Shallow-to-Deep Convective Transition in the Amazon

Principal Investigator(s):
Yolande Serra, University of Washington

Angela Rowe, University of Washington

David Adams, National Autonomous University of Mexico
Henrique Barbosa, University of São Paulo

The shallow-to-deep convective transition over land typically refers to the growth of the convective boundary layer after sunrise, followed by the development of cumulus congests clouds in the late morning/early afternoon and transitioning to deep convective clouds in the late afternoon and early evening. Under favorable conditions, this diurnal convection can result in organized mesoscale convective systems (MCSs: Houze 2004) that last through the following morning. This timing of the diurnal cycle is well documented in observations of tropical and sub-tropical continental convection (Wallace 1975; Garreaud and Wallace 1997; Dai 2001; Yang and Slingo 2001; Machado et al. 2004; Adams et al. 2013). While many studies have focused on improving this process in models (see Arakawa 2004 and references therein), the shallow-to-deep transition remains poorly represented especially over land (Betts and Jakob 2002a; Bechtold et al. 2004; Wu et al. 2009). Specifically, global models tend to initiate convection too early, failing to develop morning shallow cumulus clouds (Betts and Jakob 2002a; Guichard et al. 2004; Bechtold et al. 2004; Rio et al. 2009). Understanding the life cycle of tropical convection and its organization into MCSs, along with its interactions and feedbacks with water vapor fields, has been hampered by the lack of long-term observations at the convective (2-20 km, 5-60 min) scales. While satellite observations have been the backbone of many such studies given its spatial coverage and long-term record, these data exhibit biases due to cloud cover, heavy rain and uncertainties in land emissivity. More recently, GNSS (Global Navigation Satellite System) meteorology, which provides all weather column precipitable water vapor at high time resolution, has been used to investigate these issues (Wu et al. 2003; Kursinski et al. 2008; Ralph et al. 2010; Adams et al. 2011; 2013; 2014; 2015). The recent DOE ARM mobile facility deployment in the Amazon (, launched as part of GOAmazon (, along with a dense GNSS network supported by Universidade do Estado do Amazonas (UEA)/Instituto Nacional de Pesquisas Espaciais (INPE) and co-located with the CHUVA Project ( sites for GOAmazon, now offer a unique opportunity to examine land-based convective processes in the tropics in detail. These measurements form the basis for the proposed study, which, along with global and regional model simulations, will address the following scientific questions: 1) What is the relative importance of tropospheric moisture, atmospheric stability and local boundary layer turbulence to the shallow-to-deep convective transition over the Central Amazon? Does the shallow-to-deep convective transition in the Amazon follow the shallow-to-deep convective transition paradigm described in the literature? 2) How does the relative importance of these processes compare during the wet and dry seasons? How does the paradigm differ with season? 3) How does the relative importance of these processes compare during dynamically forced (e.g., squall line events) vs. weakly forced environments? 4) How do local river-forest circulations contribute to the formation of nonprecipitating cloud structures in the region? 5) How well are these processes captured by the global variable resolution and regional model simulations at mesoscale resolution with different convective parameterizations? How do these results change when the model is run at the convective resolving scale?