Characteristics of quasi-stationary mesoscale convective systems during the warm season in Japan

Meso-beta scale quasi-stationary mesoscale convective systems (QSMCSs) are well-known phenomena as a cause of heavy rainfall in Japan. The Shikoku Island, a southern part of Japan, is one of the regions in which rain-producing QSMCSs frequently occur. The annual mean precipitation at the Kochi city that is located on the Pacific side of the Shikoku Island during 1981 and 2010 is 2547.0 mm, while the mean rainfall during March and November, a warm season is 2324.3 mm, i.e., 91.2 % of the annual mean precipitation. Extreme events that produced heavy rain have been studied from observational, analytical, and numerical points of view through case studies. These case studies indicated that QSMCSs frequently occur under the influences of synoptic-scale disturbances such as tropical cyclones, extratropical cyclones, and fronts. On the other hand, the characteristics of these QSMCSs always differ depending on the location and season. Furthermore, there are few studies to investigate statistical characteristics of QSMCSs.

The purpose of this study is to statistically investigate the characteristics of long-lived QSMCSs during the warm season in Japan. In this study, we use operational radar data that are provided by Japan Meteorological Agency. The analysis period is March-November during 2005-2012. This data set has a high spatial and horizontal resolution, which is useful in identifying and extracting QSMCSs all over Japan. The temporal and spatial resolutions of the data are 1 km and 10 minutes, respectively. Here, we focus on the Shikoku Island where the annual precipitation is among the highest throughout the Japanese Islands. In order to extract QSMCSs, we use Algorithm for the Identification and Tracking Convective Cells (AITCC) that was developed by Shimizu and Uyeda (2012). This algorithm is capable of identifying and tracking individual convective cells within a mesoscale convective system. At the same time, the algorithm is applicable to identify and track a mesoscale convective system itself. Using this advantage, we apply a part of the algorithm for identifying QSMCSs. The procedure of how to extract QSMCSs is as follows. Firstly, we define an area of the observed precipitation intensity that exceeds 10 mm/h more than 200 km² as an MCS. Secondly, we track the extracted MCSs in time series and identify the lifetime of the QSMCSs.

The total number of the extracted QSMCSs is 13,447. The months when the total number that exceeds 2,000 are from June to September. In particular, there are two peaks in July and September in the occurrence of the QSMCSs. In order to examine what type of synoptic-scale disturbances affects in the occurrence of QSMCSs, we divide the synoptic-scale conditions into four categories: extratropical cyclones, tropical cyclones, quasi-stationary fronts, and the others (synoptically undisturbed conditions). Through the examination of weather maps, it was found that QSMCSs in April, May, and November occurred under the influence of extratropical cyclones. QSMCSs during June and September occurred under the influence of and quasi-stationary fronts. QSMCSs in August occurred under not only the influence of tropical cyclones but also synoptically undisturbed conditions. The analyses of the precipitation intensity of the extracted QSMCSs indicated that the mean value of the temporally and spatially averaged precipitation intensity within the extracted QSMCSs increased with the increase in the duration time of QSMCSs if the duration time was less than 60 min. On the other hand, the precipitation intensity within the QSMCSs with the duration time of being more than 60 min did not increase with the increase in the duration time of QSMCSs. The analyses of the area of the extracted QSMCSs showed that an area of extracted QSMCSs increased with the increase in the duration time of QSMCSs when the duration time was less than 60 min. On the other hand, for the QSMCSs with their duration time being greater than 60 min, their area did not increase with the increase in their duration time.