ATMOSPHERIC SCIENCE LETTERS Atmos. Sci. Let. 13: 103–107 (2012) Published online 23 December 2011 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/asl.367 The onset of the West African monsoon simulated in a high-resolution atmospheric general circulation model with reanalyzed soil moisture fields T. J. Yamada, 1 * S. Kanae, 2 T. Oki 3 and Y. Hirabayashi 4 1 Faculty of Engineering, Hokkaido University, Kita 13, Nishi 8, Kita-ku, Sapporo 060-8628, Japan 2 Department of Mechanical and Environmental Informatics, Tokyo Institute of Technology, 2-12-1 O-okayama, Meguro-ku, Tokyo 152-8522, Japan 3 Institute of Industrial Science, The University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8595, Japan 4 Institute of Engineering Innovation, School of Engineering, The University of Tokyo, 2-11-16 Bunkyo-ku, Tokyo 113-8656, Japan *Correspondence to: T. J. Yamada, Faculty of Engineering, Hokkaido University, Kita 13, Nishi 8, Kita-ku, Sapporo 060-8628, Japan. E-mail: tomohito@eng.hokudai.ac.jp Received: 25 June 2011 Revised: 15 September 2011 Accepted: 22 November 2011 Abstract The simulation of the onset of the West African monsoon (WAM), associated with the northward shift of the Intertropical Convergence Zone (ITCZ) occurring in May, June, and July, is a challenging task for atmospheric general circulation models (AGCMs), because of complex water and energy balance through the land-atmosphere interaction as well as atmospheric processes. We provide evidence that a combination of state-of-the-art global fields of reanalyzed soil moisture anomalies and a sufficiently high-resolution (50 km) AGCM produces a successful simulation of the northward shift of the ITCZ and the following onset of the WAM in July. Copyright 2011 Royal Meteorological Society Keywords: West African monsoon; soil moisture; African Easterly Jet; AGCM 1. Introduction The Sahel is one of the most densely populated regions of the African continent and is character- ized meteorologically as a monsoon region. The management of water resources across the Sahel in the monsoon season is important for agriculture and other human activities. The Intertropical Conver- gence Zone (ITCZ) shifts abruptly northward from a quasi-stationary location at 5 N (Gulf of Guinea) in May–June (MJ) to another quasi-stationary location at 10 N in July–August (JA; Sultan and Janicot, 2000, 2003), bringing about the onset of the monsoon over the Sahel. The monsoon onset is routinely captured in regional model simulations (Gall´ ee et al., 2004; Hagos and Cook, 2007) by realistic lateral boundary con- ditions of the atmosphere. On the other hand, many GCMs, despite good simulations of the mature West African monsoon (WAM) in August, have failed to accurately simulate the ITCZ rapid northward transi- tion from the Gulf of Guinea (d’Orgeval et al., 2006). This article shows a successful example of the onset simulation in a high-resolution atmospheric general circulation model (AGCM) that is not forced with realistic lateral boundary conditions and atmospheric initial conditions. 2. The model and experimental design Our study involves two sets of eight AGCM simula- tions in 1998, using the AGCM designed for the Earth Simulator (AFES; Ohfuchi et al., 2004). The AFES initially adopted physical and dynamical cores from the Center for Climate System Research (The Uni- versity of Tokyo)/National Institute for Environmen- tal Studies (CCSR/NIES) AGCM (Numaguti et al., 1997), which were then modified and developed inde- pendently at the Earth Simulator Center. The eight AGCM simulations started at 00Z on 24–31 March using the European Centre for Medium- Range Weather Forecasts (ECMWF) Reanalysis 40 (ERA40) dataset as an atmospheric initial condition. We used a T239 (50 km) triangular truncation, which is approximately one fifth smaller than the resolu- tions (250–300 km) used in previous studies (Kanae et al., 2006; Douville et al., 2007). We used the pentad Atmospheric Model Intercomparison Project (AMIP) optimally interpolated Sea Surface Tempera- ture (OISST) data of 1998 as the boundary condition. Hereafter, we refer to the set of basic model simu- lations as the control run (hereafter CTRL-T239), in which the AMIP OISST data were prescribed at each time step. The eight simulations differed only in their initial atmospheric conditions. The ensemble mean of the eight simulations is analyzed below. Another set of eight simulations called SOIL-T239 differed from CTRL-T239 only in the treatment of subsurface soil moisture. In CTRL- T239, subsurface soil moisture was computed by mod- eled precipitation and atmospheric conditions with land–atmosphere interactions, as in the standard in AGCM simulations. Global fields of observation- based soil moisture are not available yet; therefore, in SOIL-T239, a simulated global field of reanalyzed subsurface soil moisture was prescribed at each time Copyright 2011 Royal Meteorological Society