SEPTEMBER 2013 Y O S H I K A D O 2065 Intense Summer Heat in Tokyo and Its Suburban Areas Related with Variation in the Synoptic-Scale Pressure Field: A Statistical Analysis HIROSHI YOSHIKADO Graduate School of Science and Engineering, Saitama University, Saitama, Japan (Manuscript received 21 November 2012, in final form 18 April 2013) ABSTRACT Intense summer heat in Tokyo, Japan, and its suburban areas between 1990 and 2010 was statistically analyzed. Sample days were selected from among days with a sea breeze and sufficient sunshine duration, because sea breeze is the dominant summertime meteorological system in the region. Excess in the daily maximum temperature of an inland urban site in Tokyo over a site facing the outer sea where the sea breeze originates was selected as a key index. A comparison of a group of days for which the temperature excess was large with an opposing group of days for which it was small revealed a distinct difference in diurnal wind variations: persistent southwesterly winds overwhelmed the sea breeze in areas facing the outer sea in the former case, whereas in the latter case a diurnal sea-breeze cycle was typically observed in those areas as well as in the inland areas. This difference could furthermore be connected to differences in the pressure gradient in the region, that is, differences in the synoptic-scale pressure field. As a result, slight shifts in the pressure field affect urban heat, suggesting that, in addition to general warming, changes in the pressure field resulting from future climate change can be of great importance. 1. Introduction effects remain at the present levels. The response of the urban climate to global warming requires analysis to Increased air temperature in urban areas compared cope with these problems. with surrounding rural areas is a well-known phenom- Urban climate is influenced by larger-scale meteoro- enon called the urban heat island (UHI). This phe- logical factors. First of all, global warming is expected to nomenon is most clearly observed in large cities. In the raise base temperatures in urban areas as in most regions case of Tokyo, the annual average temperature has in- 2 of the world. More specifically, however, global climate creased at a rate of 3.28C (100 yr) 1 for the period be- change is an aggregate of changes in synoptic-scale sys- tween 1931 and 2011 (JMA 2012). For the same period, tems such as pressure distributions and accompanying 17 observatories in remote small towns recorded an 2 weather characteristics in each season. These changes average temperature increase of 1.58C (100 yr) 1. This furthermore influence local wind patterns around urban increase is considered to be mainly a response to global areas of interest. Therefore, UHIs are influenced by climate change in the Japan region, and the difference in synoptic-scale meteorology through local meteorology. Tokyo’s temperature is attributed to the UHI effect. Many studies have been devoted to local winds and Special attention is paid to daily maximum tempera- thermal structures dominated by synoptic-scale systems. tures T in summer because the peak electric power max The Kanto Plain, with the Tokyo metropolitan area in demand for air conditioning is directly dependent on it. the southern part, has been the most important region in Also, there is a marked increase in the number of cases Japan for such studies. For example, Fujibe (1985) an- of heat stroke in hot summers. Global warming will alyzed deviations in local winds over the south Kanto possibly aggravate these circumstances, even if the UHI region under the influence of a synoptic-scale pressure gradient. Furthermore, Fujibe (1998) examined ex- tremely hot days in the Kanto Plain in relation to the Corresponding author address: Hiroshi Yoshikado, Graduate School of Science and Engineering, Saitama University, Saitama, pressure gradient. Nevertheless, few researchers have 338-8570, Japan. analyzed the two-stage reaction, between synoptic-scale E-mail: [email protected] and local-scale systems as the first stage and between DOI: 10.1175/JAMC-D-12-0315.1 Ó 2013 American Meteorological Society Unauthenticated | Downloaded 09/30/21 09:21 AM UTC 2066 JOURNAL OF APPLIED METEOROLOGY AND CLIMATOLOGY VOLUME 52 local-scale and urban-scale systems as the second stage, The present study statistically analyzes data observed specifically for the UHI around Tokyo. This is probably in Tokyo and its suburban areas with the aim to detect due to the preconception that urban-scale phenomena a two-stage relationship that connects synoptic-scale are not sufficiently clear to be detected under dynamic structure with intense summer heat through local-scale changes in larger-scale systems. In reality, however, the systems, while synoptic-scale structures predominant in UHI observed in the Tokyo area persists under the the future will be predicted by GCMs. Automated Me- predominant summertime sea-breeze system and in- teorological Data Acquisition System (AMeDAS) data fluences the downwind suburban climate (Yoshikado from the Japan Meteorological Agency (JMA) for 21 1990). Shifts in the prevailing pressure field under the summers between 1990 and 2010 are employed for this influence of global climate change may therefore distort purpose, with ‘‘summer’’ defined as July and August. major sea-breeze patterns, and furthermore may alter urban and suburban climate in a more intricate way than mere warming. 2. Methodology and data used Future urban climates under global warming are a. Hypotheses currently investigated using general circulation models (GCMs) and dynamical downscaling. Kusaka et al. Some factors and mechanisms can presumably in- (2012) made detailed climate predictions, namely tem- tensify summer heat in urban areas. First, air tempera- poral and spatial variations in temperature and other tures increase when a particularly hot synoptic-scale air meteorological factors, in major Japanese urban areas mass covers the area (hypothesis 1). Second, sunshine for the 2070s. They incorporated the best estimate cur- will heat land air more on clear days (hypothesis 2). It is rently available for future local climates by a pseudo– also empirically known that successive clear-sky days global warming method (PGW) developed by Kimura result in increased daily maximum temperatures (hy- and Kitoh (2007). The PGW uses averaged temperature pothesis 2a). These are direct effects of synoptic-scale rises predicted by GCMs instead of direct GCM output meteorology on urban climate. for future climate. Iizuka et al. (2010, 2011) applied The third factor is variations in local wind patterns, a similar approach to the Nagoya region. In their work, which contribute to the above-mentioned two-stage re- however, the magnitude of urban warming is mainly action and are of major interest in the present study. presented without discussion of the specifics of change Tokyo is located in the coastal area of the Kanto Plain, in synoptic-scale effects on local-scale or urban-scale and the sea breeze penetrates there mostly on fine climate. Analyzing this two-stage reaction with statis- summer days. As the sea breeze generally cools the tical reliability, even from past data, would improve coastal areas, a delay in its inflow can promote urban understanding of future summer heat. As an example heat. Note that the time and direction of the sea-breeze of the statistical analysis of synoptic-scale data to re- inflow depends on the synoptic-scale pressure gradient search local warming, Fujibe (1998) reported an in- (Fujibe 1985). Therefore, some pressure patterns are crease in the number of days characterized by an expected to promote urban heat, and other pressure extremely warm air mass and sufficient sunshine after patterns to moderate it (hypothesis 3). the 1980s as a cause of the increase of extremely hot We confine the analyses to days on which the sea days in the Kanto Plain. breeze developed in Tokyo and its suburban areas (SB Another insufficiency in the current modeling ap- days, below) as the most representative summertime proach, as well as in statistical analysis, is the lack of meteorological system, although another notable system a relationship between year-to-year variations in larger- promoting summer heat in the Kanto Plain has also been scale climate and urban summer heat. There were ir- investigated (Takane and Kusaka 2011). The relation regularly hot summers in the past several decades. between SB days and urban summer heat is clarified in Fujibe (1996) examined the hot summer experienced in section 3a. Japan in 1994 and concluded that UHI was a minor b. Region and data used factor in the intense heat observed that year. This im- plies that a larger-scale meteorological structure was the The Tokyo metropolitan area is most heavily urban- main cause. We suppose that not only synoptic-scale ized along the shore of Tokyo Bay and landward, mainly structures but also local systems developing under their to the northwest. Approximately 18 million people live influence, such as land and sea breezes, play an impor- in this region. Downtown Tokyo is located on the tant role, and thus that combined synoptic-scale and northwest coast of the bay, where three AMeDAS sta- local-scale structures will dominate the intensity of tions are located (see Fig. 1): Shinkiba near the shore at summer heat every summer, and ultimately every day. the north end of the bay, Tokyo in the central area about Unauthenticated | Downloaded 09/30/21 09:21 AM UTC SEPTEMBER 2013 Y O S H I K A D O 2067 FIG. 2. Numbers of days in the groups of days defined in the study. There are exceptional sites in urbanized areas, which FIG. 1. The southern Kanto region and targeted AMeDAS sta- are not surrounded by wide fields.