Meso-scale pressure dips accompanied by a severe convective storm of tropical cyclones

The phenomenon of meso-scale pressure dip, consisting of a rapid decrease and subsequent increase in surface pressure, is occasionally observed in the Japan Islands during the passage of a tropical cyclone (see figure). The movement of the pressure dip with a band-like structure is recognized in tandem with the movement of the accompanying tropical cyclones. After the pioneering work of Dr. Tetsuya Fujita (1952), several studies have reported that pressure dips are commonly accompanied by a severe local storm with strong winds and drastic rainfall changes that occasionally cause severe damage over the Japan Islands. Although a number of researchers have studied the pressure dip, the structure and formation process of pressure dips remains poorly understood. The purpose of this study is to describe the typical features of pressure dips associated with tropical cyclones using available data sets. We also investigate the processes of pressure dip formation using a numerical model.

In order to identify past occurrences of pressure dips, we analyzed the temporal variations of surface pressure as recorded in the barographs at meteorological stations of the Japan Meteorological Agency. During the period from 1980 to 1998, 89 tropical cyclones passed across Japan, and seven of these tropical cyclones were accompanied by pressure dips of significant amplitude and duration. It is noted that a greater proportion of the tropical cyclones may has been accompanied by pressure dips, but we cannot detect pressure dips in those tropical cyclones that passed over the sea far from the Japan Islands. Pressure dips form in association with tropical cyclones as they approach large areas of strong baroclinicity, which is characterized by the existence of the mid-latitude westerlies associated with a trough to the west of Japan at upper levels and of the frontal structure to the northern and western sides of the tropical cyclones at lower levels. Such environment conditions commonly develop around Japan during September and October, which is consistent with the seasonality of pressure dip formation over Japan. There is no record of pressure dip development within tropical cyclones that make landfall in the absence of the above atmospheric conditions, even during the peak tropical cyclones period of late July and August. Seven tropical cyclones recorded a central pressure of less than 980 hPa at the time they approached the Japan Islands. They moved rapidly toward the northeast or north-northeast during their passage over the Japan Islands before weakening rapidly and eventually transforming into extratropical cyclones.

We use observational data relating to seven tropical cyclones to clarify the typical features of pressure dips. A pressure dip is typical dimensions of 150 km length and 20-50 km width. The amplitude of observed pressure dips varies considerably from 1 to 9 hPa, with a typical magnitude of 2 hPa. In each of the recorded pressure dips, the maximum amplitude over the lifespan of the pressure dip was 4 to 9 hPa. The duration of the pressure dips recorded at various meteorological stations was generally 20-40 minutes. The pressure dips occurred only on the western side of the cyclone centers, or on the left and rear sides with respect to the movement direction of those tropical cyclones with a distance of 50-300 km from the cyclone centers. It is interesting to note that pressure dips appear at the remarkable boundary between the cloud region of tropical cycloen and the surrounding dry region. The intense radar echo region, corresponding to intense rainfall, occurs in advance of the pressure dip, while a sudden cessation of rainfall is recorded following the passing of the pressure dip. Time series data commonly shows that surface wind speed weakened after the passage of the tropical cyclone center, but strong winds were observed immediately prior to the arrival of the pressure dip.

To investigate the processes of pressure dip formation associated with tropical cyclones, we use the PSU/NCAR MM5 with the finest grid resolution at 5 km. The model successfully reproduced pressure dip with a band-like structure in an area west of the cyclone center. The simulated pressure dips are closely linked to warm potential temperature anomalies in the lower troposphere. These warm potential temperature anomalies result from the approximately adiabatic downward intrusion of an upper air mass of higher potential temperature from the rear side of a moving tropical cyclone. The low-level warm potential temperature anomalies are related to the occurrence of pressure dips, and develop in the limited region where an upper air mass of high potential temperature intrudes into a colder region of the low-level front. This explains why pressure dips appear only on the western side of the cyclone center. In fact the localized descending flow and relatively dry regions to the west of the cyclone center extend over much greater distances than the pressure dip. The localized descending flow also contributes to the cessation of intense rainfall around the pressure dip, which was described from observational data.

Based on the observational results and numerical simulations, we consider that large-scale atmospheric conditions control the formation of pressure dips. These atmospheric conditions described above might be favorable for the transformation of the tropical cyclone structure, e.g. extratropical transition. We therefore suggest that a pressure dip is an inherent feature of the asymmetric structure of a tropical cyclone undergoing extratropical transition. Because tropical cyclones rarely move over continents and meteorological stations within the large-scale atmospheric conditions described above, pressure dips associated with tropical cyclones have frequently been reported from Japan where is remote from a continental mass. The proportion of tropical cyclones contain associated pressure dips over the sea of Pacific and Atlantic may be higher. High resolution satellite data is required to reliably detect and characterizes mesoscale atmospheric structures such as pressure dips.