Introduction
For land-use planning in urban and regional scales, which preserves better atmospheric environment such as high air quality and comfortable climate, it is important to know the characteristics of local wind and temperature fields in the region and the various topographic factors determining the wind and temperature. Most large Japanese cities are located in coastal regions and have rapidly expanded in the last three decades. People living in the region now experience a warmer and unpleasant thermal environment in the summer season, supposedly because of the extensive urbanization. It is interesting that, under fine weather with light synoptic-scale gradient wind, the highest temperatures are often observed in a rather less urbanized inland area, which is at the downstream side of the highly urbanized coastal zone in a sea-breeze situation. For example, the cities of Koshigaya and Urawa, in the Kanto Plain (Kimura and Takahashi 1991; Fujino et al. 1993), and Ichinomiya and Inuyama, in the Nohbi Plain (Kitada et al. 1991a; Kitada et al. 1992), are all located north of the highly urbanized areas of Tokyo and Nagoya, respectively, and those cities show the highest temperature in each plain area in the above-described situation.
In this study, by using a mesoscale meteorological model with a k−ε turbulence model (Kitada 1987; Kitada et al. 1991b; Takagi and Kitada 1994, 1996), we have investigated the effects of natural topography and human-modified land use on the characteristic features of wind and temperature observed on a typical land–sea-breeze day under a light gradient wind in the Nohbi Plain and Ise Bay area, central Japan (see Figs. 1a,b), in the warm season. Especially, effects of the natural topography, ranging from local coastline and small hills to the Japanese Alps, and the modified land use, such as an urban area and rice paddies, have been focused upon. Using the simulation results, we will show that the long-lasting sea breeze in the Nohbi Plain can be numerically reproduced under the influence of the Japanese Alps and will clarify the mechanism of the formation of high-temperature zone in inland suburbs in a sea-breeze situation. Previous studies related to the present subjects includes those by Kondo (1990) and Kimura and Takahashi (1991).
Numerical model and calculation domain
Governing equations
Boundary and initial conditions
No synoptic-scale pressure gradient was assumed during the simulation period. Initial wind velocity was set at zero everywhere. Initial potential temperature at the top boundary, 6.5 km high above the mean sea level, was set constant at 323.5 K. That potential temperature at the top was decreased with a constant rate of 5.5 K km−1 to the level of the fourth vertical grid, approximately 30 m above the ground surface. Below the fourth grid, the potential temperature was linearly interpolated to the observation-derived value at 1.5 m above the ground at 0700 LST. Temperature at the sea surface layer was estimated using observed data, which were taken during the cruise of the Asama-maru near the mouth of the Ise Bay on 13 May 1985 (Mie Prefectural Marine Technology Center 1985). The temperature varied between 18.6° and 20.0°C during the simulation. Relatively higher temperatures were observed during the daytime.
Calculation domain and numerical method
Two kinds of calculation domains were used for the simulations, the aims of which will be described in the next section 2d. Figure 1a stands for a narrower domain and Fig. 1b for a broader domain. The area boxed with a solid line in Fig. 1b corresponds to the region of Fig. 1a. Vertical depth of the domain is 6.5 km above mean sea level for both narrow and broad regions.
A staggered grid system is used; Θ, P, ρ, k, and ε are defined at main grid points, while U, V, and W at grid points shifted by half of the grid size from the main grid point along the coordinate curves ξ, η, and σ, respectively. Variable grid size is used for both vertical and horizontal directions. For the vertical direction, the grid size is small in the atmospheric boundary layer and is made coarser above it; the total number of grid points is 36. For the horizontal direction, uniform grid sizes of 2.29 km for ξ direction and 1.85 km for η direction are used for the narrow domain in Fig. 1a and also for the boxed area in Fig. 1b; for the broad domain shown in Fig. 1b, the grid size is increased gradually from around 2 km for the boxed area to about a maximum of 11.4 km outside of the area; the numbers of grid points are 41 × 71 and 88 × 113 for the domains in Figs. 1a and 1b, respectively. Depth of the underground region is 1 m from the ground surface; the total number of vertical grids is 13; the variable grid size is used with a minimum of 0.47 cm to a maximum of 32.25 cm.
The finite difference method was adopted for discretization of the governing equations. Spatial derivatives of advection and diffusion terms were discretized with the power-law scheme (Patankar 1980), which is nearly equal to the centered difference when the cell Reynolds number, for example, (U1Δx1)/νt, is smaller than 0.5 and is equal to the upwind difference when that number is larger than 6. For time integration, a fully implicit method is used with a time step of 30 s. The method of SOR (successive over relaxation) is applied for the solution of a set of linear algebraic equations.
Simulation cases
Three simulations were conducted (see Table 1). Case 1 uses the narrow domain (Fig. 1a) without “city”; case 2 uses the broad domain (Fig. 1b) without city; case 3 uses the broad domain with city. “With city” means to adopt realistic land use for Nohbi Plain, which is shown in Fig. 2 and represents the situation in 1985, while the“without city” case assumes “forest” for the whole land area. Comparison between cases 1 and 2 should show the effects of large-scale topography, such as the Japanese Alps on the meteorological quantities over the Nohbi Plain. Similarly, differences between cases 2 and 3 should reflect the effects of change of the human-modified land use, that is, urbanization and agricultural land use. In Fig. 2 the “water” over land area indicates the rice paddy field primarily and additionally the river, water channel, and pond. Table 2 lists surface parameters assigned for each land-use type; the parameters are used for the calculation of fluxes within the surface layer. The values of albedo and emissivity in Table 2 were determined mostly by considering those listed in Oke (1978). For example, the albedo for forest was assumed to be 0.2, which is assigned for leaved deciduous forests in Oke (1978). For city, the albedo was set at 0.15, which is based on the value estimated for the cities of Nagoya, Gifu, and Takayama in central Japan (Nomoto 1991). The Bowen ratio for forest was assumed to be 0.5 and was estimated from the observations in central Japan in the summer of 1995 (Kondo et al. 1996). For the surfaces of “garden” and city, the values of the Bowen ratio were assumed to be 1.5 and 3.0, respectively, using information in Oke (1982). Roughness length for “water” was set at 0.01 m. The water surface represents both the rice paddy field and sea surface. The values of 0.01 m might be a little large for open sea surface, that is, nearly the largest end of the roughness length for sea surface listed in Davenport (1982).
Each simulation was started at 0700 LST without a synoptic-scale pressure gradient and was continued for two days. Calculation results from the second day will mainly be discussed in a later section.
Characteristics of local winds in the Nohbi Plain–Ise Bay area
When the central part of Japan is covered by an anticyclone with a weak pressure gradient in the warm season, characteristic local winds develop over the Nohbi Plain; the flow pattern at surface level shows three stages from morning to midnight (Kitada et al. 1991a; Mori et al. 1994). At the first stage up to 1100 LST, local winds such as sea breeze, valley wind, and upslope wind blow just locally under the influence of local topography. At the second stage, these local winds are organized to “southwesterly,” which we call the “Ise Bay sea breeze” over the central Nohbi Plain (Fig. 3a shows the flow field at this stage). Finally at the third stage from late afternoon to midnight, via a transition stage at around 1500 LST shown in Fig. 3b, the sea breezes over the plain turn to “southeasterly” and are dominated by those from the Pacific Ocean, that is, the “Enshu-Nada sea breeze” (Figs. 3c,d correspond to this stage at 1800 and 2100 LST, respectively). The sea breeze at this third stage, especially after sunset, is somewhat strange, since the temperature at surface level over the Nohbi Plain is lower than that over sea surface, and the situation is not preferable for the sea breeze. This suggests that the sea breeze blowing even at night may be maintained by a rather large-scale topographic effect. In the next section we will show this sea breeze is significantly affected by the Japanese Alps, which is the largest topographic feature in the region.
Effect of large-scale mountains on local flows
By comparing the results of case 1 (Fig. 1a for its calculation domain) with those of case 2 (Fig. 1b for the domain), we have tried to elucidate the effects of various topographic scales on the flow and temperature over the Nohbi Plain; the smaller-scale topography includes Ise Bay and the mountains just surrounding the plain area, while the larger-scale topography includes the Japanese Alps and the Pacific Ocean.
Figures 4a and 4b show computed flow and potential temperature at 10 m above the ground over the Nohbi Plain at 1500 LST for cases 1 and 2, respectively. Figure 4c is the same as Fig. 4b but for the whole area, and Fig. 4d is also for the whole area but for case 3 and will be discussed in a later section. The flow patterns over the Nohbi Plain shown in Figs. 4a,b are quite similar to each other, and both cases 1 and 2 qualitatively reproduce observations in Fig. 3b. However, a close investigation shows that wind velocities in Fig. 4a (case 1) are weaker in both southern and northeastern parts of the domain compared to those in Fig. 4b (case 2). The observations in Fig. 3b support case 2 results (Fig. 4b), that is, strong winds over the Chita and Atsumi peninsulas in the southern part of the domain near the Pacific Ocean. To see this more quantitatively, Fig. 5 illustrates comparison of wind vectors at 1500 LST between observations and case 1 and 2 simulations at three sites over the Ise Bay, and the Chita and Atsumi peninsulas, the locations of which are marked with open circles in Fig. 3b. Figure 5 shows wind vectors calculated in case 2 agree much better with observation than those in case 1 do.
Figures 6a and 6b show computed surface winds and potential temperature at 10 m above the ground at 2100 LST for cases 1 and 2, respectively. Figure 6b is the partial area of the whole wind field shown in Fig. 6c; Fig. 6d is the same as Fig. 6c but for case 3 and will be discussed later. At 2100 LST the observed wind field is at its third stage and shows clear southeasterly flow over the Nohbi Plain, as in Fig. 3d. A comparison between Figs. 6a (case 1) and 6b (case 2) indicates that the winds in Fig. 6b simulate the observed flow in Fig. 3d much better than those in Fig. 6a. The reason may be that the high mountains of the Japanese Alps, which are included in case 2 but not in case 1, cause plain–plateau circulating flows, then the weak subsidence associated with the return flow warms the air mass over the Nohbi Plain during daytime, and the warmed air contributes to maintain the pressure gradient at the surface level that drives the southeasterly sea breeze even after sunset. To see this, vertical profiles of pressure differences between site D and sites A, B, and C (see Fig. 4a for locations of sites A, B, C, and D) are plotted in Figs. 7a (1500 LST) and 7b (2100 LST) for case 1 and in Figs. 8a (1500 LST) and 8b (2100 LST) for case 2. The site D represents “ocean,” A and B stand for northwest and northeast corners of the Nohbi Plain, and C is the south end of the plain and represents the location of Nagoya, which is the largest city in the plain. The sites A, B, and C were chosen conveniently to see the effect of the pressure gradient on the two important sea breezes in the Nohbi Plain, that is, southwesterly from the Ise Bay and southeasterly from the Enshu-Nada, the Pacific Ocean. In these figures, for example, the profile at A shows the pressure difference, that is, the pressure at A minus that at D. Thus, the negative number indicates that the pressure gradient is suitable for the sea breeze, that is, wind from D to A. Figure 8b shows that the pressure gradients between D and A and also D and C in the lower layer below 1 km in altitude still support sea breeze, while those in Fig. 7b do not. In addition, comparison of the profiles between A and B (or C and B) in Fig. 8b indicates that the pressure at B is higher than those at A and C near surface level and thus only a southeasterly flow is allowed over the Nohbi Plain, as shown in Figs. 6b and 3d (observation). To see the difference of warming of the air mass over the Nohbi Plain between cases 1 and 2, vertical profiles of the difference of potential temperature between cases 1 and 2 are plotted in Figs. 9a (1500 LST) and 9b (2100 LST);in these figures, for example, the profile at A shows ΔΘ.DIFF = (ΔΘ for case 2 − ΔΘ for case 1) at A, where ΔΘ is (potential temperature at A − that at D);thus positive ΔΘ.DIFF denotes that air of the corresponding height at A is warmer in case 2 than in case 1. Figure 9 indicates that at 1500 LST (see Fig. 9a), air mass over the Nohbi Plain below 2 km above the ground is warmed much more in case 2 than in case 1 and that the warmer air in case 2 still remains at 2100 LST (see Fig. 9b). Thus the profiles in Fig. 9 suggest that the warming of the air mass over the Nohbi Plain in case 2 is due to subsidence associated with the plain–plateau circulation caused by the larger topographic feature of the Japanese Alps. Figure 10 shows the north–south vertical cross section at x = 176 km (see Fig. 1b) of wind vectors and potential temperature at 1500 LST in case 2. In Fig. 10, the “narrow” region for case 1 is shown by two upward arrows on the y axis, and the location of Nagoya is also indicated. Clear return flow and subsidence are found in the layer between 1 and 2 km high over the “narrow region” in Fig. 10. The potential temperature contour over this region below 2 km high shows that the return flow and subsidence contribute to the warming of air mass over the Nagoya area. These flow features could not be found or be much weaker without high mountains located at y = 300 km.
Effects of urbanization on temperature and flow fields
Temperature fields: Formation of high temperature zone
By analyzing observed temperature and land-use distributions during 10 years from 1975 to 1985, it was found in Kitada et al. (1991a) that, on a typical sea-breeze day in the warm season, the urbanization extended during the decade in the coastal area of the Nohbi Plain caused an increase in the daily maximum temperature in the inland area. The inland area is located north-northeast of greater Nagoya (see Fig. 1) and downwind in the sea breeze. The zone of the highest daily maximum temperature moved from the highly urbanized city center of Nagoya in 1975 to its inland suburbs in 1985.
We have investigated the reason for this by comparing results of cases 2 and 3 (see Table 1). Figure 11 illustrates observation points for temperature in the Nohbi Plain that are routinely operated by the Aichi Prefectural Government, and contours drawn using these temperature data will later be compared to numerical simulations. In the warm season from April to October, central Japan usually has around a total of 70 days for typical sea breeze (Mori et al. 1994). The day of 17 May 1985 was such a typical sea-breeze day with weak pressure gradient in the synoptic scale, whose surface wind is already shown in Fig. 3. The target day for qualitative comparison with the present simulation was 17 May 1985. Although the detailed tuning to simulate the meteorology of that particular day was not done, the basic vertical profiles of potential temperature and sea surface temperature were determined, as already mentioned, using information on that day. Figures 12a–c show computed wind and temperature for case 2 (i.e., without city) at 3 m above the ground at 1200, 1300, and 1400 LST, respectively. Similarly, Figs. 13a–c are for case 3 (i.e., with city). Clear differences between Figs. 12 and 13 are that the potential temperature for case 3 (Fig. 13) shows its local maximum at 1200 and 1300 LST, which is located northeast of the highly urbanized area of the greater Nagoya, and this local maximum moves northeastward, finally leaving the region at 1400 LST. Another feature, which can be seen in Fig. 13 and not in Fig. 12, is the existence of a very weak wind zone in the downwind side of the local maximum of potential temperature. This local maximum of potential temperature may be what we previously found from observation data acquired in 1985 (Kitada et al. 1991a 1992). To verify this, Fig. 14 compares contours of observation-derived temperatures (on 17 May 1985) with those of computed temperatures (not potential temperatures) at 1200, 1300, and 1400 LST, where the observation derived is expressed in contour lines, and the computed is in different colors. Local maxima can be found in both observed and computed temperatures (Figs. 14a–c), and locations of the maxima and their magnitudes in observed and computed temperatures coincide well with each other at each time, though the observation points (see Fig. 11) are concentrated mostly in the central part of the domain and thus the contours of observed temperatures likely include errors in the marginal area in the domain. As a result, it can be judged that the case 3 simulation reproduces well the characteristics of real temperature distribution over the Nohbi Plain on a typical “sea-breeze day” in May 1985; as mentioned previously, the case 3 simulation uses land use and anthropogenic heat sources for the Nohbi Plain in May 1985.
To see the vertical structure of the above high-temperature zone by using the case 3 simulation, we have plotted cross sections of computed wind and potential temperature along a diagonal line (see Fig. 11) at 1200, 1300, and 1400 LST in Figs. 15a–c, respectively. The mixed layer develops over the area of the weak surface wind (Figs. 15a,b), the area that appears ahead of the sea-breeze front. To show this more clearly, temporal evolution of the vertical profile of potential temperature at Y = 110.88 km in Fig. 15 is illustrated in Fig. 16. The location of the point is marked with a thick solid upward arrow in Figs. 15a–c and also with a solid circle in Figs. 13a–c. Figure 16 indicates that height of the mixed layer rapidly increases from 750 m at 1100 LST to around 1400 m, when the point at Y = 110.88 km comes into the weak wind zone at 1200 LST (see Fig. 13a). Then, after the sea-breeze front arrived at the point around 1300 LST as shown in Figs. 13b and 15b, the mixed-layer height decreases to 650 m at 1400 LST (see Fig. 16). The shape of the profile of potential temperature at 1300 LST in Fig. 16 may represent the transition state to the sea-breeze-dominated situation. Figures 13, 15, and 16 suggest the mechanism with which temperature is raised at the downwind side of an urban area in a sea-breeze situation: 1) Once an urban area is placed as a heat source in the sea-breeze/valley wind induced by topography, then the urban area tends to form a pressure gradient adverse to the sea-breeze/valley wind at its inland side, and thus weakens both the advancing speed of a sea-breeze front and also the valley wind; 2) the mixed layer strongly develops over this weak wind convergence area and warms the air mass. Hence, the high-temperature zone forms, and 3) this high-temperature zone moves inland with penetration of a sea breeze. Figure 17a shows a vertical cross section of potential temperature difference between case 3 (with city) and case 2 (without city) at 1200 LST (that is, Θ for case 3 − Θ for case 2), and similarly Fig. 17b for the wind velocity difference. The area of the largest heating associated with city shown in Fig. 17b coincides with the weak wind area in Figs. 13a and 17b. Thus, these figures also support the above explanation for the formation of the inland high-temperature zone.
Another reason for the formation of the highest temperature zone at the downstream side of the highly urbanized area in a sea-breeze situation is horizontal heat transport from the urbanized area by the sea breeze; the sea breeze that has traveled over the urbanized area for a long time accumulates heat within its layer and results in the highest temperature at the downwind side of the urban area. Figure 18 shows the vertical cross section of potential temperature difference between case 3 and case 2 as in Fig. 17a but for 1400 LST. The sea-breeze front at 1400 LST has almost passed over the domain, and in addition there is no urbanized area beyond 115 km; nevertheless, high temperatures that cannot be found in case 2 are observed at around 125 km (see contour line of 1.6°C in Fig. 18), demonstrating possible formation of the highest temperature zone caused by horizontal heat transport from the urban area.
Influence of rice paddy fields on flow and temperature
Rice paddy fields extend in the western part of the Nohbi Plain, and they usually hold water from mid-May to early September; thus, they are illustrated as inland water in Fig. 2. Effects of these rice paddy fields on flow and temperature have been investigated by comparing case 2 and case 3 results. Figures 19a and 19b show horizontal distributions of the differences of potential temperature and wind velocity at 1500 LST between cases 2 and 3, respectively; the differences are defined as (Θ for case 3 − Θ for case 2) and (wind velocity for case 3 − wind velocity for case 2). Potential temperatures over the rice paddy in case 3 are lower compared to those over the forest in case 2, indicated by −0.4° in Fig. 19a. These lower temperatures are due to both the thermal nature of the water surface of the rice paddy and the aerodynamic nature of the relatively small roughness length. In the simulations, we assumed the Bowen ratio to be smaller for water surface (rice paddy field) than for forest (see Table 2). Thus, the local thermal property will work for lower potential temperatures over the rice paddy in case 3. However, another important factor for the lower temperatures is that fast penetration of the sea-breeze front and larger sea-breeze velocity, over the relatively smooth surface of the rice paddy (see Table 2), brought cooler marine air over the Ise Bay deep into the inland area. Figure 19b shows faster wind velocity over the rice paddy in case 3 and thus supports the above explanation. The same large sea-breeze velocity over the rice paddy gives an opposite effect on temperature at night. Figure 20a is the same as Fig. 19a but for 2100 LST and shows, in contrast to Fig. 19a, a higher temperature for the rice paddy. As shown in Table 2, the Bowen ratio for all surfaces over land area were assumed to be equal during nighttime. Then this higher temperature is largely due to the transport of warmer marine air by a fast sea breeze over the smooth rice paddy fields (see Fig. 20b for higher wind velocity over the rice paddy).
Flow fields
As discussed above, change of land-use type in the regional scale, for example, 80 km × 100 km, significantly modifies wind and temperature in that scale. However, the characteristic diurnal pattern of winds over the Nohbi Plain, such as the three stages of the sea breeze discussed earlier, seems to be unchanged by the land-use modification (Figs. 4c,d and Figs. 6c,d). Figures 4c and 4d compares winds at 10 m above the ground in the whole domain at 1500 LST in cases 2 and 3 and also Figs. 6c,d at 2100 LST. These figures demonstrate that natural topography dominates local flows over the Nohbi Plain, though some enhancement of wind speed affected by the land use over the plain can be seen in Figs. 4d and 6d (case 3), as discussed in the previous section.
Conclusions
The influence of natural topography in various spatial scales, which range from the local coastline and small hills to the Pacific Ocean and the Japanese Alps, and human-modified land use such as the urbanization and rice paddy fields, on flow and temperature fields over the Nohbi Plain of central Japan was investigated utilizing a mesoscale meteorological model with a k−ε turbulence model. Conclusions based on the simulation results are as follows.
First, the relative importance of natural topography in various spatial scales and human-modified land use for the formation of characteristic diurnal patterns of wind and temperature in the plain area has been clarified. The Japanese Alps, which is the largest topographic feature in central Japan and often called the roof of Japan, gave the most important influence on the wind at surface level, although the mountains are located quite far, around 100 to 200 km, from the Nohbi Plain. The effect of the high mountains on the flow over the Nohbi Plain was caused by the weak subsidence associated with the return flow of the plain–plateau circulation; the subsidence has a warmed air mass over the Nohbi Plain below 2 km in altitude during the daytime; then this warmed air has contributed to maintain the pressure gradient force directed from ocean to land from sunset until about midnight.
The urbanization in the coastal Nohbi Plain showed little effect on the characteristic diurnal flow pattern, which was affected most by the large-scale topography, as described above. However, local flow velocity and temperature over the Nohbi Plain have been largely affected by the change of land use.
Second, under the high pressure system with light gradient winds and sunny skies in the summer season, it is known that the extensive urbanization in the coastal area such as the Kanto Plain and the Nohbi Plain resulted in the shift of the highest temperature zone from the city center to the inland suburbs (Kitada et al. 1991a;Kimura and Takahashi 1991).
This formation of the inland high-temperature zone due to coastal urbanization in the sea-breeze situation can be explained as follows: 1) an urban area, that is, a local heat source due to its surface nature and anthropogenic energy use, once placed in the topographically induced sea-breeze/valley wind causes a weak wind zone at the downwind side of the urban area due to pressure gradients adverse to the sea-breeze/valley wind; 2) in this weakwind convergence zone, the mixed layer rapidly develops and the air mass over the zone is strongly heated from the ground surface before the arrival of the sea breeze; and 3) this high temperature area moves ahead of a sea-breeze front (Fig. 15).
Another reason is the horizontal heat transport by the sea breeze that is heated during its passage over the coastal urbanized area. By the two mechanisms, the inland high-temperature zone forms.
Third, the rice paddy fields, which extend from the coast to the inland along rivers such as the Kiso and Ibi Rivers in the western part of the Nohbi Plain (see Figs. 1 and 2), contribute to cool inland air during daytime and to warm it at night in the “May” situation. These are due to the smooth surface (low roughness length) of the rice paddy as well as its thermal property; because of low surface resistance, the sea breeze has quickly transported marine air over the Ise Bay deep into the inland area where the marine air is cooler during the daytime and warmer at night than the inland air.
Acknowledgments
This work was supported in part by the Ministry of Education, Culture, and Science of Japan through Grants 06680490, 06302051, and 08650645.
REFERENCES
Davenport, A. G., 1982: The interaction of wind and structures. Engineering Meteorology, E. Plate, Ed., Elsevier, 527–572.
Fujino, T., T. Asaeda, A. Wake, and Y. Meng, 1993: Characteristic of suburban heat island with an example of northern part of Tokyo (in Japanese). Annu. J. Hydraul. Eng.,37, 59–64.
Kimura, F., and S. Takahashi, 1991: The effects of land-use and anthropogenic heating on the surface temperature in the Tokyo metropolitan area: A numerical experiment. Atmos. Environ.,25B, 155–164.
Kitada, T., 1987: Turbulence structure of sea breeze front and its implication in air pollution transport—Application of k−ε turbulence model. Bound.-Layer Meteor.,41, 217–239.
——, K. Kunii, and S. Kubota, 1991a: Effects of urbanization on the climate and air quality in regional-scale: An analysis of historical data in 1975 and 1985 in Nohbi plain, central Japan (in Japanese). Proc. Environ. Eng. Res.,27, 117–227.
——, H. Takagi, K. Kunii, and H. Kato, 1991b: Numerical investigation of the coastal atmospheric environment influenced by small-scale peninsula. Energy Build.,15/16, 979–992.
——, S. Kubota, and K. Kunii, 1992: Numerical analysis of the effects of change of land use on temperature distribution in Nohbi plain (in Japanese). Preprints, Autumn Meeting of Meteorological Society of Japan, Sapporo, Japan, Meteorological Society of Japan, 284 pp.
Kondo, H., 1990: A numerical experiment on the interaction between sea breeze and valley wind to generate the so-called “Extended Sea Breeze.” J. Meteor. Soc. Japan,68, 435–446.
——, S. Yamamoto, and S. Murayama, 1996: An observational study of heat budget in the deciduous temperate forest over the complex terrain (in Japanese). J. Nat. Inst. Resour. Environ.,5, 27–37.
Launder, B. E., and D. B. Spalding, 1974: The numerical computation of turbulent flow. Comp. Methods Appl. Mech. Eng.,3, 269–289.
Mie Prefectural Marine Technology Center, 1985: Observation record for Ise Bay in May 1985. 3 pp. [Available from Mie Prefectural Marine Technology Center, 3564-3 Hamajima, Hamajima-cho, Shima-gun, Mie Prefecture 517-0404, Japan.].
Mori, H., H. Ogawa, and T. Kitada, 1994: Characteristics of land and sea breezes in the Nohbi plain and the conditions of occurrence of the “extended sea breeze” (in Japanese). Tenki,41, 379–385.
Nakamichi, K. 1992: Decadal trend (1975–1985) of anthropogenic heat sources in the Nohbi Plain and its impact on the atmospheric environment (in Japanese). B.S. thesis, Toyashi University of Technology, 37 pp.
Nomoto, S., 1991: Distribution of surface albedo in and around Nagoya, Gifu and Takayama cities (in Japanese). Geographic Information Systems for Environmental Change in Modern Japan, Tech. Report of Grant-in-Aid for Scientific Research on Priority Areas 101, Ministry of Education, Science and Culture, Tokyo, Japan, 318 pp.
Oke, T. R., 1978: Boundary Layer Climates. Methuen and Co., 372 pp.
——, 1982: The energetic basis of the urban heat island. Quart. J. Roy. Meteor. Soc.,108, 1–24.
Panofsky, H. A., and J. A. Dutton, 1984: Atmospheric Turbulence. John Wiley and Sons, 397 pp.
Patankar, S. V., 1980: Numerical Heat Transfer and Fluid Flow. Hemisphere, 197 pp.
Pielke, R., 1974: A three-dimensional numerical model of the sea breezes over south Florida. Mon. Wea. Rev.,102, 115–139.
Rodi, W., 1985: Calculation of stably stratified shear-layer flows with a buoyancy-extended k−ε turbulence model. Turbulence and Diffusion in Stable Environments, J. C. R. Hunt, Ed., Oxford University Press, 112–143.
Sha, W., T. Kawamura, and H. Ueda, 1991: A numerical study on sea/land breezes as a gravity current: Kelvin–Helmholtz billows and inland penetration of the sea-breeze front. J. Atmos. Sci.,48, 1649–1665.
Swinbank, W. C., 1963: Long-wave radiation from clear skies. Quart. J. Roy. Meteor. Soc.,89, 339–348.
Takagi, H., and T. Kitada, 1994: Vertical profiles of turbulent kinetic energy observed with Doppler sodar and their analysis using k−ε turbulence model (in Japanese). Tenki,41, 827–846.
——, and ——, 1996: Transport of turbulent kinetic energy generated over small hills in a sea breeze—A numerical simulation with a k−ε turbulence model (in Japanese). Tenki,43, 289–302.
(a) Calculation domain for case 1. Latitude and longitude at the southwest and northeast corners are 34.5°N, 136.4°E and 35.7°N, 137.4°E, respectively. Minimum contour value is 100 m, and the contour interval is 100 m. (b) As in Fig. 1a but for cases 2 and 3. Southwest corner: 33.4°N, 134.9°E; northeast corner: 37.7°N, 141.3°E. The area boxed with a solid line corresponds to the area shown in Fig. 1a. Minimum contour value is 250 m, and the contour interval is 250 m.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Land use in the Nohbi Plain/Ise Bay area. Applied for case 3 simulation (see Table 1). Purple: sea, blue: inland water, green: forest, yellow: garden, red: city.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Observed surface winds for a typical land- and sea-breeze day: 17 May 1985 (Mori et al. 1994): (a) 1200, (b) 1500, (c) 1800, and (d) 2100 LST. Open circles in Fig. 3b denote the sites over the Ise Bay and Chita and Atsumi peninsulas from left to right; wind vectors are discussed in the text and plotted in Fig. 5.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Computed winds and potential temperature (K) at a height of 10 m above ground at 1500 LST: (a) case 1, (b) case 2 (partial area:the area boxed in Fig. 1b), (c) case 2 (whole area: the area shown in Fig. 1b), and (d) case 3 (whole area: the area shown in Fig. 1b). The contour interval is 1 K. Symbols A, B, C, and D in Fig. 4a show key locations where vertical profiles of potential temperature and pressure are discussed in text and Figs. 7, 8, and 9. Contour interval for potential temperature is 1 K.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
An example of the comparison between observed and simulation-derived wind vectors at 1500 LST at sites over the Ise Bay and Chita, and Atsumi peninsulas, the locations of which are shown with open circles in Fig. 3b. The letter “O” denotes observation, and“1” and “2” the results of cases 1 and 2, respectively (see Table 1 for cases 1 and 2).
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Same as in Fig. 4 but for 2100 LST.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Vertical profiles of pressure difference, ΔP, among the points A, B, or C, and point D at (a) 1500 and (b) 2100 LST for case 1. See Fig. 4a for the points’ locations. The profile A denotes ΔP at A; that is, pressure at A − pressure at D. Similarly, ΔP is shown at B and C, respectively.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Same as in Fig. 7 but for case 2.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Vertical profiles of the difference of potential temperatures, ΔΘ.DIFF, between case 2 and case 1 at points A, B, and C at (a) 1500 and (b) 2100 LST. The profile A denotes ΔΘ.DIFF at A, that is, [(Θ at A − Θ at D) for case 2 − (Θ at A − Θ at D)] for case 1. Similarly, ΔΘ.DIFF is shown at B and C, respectively.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
North–south vertical cross section at x = 176 km (see Fig. 1b) of wind vectors and potential temperatures at 1500 LST in case 2. The region surrounded by the two upward arrows on the Y axis corresponds to the “narrow” area in Fig. 1a.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Sites for the meteorological observation, that is, the temperature and wind at surface level. Vertical cross sections of computed temperature and wind on the diagonal line are discussed in Figs. 15, 17, and 18.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Horizontal distribution of computed temperature and wind at the surface level for case 2: (a) 1200, (b) 1300, and (c) 1400 LST.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Same as in Fig. 12 but for case 3. Solid circles denote the location at which temporal evolution of the vertical profile of potential temperature is analyzed, as discussed in the text and plotted in Fig. 16.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Horizontal distributions of the observation-derived (contour lines) and the computed temperatures (the colors, case 3) at around 3 m above the ground: (a) 1200, (b) 1300, and (c) 1400 LST.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Vertical cross sections of computed potential temperature and winds on the diagonal line in Fig. 11 for case 3: (a) 1200, (b) 1300, and (c) 1400 LST. Thick solid arrows denote the location at which temporal evolution of the vertical profile of potential temperature is analyzed, as discussed in the text and plotted in Fig. 16.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Temporal evolution of the computed vertical profile of potential temperature (case 3) at the site marked with a solid circle in Fig. 13 and with thick solid upward arrows in Fig. 15. The sea-breeze arrival at this site at 1300 LST can be found in the change of the profile.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Vertical cross sections, on the diagonal line in Fig. 11, of the differences between (a) potential temperatures for cases 3 and 2, that is, Θ for case 3 − Θ for case 2, and (b) wind velocities for cases 3 and 2.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Same as in Fig. 17a but for 1400 LST.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Horizontal distributions of the differences of the potential temperatures and the wind velocities at 1500 LST between cases 2 and 3: (a) potential temperature difference and (b) wind velocity difference.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Same as in Fig. 19 but for 2100 LST.
Citation: Journal of Applied Meteorology 37, 10; 10.1175/1520-0450(1998)037<1026:EOTAUO>2.0.CO;2
Simulation case.
Land-use parameters.
