Object-Based Analyses of Mesoscale Convective Systems and Embedded Storms over the Indian Monsoon Zone Using Datasets from Satellite, Radar and Model Simulations

Mesoscale convective systems (MCSs), the largest type of deep convective storms are formed when convection aggregates and grows upscale, forming a distinct mesoscale circulation through the interaction of multiple storms. Thus, storms play an important role in MCS organization. Due to their large size, longer duration, and larger precipitation, MCSs cause high-impact extreme weather events like lightning, damaging hail, gusty winds, and flooding. During the Indian summer monsoon (June-September), synoptic-scale weather systems move across the monsoon zone (MZ), causing MCSs to form frequently. MCSs often produce widespread and heavy rain throughout the MZ. Hence, studies on structure and evolution of MCSs highlighting the organization of convection are needed for an improved understanding of MCS.

In this study, we used an object-based cloud-tracking method (Feng et al., 2018) to identify and track MCSs and embedded storms in remote sensing observations and numerical simulations. The work is divided into three parts. In the first part, we tracked MCSs over the monsoon zone using geostationary satellite infrared brightness temperature (IRTb) and GPM IMERG precipitation from June-September, 2014 to 2019 and examined various aspects of observed MCSs (n=2092) such as spatial coverage, diurnal cycle, rainfall amount, and land-ocean contrast. The majority of MCSs are positioned in the monsoon trough's southeast-northwest stretch and account for more than 60% of total precipitation. For MCSs with short and long lifespans, there was a clear land-ocean divide and varied lifecycle trends. Oceanic MCSs last longer, are deeper, and provide more rainfall over a larger area than land-based MCSs.

In the second part of the study, we explored embedded storm structures for those MCSs that exist within the radar domain (n=65) by applying a storm classification algorithm to the S-band Doppler radar observations during June-September 2015. We observed that an MCS contains many precipitation features, especially during early stages of development when multiple convective clusters begin to amalgamate. Furthermore, we investigated the co-evolution of numerous storm parameters (e.g., areas of convective/stratiform precipitation, convective core length, and top heights) as a function of MCS lifetime. Distinct vertical structures are observed for the convective, stratiform, and anvil components of MCSs.

In the third part of this work, we examine the ability of a convection-permitting Weather Research Forecast (WRF) model in simulating MCSs and their characteristics (initiation, size, intensity, lifetime, propagation) during June-September 2015. A similar cloud-tracking algorithm is applied to WRF-simulated data (reflectivity, IRTb, and precipitation) to identify and track MCS in the simulation. Although the model underestimated the number of observed MCSs, the composite evolution and frequency distribution of convective area, precipitation amount, MCS propagation speed produces reasonable agreement with observations but underestimate stratiform areas. Consistent with observations, the simulated MCS properties showed a gradual increase from convective initiation to around the first half of the MCS lifetime. We observed that an MCS contains multiple precipitation features, particularly during the initial development stage when multiple convective clusters begin to aggregate. More details on observed MCSs and embedded storm structures, as well as their representation in simulation, will be presented.