Multi-Scale Modeling of Moist Convection Over India

Abstract

Clouds exist on all scales from turbulent processes to global scale and are organized on all the scales in between. They truly are multi-scale in nature. This makes them a very challenging simulation problem for atmospheric models. No model can resolve all of the scales relevant for cloud formation. Any attempt to develop a theory to resolve or represent all these processes require a deeper understanding of cloud scale processes at all the scales. It also requires the understanding of the state of the art in modeling these processes. The present research focuses on the moist convective representation in a hierarchy of models from General Circulation Model (GCM) to Large Eddy Simulations (LES). In finding out the missing gap between the different convective representations, we improve the simulation skill of GCM over India during monsoon, we propose a mechanism for the southward propagating meso-scale convective systems over Bay of Bengal (BoB) using Cloud Resolving Model (CRM), and explore the role of sea surface temperature (SST) in modifying the boundary layer stability using LES. The three types of studies are described below. In the rst part of the study we carry out a 28 year long simulation using a GCM. We focus on the intra-seasonal and inter-annual variability of Indian monsoon and its teleconnection with El Nino Southern Oscillation and Indian ocean dipole in the model. To perform a sensitivity study, we change the cloud adjustment time scale (CATS) in the Relaxed Arakawa-Schubert (RAS) cumulus parameterization. RAS is designed to perform well in a typical GCM resolution. CATS ( adj) implicitly determines the rate of change of convective available potential energy (CAPE) in the model by the clouds. We nd that changing the adj not only improves the mean monsoon predictability, it also in uences the teleconnections of Indian monsoon with El-nino Southern Oscillations and Indian Ocean dipole. It is realized in this part of the study that a better understanding of relationship between CAPE and precipitation is crucial to improving simulation skill of the model. To analyze the relationship between CAPE and deep convection, we carry out a season long CRM simulation over India. We nd that mesoscale convective systems (MCS) constitute a major portion of the overall precipitation over India. As MCS consume CAPE in a characteristically di erent way from what is proposed in a typical cumulus parameterization used in a GCM, a GCM is unable to correctly simulate most of the Indian monsoon precipitation. Present day cumulus parameterization lack representation of MCS. In our CRM simulation, one of the most signi cant MCS we found was the southward propagating MCS over the BoB. These MCSs move south within northward moving synoptic scale cloud clusters over the BoB. These systems have gravity currentlike structure and propagate orthogonal to lower tropospheric winds. High-resolution and cloud microphysics schemes are necessary to simulate these events using numerical models. A model with cumulus parameterization is unable to simulate the updraft-downdraft pair and the gravity current structure of this southward propagating mesoscale system. We nd that high model resolution is needed to resolve the updraft-downdraft pair and cumulus parameterization assumptions break down at such high resolutions. The southward propagating MCSs over the BoB sometimes originate over land (due to diurnal land heating) and sometimes over ocean. To understand the process of convective initiation over the ocean, it is necessary to understand the relationship of lower atmospheric stability as a function of SST. In the last part of the thesis, we use LES to simulate the boundary layer structure in response to changing sea surface temperature. A threshold value of SST (28 C) is found in the tropics above which there is a high probability of getting convection. LES simulations show that above this SST, the mixed layer depth is higher than the lifting condensation level.