Current and perturbed climate as simulated by the second-generation Canadian Regional Climate Model (CRCM-II) over northwestern North America Received: 8 July 2002 / Accepted: 25 April 2003 / Published online: 23 August 2003 Ó Springer-Verlag 2003 Abstract An updated version of the Canadian Regional Climate Model (CRCM-II) has been used to perform time-slice simulations over northwestern North Amer- ica, nested in the coupled Canadian General Circulation Model (CGCM2). Both driving and driven models are integrated in a scenario of transient greenhouse gases (GHG) and aerosols. The time slices span three decades that were chosen to correspond roughly to single, double and triple current GHG concentration levels. Several enhancements have been implemented in CRCM-II since the CRCM-I climate-change simulations reported upon earlier. The larger computational domain, extending further to the west, north and south, allows for a better spin-up of weather systems as they enter the regional domain. The increased length of the simula- tions, from 5 to 10 years, strengthens the statistical robustness of the results. The improvements to the physical parameterisation, notably the moist convection scheme and the diagnostic cloud formulation, cure the excessive cloud cover problem present in CRCM-I, re- duce the warm surface bias and prevent the occurrence of grid-point precipitation storms that occurred with CRCM-I in summer. The dynamical ocean and sea-ice components of CGCM2 that is used to provide atmo- spheric lateral and surface boundary conditions to CRCM-II, as well as the use of transient rather than equilibrium conditions of GHG and the inclusion of direct aerosols forcing, in both CGCM2 and CRCM-II, increase the realism of the CRCM-II climate-change simulation. 1 Introduction General circulation models (GCMs), including land- surface processes and coupled with dynamical ocean and sea-ice models (hereinafter termed CGCMs), currently provide the most sophisticated, physically based ap- proach to simulate the large-scale response of the cli- mate system to projected scenarios of increasing greenhouse gases (GHG) and aerosols concentrations. Because of the computational load of integrating com- plex CGCMs over simulated periods of several centu- ries, the horizontal resolution has to be limited to grid meshes of the order of a few hundred kilometres. Such resolution is inadequate to resolve atmospheric pro- cesses operating at mesoscale, such as frontal zones, and small-scale surface forcing due to local orography or in- land water bodies (IPCC 1995). Surface hydrology and atmospheric moist processes in particular are known to exhibit high variability at small (below the CGCMs’ resolved) scales. Furthermore, climate-change impact and adaptation studies deal with processes operating on much finer scales than currently affordable with CGCMs. Nested limited-area regional climate models (RCMs) represent an appealing approach to achieve finer spatial resolution climate and climate-change simulations at an affordable computational cost. Nested with atmospheric data simulated by CGCMs and integrated for long periods of time, RCMs are potentially useful tools for identifying effects of anthropogenic forcing at regional scale (e.g. Pan et al. 1999). These RCMs can be run at fairly high resolutions (with grid meshes of a few tens of kilometres) over an area of interest covering typically several millions of square kilometres. Such RCM cli- mate-change simulations have been made for various parts of Europe (e.g. Jones et al. 1997; Machenhauer et al. 1998; Jones and Reid 2001; Durman et al. 2001; Christensen et al. 2001; Frei et al. submitted 2002), of North America (e.g. Giorgi et al. 1998; Laprise et al. Climate Dynamics (2003) 21: 405–421 DOI 10.1007/s00382-003-0342-4 R. Laprise ® D. Caya ® A. Frigon ® D. Paquin R. Laprise (&) ® D. Caya ® A. Frigon ® D. Paquin CRCM/UQAM – Ouranos, 550 West Sherbrooke St, 19th floor, Montre´al (Que´bec), Canada H3A 1B9 E-mail: laprise.rene@uqam.ca