Assessing PV Streamer Activity and Its Relationship with TC Predictability in Subseasonal Forecasts

Potential vorticity streamers (PVSs) are elongated filaments of high PV air that can serve as channels for tropical-extratropical air mass exchange. This air mass exchange in PVSs is facilitated through anticyclonic Rossby wave breaking (AWB), where upstream low PV air is advected poleward over downstream high PV air in the upper troposphere. This flow evolution results in an elongated positively tilted upper-tropospheric trough (i.e., PVS) downstream of the AWB axis. In the warm season, subtropical PVSs modify the tropospheric environment by enhancing vertical wind shear (VWS) and moisture anomalies in their vicinity. These environmental changes play an important role in enhancing or suppressing nearby tropical cyclone (TC) activity in oceanic basins. Given the importance that PVSs play in modifying their local environment in the subtropics, predicting the correct magnitude and location of PVS activity on subseasonal timescales may also be important in determining corridors where TC activity is impacted.

This study compares the Navy Earth System Prediction Capability (Navy-ESPC) model to the Climate Forecast System version 2 (CFSv2) in order to investigate subseasonal predictability of PVSs and TCs in the northern hemisphere during the warm season (June–November). 45-day reforecasts from 2009-2015 are used to determine predictability of PVSs compared to the verifying analysis. PVSs are identified in these reforecasts on the 350-K isentropic surface bounded by the 2-PVU contour, where PVSs are defined as the high PV trough axis that occurs downstream of the AWB axis. PVS events are also categorized by their area and intensity in order to show regions where repeated PVS occurrence is prevalent in the Navy-ESPC, and also to compute an activity metric incorporating PVS size and intensity for a reforecast period. Skill scores and mean state biases are used to determine the predictability of PVS activity between 1–6 week lead times. Finally, these results are compared to TC activity in these subseasonal models to see if increased/decreased skill in predicting PVS activity is also related to similar skill changes in TC activity forecasts.