Automated tracking of dry intrusions on satellite water vapour imagery and model output

A satellite image-processing technique has been developed for the identification and tracking of upper-tropospheric features related to midlatitude synoptic-scale cyclogenesis. Persistent warm radiance features are detected on water vapour images using an adaptive thresholding technique, tracked using estimations of the motion of the features through displacement and cross-correlation of successive images, and then screened using both image-based and model-based criteria. The aim is the characterization of jet dynamical features and dry intrusions of stratospheric air into the upper troposphere. On a selected sample of events, the resulting trajectories prove to be very consistent with the subjective identification of cyclogenesis events on imagery. In accordance with potential vorticity theory, the detected warm features are correlated with positive anomalies of potential vorticity. This identification technique can be applied to model output, which suggests future applications in forecast verification and data assimilation of midlatitude cyclones.