• 検索結果がありません。

Observation of downburst event in Gunma prefecture on August 11, 2013 using a surface dense observation network

N/A
N/A
Protected

Academic year: 2021

シェア "Observation of downburst event in Gunma prefecture on August 11, 2013 using a surface dense observation network"

Copied!
11
0
0

読み込み中.... (全文を見る)

全文

(1)

Observation of downburst event in Gunma prefecture on August 11,

2013 using a surface dense observation network

Keiko Norose1, Fumiaki Kobayashi1, Hirotaka Kure2, Takuya Yada2, Hiroyuki Iwasaki3 1

National Defense Academy, Yokosuka, Japan,

2

Meisei Electric Co., Ltd, Isesaki, Japan

3

Gunma University, Maebashi, Japan

Abstract. On the evening of August 11, 2013, a severe thunderstorm passed over Takasaki and Maebashi City in Gunma Prefecture, Japan, causing extensive wind damage due to gusts. Changes in surface weather elements associated with the gusts were recorded by the dense network of surface meteorological observation stations (“POTEKA”), which the compact weather stations were set up at primary schools and convenience stores. We examined the development and propagation of a gust front and a downburst by analyzing the characteristics of the pressure field recorded by the POTEKA network. Temporal changes in surface pressure observed in the damaged area were characterized by two episodes of marked increases in atmospheric pressure. The propagation of two marked increases in pressure (0.5 to 1.5 hPa for peak 1, 1.5 to 2.5 hPa for peak 2) was revealed and a cold-air depth of 200 to 600 m and an average propagation speed of 10 ms-1. Based on the time of damage and meteorological observations, the second increase in

pressure coincided with the occurrence of both damage and a gust. On the basis of the POTEKA data, the cause of the damage was attributed to the downburst.

Key words

: downburst, gust front, surface observation network, pressure jump 1. Introduction

The observation of meso- or micro-scale weather phenomena, such as tornadoes, downbursts and gust fronts, is extremely difficult at the surface. In Japan, the Automated Meteorological Data Acquisition System (AMeDAS) operated by the Japan Meteorological Agency (JMA) covers the entire country at a spatial resolution of 17 km. However, an observation network with a resolution of at least 1 km is required in order to observe and analyze the structure of micro-scale disturbances. The structures of gust fronts have been discussed from field observations, laboratory experiments and numerical simulations. Fujita (1986) investigated downbursts / microbursts and explained ground-damage patterns. Wakimoto (1982), Wilson et al. (1984) and Hjelmfelt (1988) presented the life cycle of a thunderstorm outflow and a microburst based on Doppler radar observations. Schematic cross section through a gust front was showed by Charba (1974). Detailed structures of a gust front and airflow of the gust front head using Doppler radar, Doppler sodar and wind-profiler have been reported (e.g., Klingle et al. 1987, Ralph et al. 1993,

(2)

Martner 1997, May 1999). While, Simpson (1994) demonstrated laboratory experiments of a gravity current and showed the similarity between a flow pattern of laboratory gravity current and observational gust fronts / sea breeze fronts. Droegemeier and Wilhelmson (1987) showed numerical-model simulation of a thunderstorm outflow structure. They discussed fine structures of airflow pattern, pressure and temperature perturbation associated with a thunderstorm outflow and gust front. However, the detailed structures of gust fronts are not well known, such as gusts neat the surface, the vertical structure of the head rotation and the structure of the arc cloud, because of few observational data.

Especially, gust wind which is accompanied by gust front generates near the developing cumulonimbus and often causes damage. Downburst/microburst damage have been reported in Japan (e.g., Kobayashi and Kikuchi 1989, Ohno et al. 1994) and some damage to temporary structures caused by a gust front occurred. Despite the large number of such events, however, the frequency, morphology, structure, and gust winds of a gust front remain largely unknown. So, no paper reported regarding as the propagation of gust fronts concerned with a damage area using surface observation network data. To resolve this issue, a dense network of surface observation stations called the POTEKA (POint TEnki KAnsoku) network was established in Gunma Prefecture in August 2013 (Maeda et al. 2014). The network consists of a total of 55 compact weather stations which have been set up at intervals of 1.5 to 4 km at primary schools and convenience stores in 2013.

Fig. 1. The geographical locations of the observation sites. Dotted circles and closed circles (yellow) denote POTEKA stations and AMeDAS sites, respectively. An ellipse and a cross mark indicate the damage area and the most intensive damage point (F1), respectively. Letters A ~ D indicate the POTEKA stations which show weather elements in Figs. 2 and 3.

(3)

In this study, we examined the spatiotemporal changes in pressure and temperature field formation associated with a downburst event that occurred on August 11, 2013. This paper presents temporal and spatial relationship between the propagation of downburst outflow (gust front) and surface damage.

2. The dense network of surface observation stations in Gunma

Figure 1 shows the location of the individual POTEKA observation points in Gunma Prefecture at the end of August 2013. The 55 observation stations that comprise the POTEKA network are positioned to give a spatial resolution of 1.5 to 4 km in an area measuring 20 km (north-south) by 30 km (east-west) in southeastern Gunma Prefecture. The POTEKA network is used to monitor the five weather elements of pressure, air temperature, relative humidity (RH), rain detection, and solar radiation. The data is averaged every 60 s and sent wirelessly to the cloud where it is stored and analyzed. The resolutions of the obtained data are 0.1 hPa (pressure), 0.1°C (temperature), and 0.1% (relative humidity), respectively. The meteorological element data obtained at 48 of the POTEKA stations with the exception of data loss, were combined with JMA radar data and used for this analysis. The spatial distribution of changes in pressure was expressed as isopleth of time, which were detected as the pressure rise over 0.5 hPa during 10 minutes.

Fig. 2. Time series of meteorological elements from 17:30 to 19:00 JST, August 11, 2013 recorded by the POTEKA point A in Fig. 1. Solid red, blue and green lines denote pressure, temperature and mixing ratio, respectively. Shaded blue area indicates the duration of precipitation.

(4)

Strong gusts of wind occurred in Takasaki and Maebashi City in Gunma Prefecture from 18:00 to 18:30 Japan Standard Time (JST) on August 11, 2013. The wind storm associated with the thunderstorms that passed over Maebashi caused widespread damage, injuring 3 people, breaking windows, and lifting the roofs off several houses (Fig. 1). A concrete telephone pole was also blown over. Most of the objects were scattered in a northeasterly direction by the wind. According to a JMA report (Maebashi Local Meteorological Observatory 2013), damage occurred over a length of 23 km and wind speeds on the Fujita scale (e.g., Fujita 1981) were estimated to be F1 (33-49 ms-1). The cause of the damage was concluded to be a “downburst or gust front”.

3. Change in weather elements near the damage area

Figure 2 shows the changes in each weather element over time at the observation point nearest to the site of the damage (Point A in Fig. 1). The figure clearly shows an abrupt increase in atmospheric pressure and an abrupt decrease in temperature, respectively. The first increase of 1.4 hPa was observed from 17:42 to 17:56 JST (pressure jump 1), and the second increase of 2.1 hPa was recorded from 18:02 to 18:07 JST (pressure jump 2). The first increase corresponded closely with a drop in temperature (17:53 and 17:55 JST), the peak of the pressure jump 1 (peak 1 in Fig. 1) occurred 3 minutes after the beginning of the temperature drop (17:53 JST). The temperature decreased for 11 °C between the first (17:56 JST) and second (18:07 JST) jumps in pressure. No pressure nose (e.g., Fujita and Wakimoto 1981) was observed in this case because of the time resolution of the system (60 seconds average). The decrease in humidity (humidity dips; Kobayashi et al. 2007) was observed at the time of the pressure jumps, and the mixing ratio rose approximately 1.5 gkg-1 before and after the pressure jumps in this case. Precipitation began after

the pressure jump 1 (18:00 JST) at the point A.

The same trend in atmospheric pressure conditions was observed at other stations to the eastward of point A. Kobayashi et al. (2007) reported that such abrupt changes in pressure are characteristic of a gust front head, with a decrease in pressure (pressure dip) observed before the peak of the gust and then an increase in pressure 3 minutes after the temperature drop (pressure jump). Our findings correspond with theirs, with the first increase in pressure (Peak 1) being relatively smaller than the second, more conspicuous, increase (Peak 2). In addition, three dips in pressure, indicative of meso-lows, were observed before and after each jump in pressure.

Figure 3 shows the changes in each weather element over time at four observation points on the path of the gust damage (points A, B, C, D in Fig. 1). The distance from point A to point B, B to C and C to D is 4.5, 2.6 and 11.2 km, respectively. The changes in pressure at points B, C and D were similar to that at A, with two pressure jumps observed. At points B, C and D, the first pressure jump was associated with changes of 0.9, 1.4 and 1.7 hPa, and the secondary pressure jump was associated with changes of 2.5, 1.3 and 1.4 hPa, respectively.

The increases in pressure associated with Peak 1 were larger than 0.5 hPa (0.5 to 1.5 hPa), and were observed from the west edge of the observational field (point A) to 20 km east of the point A. On the other hand, the abrupt increases in pressure associated with Peak 2 (1.0 to 2.5 hPa) were observed at wider stations. As for the other weather elements, a temperature drop of 10 to 12.5°C was recorded from 17:53 JST to 18:08 JST (Points A, B, C), successively. Precipitation fell after the first increase in pressure (Peak 1), no precipitation lasting for several minutes after the decrease in temperature. The humidity dips were observed at points B and C at the time of the second increase in pressure (Peak 2), but no marked change was observed in the mixing ratio

(5)

before and after the increase in pressure at the other points. These findings indicated that, in this case, the absolute humidity of the colder air behind the gust front and the warmer air ahead of the gust front was similar.

(6)

Fig. 4. Time sequence of pressure fields from 18:00 to 18:40 JST. Bold color lines P1, P2, TD denote isopleths of Peak 1, Peak 2 and temperature drop, respectively. Solid lines denote isopleths of radar echo intensity 20, 50 and 80 mm/h, respectively. An ellipse and a cross mark indicate the damage area and the most intensive damage point (F1), respectively (Fig. 1).

(7)

Figure 4 shows time sequence of pressure fields from 18:00 to 18:40 JST. The isopleths of TD (temperature drop) and P1 (Peak 1) were well coincided with the positions of the leading edge of high pressure region (18:10 JST) and the edge of radar echo (20 mm/h), respectively. The isopleth P2 (Peak 2) moved eastward and coincided with the center of the high pressure area. In the same manner, those of TD and P1 well coincided with the leading edge of the lower temperature region (28 ~ 30 °C, Fig. 5). The isopleths P2 existed in the lower temperature region below 28 °C.

Fig. 6. Time change of the isopleths associated with both peaks (Peak 1 and Peak 2) in pressure jumps. Bold red and black lines denote isopleths of Peak 1 and Peak 2 at 18:20 JST, and dotted lines denote isopleths of Peak 2 at 18:30 and 18:40 JST, respectively. Color image indicates radar echo pattern (JMA) at 18:20 JST. An ellipse and a cross mark indicate the damage area (JMA) and the most intensive damage point (F1), respectively (Fig. 1). A circle indicates the location of Maebashi Local Meteorological Observatory.

4. Propagation of gust front

Figure 6 shows the time change of isopleths associated with both increases in pressure (Peak 1 and Peak 2) superimposed on the JMA radar echo pattern at 18:20 JST; the isopleths were interpolated at 10-min intervals and are shown for the period from 18:20 to 18:40 JST. In this study, pressure jumps were defined as increases of more than 0.5 hPa compared to the previous pressure dip. The position of Peak 1 (18:20 JST) was associated with an increase in pressure of 0.5 to 1.5 hPa and a drop in temperature that was limited to the main echo area which was located along the forward edge of the developing echo cell which moved toward the northeast.

Time sequence of the isopleths associated with Peak 2 in pressure jump and radar echo patterns are shown in Fig. 7. The increase in pressure associated with Peak 2 was 1.5 to 2.5 hPa and was due to the intensive radar echo passing. The isopleths of Peak 2 had a horizontal scale of 20 km and corresponded with the echo core region. The propagation speed of Peak 1 was 7.0 ms-1

from 18:00 to 18:10 JST and 12.5 ms-1 from 18:10 to 18:20 JST. The average speed of the echo cell

was approximately 10 ms-1 (from 18:10 to 18:40 JST), which corresponded with the average for

(8)

decreases due to friction with the ground over time; however, in this case, it increased with time. The depth of cold air near the gust front, calculated based on the pressure, and propagation speed for Peak 1 - assuming no hydrostatic contribution due to the penetration of the colder air at the face of the gust front - increased from 200 to 600 m.

The isopleths of Peak 2 moved northeasterly at an average speed of 12.3 ms-1 (from 18:10 to

18:40 JST) , with the propagation speed measuring 11.2 ms-1 from 18:10 to 18:20 JST, 9.8 ms-1

from 18:20 to 18:30 JST, and 16.0 ms-1 from 18:30 to 18:40 JST. Additionally, the position of Peak

2 at 18:20 JST corresponded to a site where damage occurred; there were numerous eyewitness accounts of the gust and the damage in this region. A strong gusty wind with a maximum instantaneous wind velocity of 23.4 ms-1 was recorded at the Maebashi Local Meteorological

Observatory at the same time (18:20 JST). Peak 2 could thus be regarded as the second gust front formed by the downburst immediately above the echo core, and the damage in Maebashi City was considered to be caused by the downburst.

In summary, a schematic of outflow development in the thunderstorm is shown in Fig. 8. The propagation of two marked increases in pressure (Peak 1 and Peak 2) was observed in this study. For Peak 1, a propagation speed increased from 7 to 12.5 ms-1. Peak 2 was associated with an

increase in pressure of 1.5 to 2.5 hPa and its propagation speed increased from 10 ms-1 to 16 ms-1

over a 30-min period. A cold-air depth of outflow increased from approximately 200 m (18:00 JST) to 600 m (18:20 JST). The damaging wind (downburst) occurred just after the second pressure jump was observed at 18:20 JST.

5. Conclusions

On August 11, 2013, very strong winds were experienced in Takasaki and Maebashi City in Gunma Prefecture, Japan. Based on data collected by the POTEKA network in Gunma, pressure field characteristics were occurred in conjunction with two marked increases in pressure detected before the passage of the developing cumulonimbus echo core. The propagation of two marked increases in pressure (Peak 1 and Peak 2) were also captured, both of which exceeded 0.5 hPa. Specifically, the increases in pressure were 0.5 to 1.5 hPa for Peak 1, which had a cold-air depth of 200 to 600 m and an average propagation speed of 10 ms-1. Peak 2 was associated with an

increase in pressure of 1.5 to 2.5 hPa and its propagation speed increased from 10 ms-1 to 16 ms-1

over a 30-min period. Peak 1 and Peak 2 were located ahead of, and directly under, the core of the developing radar echo cell, respectively. Thus, the first increase in pressure was associated with the gust front preceding the parent cloud echo core, and the second increase in pressure was caused by the downburst. Based on reports of damage occurrence and records of the gust by the meteorological observatory, the second pressure increase was considered to coincide with the position of F1 damage and the time of the gust observed at the Maebashi Local Meteorological observatory. On the basis of the surface pressure data, the damage occurred at the time of the second pressure jump, not the gust front (first pressure jump). The results imply that it is possible to discriminate the cause of damage associated with thunderstorms at the time of judgments of the wind disaster identify using the surface dense weather observation network data.

It is not well known whether the thunderstorm development on the gust front is related the way of the gust front propagation. Further studies are needed to accumulate the detailed data on the gust front structure, surface gust winds and each meteorological element at the passage of downbursts and gust fronts.

(9)

Fig. 7. Time sequence of radar echo patterns and the isopleths associated with Peak 2 (red bold) in pressure jump from 18:10 to 18:40 JST.

(10)

Fig. 8. Schematic of outflow development in the thunderstorm.

Acknowledgments

The authors would like to thank the Japan Meteorological Agency (JMA) for providing the meteorological data. This study was partly supported by a Grant-in-Aid for Scientific Research (C) 23510232.

References

Charba, J., 1974: Application of gravity current model to analysis of squall-line gust front, Mon. Wea. Rev., 102, 140-156.

Droegemeier, K. K., and R. B. Wilhelmson, 1987: Numerical simulation of thunderstorm outflow dynamics. PartⅠ: Outflow sensitivity experiments and turbulence dynamics, J. Atmos. Sci., 44, 1180-1210.

Fujita, T. T., 1981: Tornadoes and downbursts in the context of generalized planetary scales. J. Atmos. Sci., 38, 1511-1534.

Fujita, T. T., 1986: DFW microburst, University of Chicago, 155pp.Goff, R. C., 1976: Vertical structure of thunderstorm outflows, Mon. Wea. Rev., 104, 1429-1440.

Fujita, T. T. and R. M. Wakimoto 1981: Five scale of airflow associated with a series of downbursts on 16 July 1980. Mon. Wea. Rev., 109, 1438-1456.

(11)

Hjelmfelt, M. R., 1988: Structure and life cycle of microburst outflows observed in Colorado, J. Appl. Meteor., 27, 900-927.

Klingle, D. L., D. R. Smith, and M. M. Wolfson, 1987: Gust front characteristics as detected by Doppler radar, Mon. Wea. Rev., 115, 905-918.

Kobayashi, F., and K. Kikuchi, 1989: A microburst phenomenon in Kita Village, Hokkaido on September 23, 1986, J. Meteor. Soc. Japan, 67, 925-936.

Kobayashi, F., K. Suzuki, H. Sugawara, N. Maeda and S. Nakato, 2007: Strong wind structure of gust front. J. Wind Engineering, 32, 21–28 (in Japanese with English abstract).

Maebashi Local Meteorological Observatory, 2013: On the gust damage in Takasaki and Maebashi City, Gunmaᴾ Prefecture on August 11, 2013 (in Japanese). Quick Report of Damage Investigation, Japan Meteorological Agency, (online), available from (http://www.data.jma.go.jp/obd/stats/data/bosai/tornado/new/2013081102/20130811_gumma.p df), (Last accessed 2014-11-18) .

Maeda, R., M. Suzuki, H. Kure, T. Morita and H. Iwasaki, 2014: Outline and results of the surface dense observation stations “POTEKA” (Point Tenki Kansoku) in Gunma Prefecture. Society of Atmospheric Electricity of Japan, 8, 44 (in Japanese).

Martner, B. E., 1997: Vertical velocities in a thunderstorm gust front and outflow, J. Appl. Meteor., 36, 615-622.

May, P. T. 1999: Thermodynamic and Vertical Velocity Structure of Two Gust Fronts Observed with a Wind Profiler/RASS during MCTEX, Mon. Wea. Rev., 127, 1976-1807.

Ohno, H. et al. Ohno, H., O. Suzuki, H. Nirasawa, M. Yoshizaki, N. Hasegawa, Y. Tanaka, Y. Muramatsu, and Y. Ogura, 1994: Okayama downburst on 27 June 1991:Downburst identification and environmental conditions, J. Meteor. Soc. Japan, 72, 197-222.

Ralph, F. M., C. Mazaudier, M. Crochet, and S. V. Venkateswaran, 1993: Doppler sodar and radar wind-profiler observations of gravity-wave activity associated with a gravity current, Mon. Wea. Rev., 121, 444-463.

Simpson, J. E., 1994: Sea breeze and local wind, Cambridge University Press, 234pp.

Wakimoto, R. M., 1982: The life cycle of thunderstorm gust fronts as viewed with Doppler radar and rawinsonde data, Mon. Wea. Rev., 110, 1060-1082.

Wilson, J. W., R. D. Roberts, C. Kessinger, and J. McCarthy, 1984: Microburst wind structure and evaluation of Doppler radar for airport wind shear detection, J. Clim. Appl. Meteor., 23, 898-915.

Corresponding author: Fumiaki Kobayashi

Department of Geoscience, National Defense Academy, Yokosuka 239-8686, Japan. E-mail: kobayasi@nda.ac.jp

(Received August 14, 2015; revised January 13, 2016;

accepted January 13, 2016)

Fig. 1. The geographical locations of the observation sites. Dotted circles and closed circles (yellow)  denote POTEKA stations and AMeDAS sites, respectively
Figure 1 shows the location of the individual POTEKA observation points in Gunma Prefecture  at the end of August 2013
Fig. 3. Same as Fig. 2 except for points A to D in Fig. 1.
Fig. 4. Time sequence of pressure fields from 18:00 to 18:40 JST. Bold color lines P1, P2, TD denote  isopleths of Peak 1, Peak 2 and temperature drop, respectively
+4

参照

関連したドキュメント

We present sufficient conditions for the existence of solutions to Neu- mann and periodic boundary-value problems for some class of quasilinear ordinary differential equations.. We

In Section 13, we discuss flagged Schur polynomials, vexillary and dominant permutations, and give a simple formula for the polynomials D w , for 312-avoiding permutations.. In

Analogs of this theorem were proved by Roitberg for nonregular elliptic boundary- value problems and for general elliptic systems of differential equations, the mod- ified scale of

The commutative case is treated in chapter I, where we recall the notions of a privileged exponent of a polynomial or a power series with respect to a convenient ordering,

Then it follows immediately from a suitable version of “Hensel’s Lemma” [cf., e.g., the argument of [4], Lemma 2.1] that S may be obtained, as the notation suggests, as the m A

Correspondingly, the limiting sequence of metric spaces has a surpris- ingly simple description as a collection of random real trees (given below) in which certain pairs of

Q discrep : Predefined empirical constant corresponding to the minimum value of the module of total discrepancy between estimated gas supply volumes, which is of practical

Based on sequential numerical results [28], Klawonn and Pavarino showed that the number of GMRES [39] iterations for the two-level additive Schwarz methods for symmetric