Properties of the Ensemble Transform Kalman Filter adaptive sampling strategy for tropical cyclones

The ensemble transform Kalman filter (ETKF) has been provided as one of the adaptive sampling strategies for aircraft surveillance mission planning in the synoptic environment of tropical cyclones (TCs). In the original version of the ETKF, the targets usually coincided with the spread of the TC location, given the >2-day lead-time necessary for mission planning. As a result, only minimal useful guidance on targeting was derived in the environment of the TC. To correct for this, a new version of the ETKF has been developed. In this version, the axisymmetric part of the TC has been removed from each ensemble member, and all the ensemble members have been shifted into a storm-relative frame. The goal is to identify sensitivity associated with asymmetries in the TC system, its immediate vicinity and far field environment.

Several issues about the characteristics of sensitive areas based on the ETKF will be discussed, including sensitivity to observations of winds, temperature and specific humidity targeted at different vertical levels, sensitivity to different global model ensembles or their combination, sensitivity to the forecast lead time, and comparison with the ensemble variance. In addition, the evolution of ETKF sensitivity structure through various stages of the TC life cycle will also be presented. The data used are from the THORPEX Interactive Grand Global Ensemble (TIGGE) database and the primary case examined is Typhoon Sinlaku in 2008.