compared to La Nina years, and the Noah model performed better than the OSU model in monsoon predictions for individual years. These improvements occurred not only in relation to monsoon onset in late spring but also to monsoon intensity in summer. Our analysis of the monsoon features over the India peninsula, the Indo-China peninsula, and the South Chinese Sea indicates different degrees of improvement. Furthermore, a change in the land models led to more remarkable improvement in monsoon prediction than did a change from the GR2 land ICs to the GLDAS land ICs.