Large-Scale Circulation Variability over the Tropical Western North Pacific. Part II: Persistence and Transition Characteristics

View More View Less
  • 1 Department of Meteorology, Naval Postgraduate School, Monterey, California
© Get Permissions
Restricted access

Abstract

The variability of the large-scale circulation over the tropical western North Pacific is described within a framework defined by recurrent 700-mb circulation patterns that were defined by a fuzzy cluster analysis. Individual cluster patterns (defined in Part 1), which represent instantaneous depictions of the circulation variability, define favorable and unfavorable regions for tropical cyclone genesis and preferred track types. The fuzzy cluster coefficients, which describe the time variability of 700-mb large-scale circulation anomalies, are used to identify the basic persistence properties of the recurrent, anomalous circulation patterns. It is found that recurrent circulation patterns that are defined by small anomalies (i.e., close to the center of the cluster analysis phase space) are less persistent than recurrent patterns that represent distinct circulation anomaly patterns. Furthermore, the persistence of a particular sequence of anomaly maps that pass through a cluster is dependent upon the size of the cluster coefficients, which define how well the cluster pattern represents individual anomaly maps.

Analysis of transitions between clusters reveals that a rather limited set of transition paths exist. The most significant transition paths occur across a boundary within the cluster analysis phase space that separates circulation patterns that represent an active monsoon trough from patterns that represent an inactive monsoon trough. Physical descriptions of the significant transition paths are based upon 700-mb and 200-mb streamfunction and velocity potential anomalies, and anomalies of outgoing longwave radiation. The primary transition paths are found to be dependent upon interrelationships between several spatial and temporal scales of atmospheric variability. Furthermore, specific relationships were found to be critical for determining which transition path is followed. Secondary transition paths, which occur less frequently, are more dependent upon regional characteristics such as circulations within the tropical upper-tropospheric trough.

Physical associations between cluster patterns and tropical cyclone characteristics that were defined in Part I remain intact during transitions between the individual clusters. This is a significant result since the variability of the large-scale circulation within the cluster framework, which is defined by the cluster membership coefficients, can be used to infer sequences of persistent or transitioning circulation patterns. The potential application of the cluster framework for estimation of the stability of large-scale atmospheric circulation patterns and expected durations and transition paths is discussed in relation to the predictability of tropical cyclone characteristics.

Abstract

The variability of the large-scale circulation over the tropical western North Pacific is described within a framework defined by recurrent 700-mb circulation patterns that were defined by a fuzzy cluster analysis. Individual cluster patterns (defined in Part 1), which represent instantaneous depictions of the circulation variability, define favorable and unfavorable regions for tropical cyclone genesis and preferred track types. The fuzzy cluster coefficients, which describe the time variability of 700-mb large-scale circulation anomalies, are used to identify the basic persistence properties of the recurrent, anomalous circulation patterns. It is found that recurrent circulation patterns that are defined by small anomalies (i.e., close to the center of the cluster analysis phase space) are less persistent than recurrent patterns that represent distinct circulation anomaly patterns. Furthermore, the persistence of a particular sequence of anomaly maps that pass through a cluster is dependent upon the size of the cluster coefficients, which define how well the cluster pattern represents individual anomaly maps.

Analysis of transitions between clusters reveals that a rather limited set of transition paths exist. The most significant transition paths occur across a boundary within the cluster analysis phase space that separates circulation patterns that represent an active monsoon trough from patterns that represent an inactive monsoon trough. Physical descriptions of the significant transition paths are based upon 700-mb and 200-mb streamfunction and velocity potential anomalies, and anomalies of outgoing longwave radiation. The primary transition paths are found to be dependent upon interrelationships between several spatial and temporal scales of atmospheric variability. Furthermore, specific relationships were found to be critical for determining which transition path is followed. Secondary transition paths, which occur less frequently, are more dependent upon regional characteristics such as circulations within the tropical upper-tropospheric trough.

Physical associations between cluster patterns and tropical cyclone characteristics that were defined in Part I remain intact during transitions between the individual clusters. This is a significant result since the variability of the large-scale circulation within the cluster framework, which is defined by the cluster membership coefficients, can be used to infer sequences of persistent or transitioning circulation patterns. The potential application of the cluster framework for estimation of the stability of large-scale atmospheric circulation patterns and expected durations and transition paths is discussed in relation to the predictability of tropical cyclone characteristics.

Save