The Heated Condensation Framework as a Convective Trigger in the NCEP Climate Forecast System version 2

In a recent work, Bombardi et al. (2015) modified the convective triggering mechanism of the deep convection scheme of the National Centers for Environmental Prediction (NCEP) atmosphere-ocean global coupled model (AOGCM) used for seasonal prediction, the Climate Forecast System version 2 (CFSv2; Saha et al. 2014a). These modifications resulted in improvements in the CFSv2 representation of the Indian summer monsoon as well as the monsoon onset timing. Here, we present a follow-up work using an updated version of the convective triggering mechanism called the Heated Condensation Framework (HCF). Our hypothesis is that the CFSv2 precipitation biases can be reduced by improving the timing of convection through the improvement of the convective trigger function. The HCF triggering mechanism was developed by Tawfik and Dirmeyer (2014) and elaborated upon in Tawfik et al. (2015a,b). The trigger is based on a diagnostic analysis of land-surface controls of boundary layer properties and cloud formation. The HCF was envisioned as an alternative diagnostic to represent the atmospheric background state with respect to convection, which can be defined from standard profiles of temperature and specific humidity. These two variables quantify how conditioned the atmosphere is to moist free convection due to surface heating. Instead of lifting a hypothetical unmixed parcel using parcel theory, the HCF constructs a hypothetical boundary layer by incrementally inputting heat at the surface. An updated version of the HCF (HCFv2) is implemented as a convective triggering mechanism into the CFSv2. The new trigger replaces the original triggering mechanism in both the deep (Simplified Arakawa-Schubert – SAS) and shallow (SAS based) convective schemes (Han and Pan 2011). The performance of the original and new triggering mechanisms is first compared using radiosonde observations. Then, a series of short (two-week-long) and seasonal (seven-month-long) hindcasts are performed to evaluate the influence of the triggering mechanisms in the CFSv2 representation of the diurnal cycle of precipitation and seasonal precipitation. The observational analysis show that the new trigger corrects a common problem in the original triggering mechanism in SAS, which is that the original trigger initiates convection too often. When implemented in the CFSv2, the impact of the new trigger on the diurnal cycle of precipitation seems to be small and confined to limited land regions. There are improvements in the phase of the diurnal cycle of precipitation over the western and southern United States. The seasonal simulations show that the HCF trigger improves the representation of seasonal precipitation over most of North America and over the Indian subcontinent. In addition, the new trigger improves the representation of the monsoon onset date over India. Since the HCF convective trigger shows improvements in the representation of convection in the CFSv2 and considering that the HCF is more physically appropriate than the original trigger in SAS, this method holds promise for future operational prediction systems.