Fujibe (2007) reviewed urban climate research in Japan, and presented five formation factors of the urban heat island: (1) anthropogenic heat, (2) decreased evapotranspiration due to urbanization (especially removal of vegetation), (3) change of heat balance equation for urban canopy, (4) greenhouse effect due to air pollution, and (5) dynamic mixing due to buildings. Fujibe (2007) assumes the latter two formation factors of the urban heat island (4) and (5) are not determinate, although he concluded that the former three formation factors from (1) to (3) are correct. Although Fujibe's (2007) review is excellent, the above-mentioned factors were already described clearly in Oke's (1978) textbook, and are described from a viewpoint that is biased towards changes in the urban surface heat balance. The above factor (5), dynamic mixing due to buildings, is not a change in the urban surface heat balance, but considers changes in the thermal capacity of the urban boundary layer. The present paper reviews studies on the formation mechanism of the urban heat island in Japan with special emphasis on the relationships between heat island intensity and boundary layer of urban areas. As a result, it is pointed out that the different wind velocity dependency of the urban heat island intensity responds to the fact that the urban boundary layer is divided into the urban dome under calm conditions and the urban plume under windy conditions.
A review of urban climate studies using the Weather Research and Forecasting (WRF) model is presented. First, a brief description of WRF is presented. Second, three urban parameterization schemes in WRF are introduced. Third, urban heat island studies are summarized, including some results in related fields. Recent studies on urban rainfall are also summarized. Finally, some future issues are described.
In this study, observation and simulation data are analyzed to examine the causes of the rapid rise of surface air temperature in the inland area of northwestern Kanto Plain, Central Japan on 20 February 2009. A numerical simulation using the Weather Research and Forecasting (WRF) model shows that the temperature rise was mainly brought about by a downward wind with adiabatic heating over the area. Backward trajectory analysis indicates that the descending air parcel came from the southward warm area, in relation to the passage of a cyclone over the Kanto Plain. At about 1600 JST, the wind at the 850 hPa surface was divided into northward and southward by the central-Japan mountains, and converged over the northwestern Kanto area, which implied the formation of a downward wind. We suggest that the downward wind had the feature of a dry foehn, which was not accompanied by a precipitation process.
To investigate the spatial and temporal variability of the urban heat island, a high-spatial density meteorological observation system was set up in the Tokyo ward area by Tokyo Metropolitan Research Institute for Environmental Protection (TMRIEP) and Tokyo Metropolitan University from July 2002 to March 2005. The observation system was named Meteorological Environmental Temperature and Rainfall Observation System (METROS) and consisted of two observation networks named METROS20 and METROS100; METROS20 was made to observe meteorological factors (wind direction and speed, pressure, rainfall etc.) on the roofs of 20 buildings; METROS100 was made to observe temperature and humidity in instruments screens of 106 elementary schools. Since April 2005, observations of temperature and humidity were continued with the instruments screens of elementary schools by TMRIEP. This observation network was maintained until March 2010. Based on their observations, temporal and spatial characteristics of thermal environment of Tokyo have been investigated such as temperature range, especially in summer. For example, warmer areas differ between daytime and nighttime as shown by spatial patterns in rate of time exceeding 30 degrees Celsius and number of sultry nights: the warmer area is located from central part to northern part of the Tokyo ward area during daytime and from central part to coastal area during nighttime.
This paper describes our newly developed high-resolution temperature observational system called Extended-METROS, which has been deployed in the Tokyo Metropolitan Area since 2006. Some climatological mean temperature charts using Extended-METROS data are analyzed in terms of urban climatology, and detailed urban heat island temperature patterns are clarified. Rainfall measurements were set up from August 2010 at 40 points in the Tokyo Metropolitan Area. The relationship between urban heat islands and local-scale heavy rainfall patterns in urban areas will also be analyzed based on our high-resolution meteorological observation system.
We analyze the influence of sea breeze on temperature distribution in the Kanto Plain (central Japan) on a day that a sea breeze front was detected (known as sea-breeze front days) using high-resolution temperature data observed by our research team. The high-temperature area on sea breeze front days moves northwest from central Tokyo, and was located at Kawagoe city (middle Kanto Plain) at 14 JST, and the northern Kanto Plain at 16 to 18 JST, respectively. This high-temperature area appears at the head of the sea breeze front to the leeward of central Tokyo, where the daily maximum temperature is highest in Kawagoe city and the northern Kanto Plain. After the sea breeze front passes, the area where the temperature is higher than that at the circumference is distributed in the shape of a wedge. This wedge-shaped area is located to the leeward of central Tokyo where the wind from Tokyo Bay and Sagami Bay forms a convergence zone. The high-temperature area around Kawagoe city, which cannot be found on days with strong winds, is formed from the hindrance of cold air advection caused by sea breeze front penetration. On the other hand, high temperatures in the northern Kanto Plain may not be related to the penetration of sea breeze fronts, which do not reach the northern Kanto Plain on days when the daily maximum temperature is recorded. However, the temperature in the northern Kanto Plain is higher on sea breeze days than on strong southerly wind days, and this suggests that local circulation plays an important role in causing high temperatures in the northern Kanto.
This study presents the minute spatial structure of both the frequency of intense rainfall (data from the 1991 to 2002, except 1993, were used) and recent trends (15-25 years until 2002) in the special wards of the Tokyo Metropolis in summer (June to September), on the basis of hourly rainfall data from a dense rain-gauge network. As this is the first step in elucidating the relationship between the distribution of the frequency of intense rainfall and that of surface roughness in metropolitan Tokyo, the averaged number of building stories within a certain area, which is referred to as the smoothed building height (SBH), was assumed to be an alternative parameter when deciding surface roughness. The distribution of the ascending rate of SBH (hereafter, the ascending rate of SBH is referred to as ARS) for wind direction was calculated by varying the averaging area for SBH, in order to compare it to the distribution of intense rainfall frequency. The results are summarized as follows. The high-frequency areas of intense rainfall appear in the western to northern parts of the area comprising the wards and along the boundary between the Tokyo Metropolis and Saitama Prefecture. The frequency of intense rainfall in these areas is two to three times as high as that in the eastern part of the area comprising the wards. Moreover, the maximum areas of intense rainfall frequency are localized in the western, northern to northwestern, and southern part of the area comprising the wards, corresponding to wind direction. These areas are situated 3-5 km from the leeward side of the area, where the ARS derived from the SBH at a 1-2 km scale is large, that is, the vicinities of Shinjuku (SNJ), Ikebukuro (IKB), and Shibuya (SBY). Accordingly, we suggest that the large surface roughness due to high-rise buildings in the western part of the area comprising the wards has the effect of increasing the frequency of intense rainfall. The increasing trend of intense rainfall is clear in the western part of the area comprising the wards, whereas a decreasing trend, although not statistically significant, is seen in the eastern part of the area comprising the wards. It is noted that observational stations with large increasing trends of intense rainfall, such as Nakano (NKN) and Shinagawa (SNG), are located 3-5 km from the leeward side of SNJ and on the shore of Tokyo Bay in the southern part of the area comprising the wards, respectively, where the ARS for easterly winds derived from the SBH at a 1-2 km scale is large.
The mitigation of the urban heat island phenomenon by local circulations of land and sea breezes has recently become a subject of major interest. This study aims to clarify the effects of sea breezes on the urban heat island phenomenon in Sendai, Northeast Japan, paying attention to the vertical mixing effects of tall buildings. In Sendai city, where the urban heat island phenomenon has developed along with sea breeze circulation, spatially dense observations of temperature were carried out with instrument screens at twenty-five elementary schools in and around the urban area from 2000 to 2004. When a sea breeze begins to blow, the air temperature in the coastal region peaks and does not rise during daytime. By comparing the warming quantity during the day when sea breezes do not blow, the cooling effects of the sea breeze are evaluated quantitatively. It was found that cooling effects are remarkable in May and June, and disappear in September. Cooling effects in the urban center do not differ from those in the suburban area, in spite of dense buildings and large number of roughness parameters. Because of the mixing function, the large number of roughness parameters is considered to be useful to pull the cool air mass of sea breezes down to the ground. Vertical observations of wind and air temperature at the Miyagi Prefectural headquarters, which is located in the central business district (CBD), were carried out from July 2007 to July 2008. When sea breezes begins to blow, downward air currents were observed at the windward walls of buildings, and the cooling effects of sea breezes were identified gradually from the tops of tall buildings to the ground. The horizontal distributions of air temperature during the day with sea breezes produce relatively cool areas near the coast and in the urban center. The cooling effects of sea breezes appear to be more remarkably in the urban center than in the residential area where there are no tall buildings.
This paper describes the vertical structure of the nocturnal heat island in a small town. Tethered balloon soundings were conducted 180 times in and around an urban area on nights with light winds from March 2003 to May 2005. Surface inversions developed not only in rural areas but also in urban areas. Surface inversions developed quickly after sunset at the rural site, whereas development was slower at the urban site. At six hours after sunset the vertical temperature profile in the urban area was the same as that in the rural area. Heat island formation in the small town was related to the delay in the development of surface inversion in the urban area. The crossover effect can be seen in only three of 23 surveys for which vertical soundings were performed in both urban and rural areas.
To clarify the effects of mountain winds on urban temperatures, meteorological observations were conducted in the summers of 2008 and 2009 in Nagano City, central Japan. Hamada et al. (2006) showed that mountain winds blew along the Susobana River, extending into the center of Nagano City. This study selected Nagano Prefectural Office (NPO), which is near the center of Nagano City as a site to observe mountain winds blowing into the city, while Nagano Local Meteorological Observatory (NLMO) at the periphery of the city was selected as an observation site not to mountain winds. Wind conditions at NPO and NLMO were classified as days with (15 cases) and without (37 cases) mountain winds, respectively. Average air temperature, wind direction and speed, and relative humidity at the two points were compared for each case. As a result, at NPO, average wind speed was about 5 m/s, air temperature dropped, and relative humidity increased on days with mountain winds. On these days, air temperature was reduced by up to 2°C, with an average drop of 0.5°C. The drop in temperature with the mountain winds was negatively correlated (r = -0.46, p < 0.001) with wind speed, and the relationship depended on the strength of radiative cooling.
The mitigation effects of green spaces in urban areas are investigated with several microclimatic observations in Tokyo. On a clear calm night, flow-out wind directions from green spaces to surrounding areas were discerned at measuring points along the boundary with a sharp temperature fall of a few degrees. These imply the accumulation of a cold air mass in the park and its gravitational flow-out into the surrounding area. Under this condition, a significant air temperature drop in an adjacent built-up area was observed within a range of 80-90 m from the boundary in the case of Shinjuku-Gyoen Park. This limit was more or less fixed throughout the night, regardless of cool-island intensity. During the seeping out of cold air, cool-island intensity increases, but sensible heat flux was almost zero. The cooling ability of parks is not directly related to cool-island intensity. Such cold-air seeping-out phenomena are observed not only in large parks but also at a green-full apartment complex, a small green slope and even a stepped roof garden.
Synchronized kite-balloons measurements revealed the development of a nocturnal stable layer in an urban green park. The vertical structure of the park cool island is discussed. A stable layer developed in the park. In the surrounding urban area, neutral stratification was maintained all night. In spite of different weather conditions on the two nights measured, the park stable layer developed up to similar height (50 m). Three temperature profiles measured at a flat area in the park showed similar vertical gradients from surface to tree height (ca. 10 m). This means that the horizontal distribution of air temperature measured at the surface level represents the lowest 10 m of the atmospheric boundary layer. Calm, clear sky conditions triggered the development of a colder layer in the park than the surrounding area. The maximum height of the colder layer was found to be 71 m. The cooling energy for cold air formation over park was estimated to be 5 Wm-2.