Verification of COAMPS-TC predictions of the synoptic-scale flow in the tropics and subtropics

Tropical cyclone (TC) track forecasts from regional TC models, such as COAMPS-TC, tend to be less accurate than those from global atmospheric models. This could be because the regional models are relatively deficient in the prediction of the synoptic-scale flow, which largely governs the movement of tropical cyclones. To diagnose the performance of COAMPS-TC on the synoptic-scale and its relationship to TC track errors, we developed a verification approach focused on the identification of systematic errors in synoptic-scale COAMPS-TC forecasts. Here, we describe this synoptic-scale verification approach and its application in the assessment of different physics options for COAMPS-TC. Results indicate that the choice of physics (cloud microphysics, cumulus, radiation, boundary layer, etc.) has profound, and often complicated, implications for synoptic-scale systematic forecast errors. For example, the control configuration of COAMPS-TC has a significant warm bias in the tropical and (especially) subtropical upper-troposphere, which induces biases in the strength and position of the subtropical high. Thus, the tropical cyclone steering flow provided by the subtropical high can be improved in COAMPS-TC by improving the upper-level temperature forecasts. Changes to the cloud microphysics, cumulus, and radiation are all shown to reduce the upper-level temperature bias and improve the representation of the subtropical high, but each change has unique and not necessarily desirable implication for other aspects of the forecast state (such as lower-tropospheric relative humidity). Finding the best combination of physics for the prediction of the synoptic-scale flow in COAMPS-TC takes careful consideration of the many aspects of the synoptic-scale forecast state relevant to TC prediction.