Dynamics and Predictability of Tropical Cyclone Genesis Evaluated through a Coupled EnKF and 4DVar Data Assimilation Method during GRIP and PREDICT 2010.

Abstract:

The genesis of Hurricane Karl (2010) and Tropical Storm Mathew (2010) are examined and compared using a coupled ensemble Kalman filter and four-dimension variational data assimilation method (E4DVar) to assimilate conventional and field campaign observations. This research makes use of dropsonde and sounding observations, which were collected during both the Genesis and Rapid Intensification Processes (GRIP) and the Pre-Depression Investigation of Cloud Systems in the Tropics (PREDICT) field campaigns. E4DVar uses a mix of climatological and flow-dependent background error covariance obtained from ensemble forecasts together with the tangent linear model and its adjoint to perform the four-dimensional data assimilation. As a result, E4DVar analyses are able to capture multi-scale features of the developing disturbances accurately, such as the evolving thermodynamic and kinematic structure of pre-genesis Karl and Mathew. Preliminary diagnostics shows that the genesis of both storms followed a bottom-up process with several mesoscale convective vorticies (MCVs) observed in the pre-genesis stages. To determine the role of these MCVs and to better understand the initial development of the surface cyclone, a two-dimensional spectral decomposition and filtering of model variables, together with a vorticity budget, are performed on high-resolution (1.5 km) simulations of both systems.