DISTRIBUTION OF SUMMERTIME THE SURFACE ROUGHNESS IN

Hideo Takahashi*, Yasuk*Tokyo Metropolitan University, Tokyo, Ja

***East Japan Railwa

AbstractAbstract

The localized high-frequency areas of summertime inleeward of high-rise buildings areas (e.g. Shinjuku,based on the calculated aerodynamic parameters (suggests that large surface roughness created byfrequency areas of intense rainfall in urban areas.

Key words: short-duration intense rainfall, surface rou

1. INTRODUCTION

Recently, the frequent occurrence of ’urban floods’ dreported in the Tokyo Metropolitan Area. It was pointeof urban area by using observed rainfall data during MTRMM data in recent years (Shepherd and Negri, 20of intense rainfall within an urban area, such as the Tof intense rainfall within an urban area, such as the Tusing high-density rainfall data. Here we present thrainfall in the Tokyo Metropolitan Area based on detaifor the localized high-frequency area of intense raincalculated using surface elevation data, which can rewind velocity.

2. DATA AND PROCEDURE

2 1 Rainfall data2.1. Rainfall data

This study utilized hourly rainfall data of 90 stationsJapan Meteorological Agency (●), East Japan RailwTokyo Metropolitan Government (×). The target pe1991 to 2002. In total, 226 cases were extracted wheor more in the central part of the Tokyo Metropolitandisturbances.

2.2. Calculation of surface aerodynamic parameter

The heights of artificial constructions were determelevation of a digital surface model (DSM) providedelevation interpolated from the 5 m interval DEMparameters (roughness length, z0, and zero-plain dispthe Tokyo Wards Area by method of Raupach (1evaluation by Grimmond and Oke (1999). One of theestimated by λF = (ΣAFi ) / AT where ΣAFi is the totaestimated by λF (ΣAFi ) / AT where ΣAFi is the totaseen by oncoming wind within plan area of AT = 1 kmzH, was calculated by omitting the roughness elemenaspect ratio λP, though it did not use in calculation, waz0 and zd representing 1 km2 areas were calculatedstructure.

* Corresponding author address: Hideo Takahashi D* Corresponding author address: Hideo Takahashi, DMinamiosawa Hachioji Tokyo, 192-0397 Japan; e-mai

INTENSE RAINFALL FREQUENCY AND THE TOKYO METROPOLITAN AREAko Nakamura**, Hiroto Suzuki***apan; **Tokyo Gakugei University, Tokyo, Japan,ay Company, Chiba, Japan

ntense rainfall in the Tokyo Metropolitan Area appear to theIkebukuro). A simple estimation of ascending wind velocity(z0 and zd) representing 1 km2 areas by using DSM datahigh-rise buildings is a possible factor for localized high-

ughness, high-rise buildings

due to summertime, short-duration intense rainfall has beened out that the intense rainfall is likely to occur to the leewardMETROMEX conducted in the 1970s (Changnon, 1978) and002). However, more detailed climatological spatial structureTokyo Metropolitan Area has not been analyzed completelyTokyo Metropolitan Area has not been analyzed completely

he spatial structure of the occurrence frequency of intenseiled hourly rainfall data. In order to discuss the causal factorsnfall, the distribution of surface aerodynamic parameters isesolve individual buildings for simple evaluation of ascending

s in the Tokyo Metropolitan Area (Fig. 1a) recorded by theway Company (△), and the Bureau of Construction of the

eriod covers the months of June to September in the yearsere at least one station recorded an hourly rainfall of 20 mmn Area without being influenced by synoptic- or meso-scale

rs

mined as elevation difference between the 2.5 m intervald by the Pasco Corporation and the 2.5 m interval ground

of the Geographical Survey Institute. The aerodynamicplacement, zd), representing 1 km2 areas were calculated in

1992, 1994, and 1995) that obtained comparatively goode required parameters, the frontal area aspect ratio λF, was

al area of roughness elements (buildings, plants and so on)al area of roughness elements (buildings, plants and so on)m x 1 km. In addition, average height of roughness elements,nts lower than one-story houses, 3.5 m, since the plan areaas consistent with the analysis of independent GIS data. Thed at the interval of 200 m to represent complicated urban

Department of Geography Tokyo Metropolitan University 1 1Department of Geography, Tokyo Metropolitan University, 1-1 il: thideo@tmu.ac.jp

The seventh International Conference on Urban Climate, 29 June - 3 July 2009, Yokohama, Japan

3. DISCUSSION

3.1. Localized high-frequency areas of intense rain

Figure 1(a) indicates the relative frequency distributiozones of intense rainfall are located along the boundaregion of the Tokyo Wards Area. Although the wind dsouth-southwest during periods of intense rainfall, ithours before the occurrence of intense rainfall. In thetoward the west from the center of Tokyo or ShinjukShinjuku). For southerly winds (Fig. 1c), a relativelyTokyo and Saitama Prefecture; the maximum frequenIn addition, a high-frequency zone of intense rainfall acase of northerly winds (Fig. 1d). It is noted that thethe leeward of high-rise buildings such as Sinjuku and

(a)( )

(c)

Figure 1. Relative frequency distribution of intense rainfthe inner frame) for all cases (a: 226 cases), easterly wand northerly wind cases (d: 29 cases)

3.2. Preliminary estimation of ascending wind velo

y ( )Relative frequency means the percentage ratio of numbrainfall cases occurred within the inner frame. Abbrevianames and location of Shinjuku, Ikebukuro, Shibuya, anand red arrows in (b)-(d) show the largest number of bAMeDAS (Automated Meteorological Data Acquisition S

Figure 2 shows the distribution of calculated z0 (a) an

nfall

on of intense rainfall among the 226 cases. High-frequencyary of the Tokyo and Saitama Prefecture, and the northwestdirection at Tokyo (Otemachi) varies between northeast andis concentrated in the easterly and southerly directions 2-3

e case of easterly winds (Fig. 1b), the occurrence increasesku; the highest frequency occurs in Nakano Ward (west ofy high-frequency zone appears along the boundary of thency occurs in the Itabashi Ward, located north of Ikebukuro.also appears south of the Shibuya-Kasumigaseki area in thelocalized high-frequency areas of intense rainfall appear to

d Ikebukuro.

(b)( )

(d)

fall (≧20 mm/hour) occurred in urban area of Tokyo (inside ofwind cases (b: 105 cases), southerly wind cases (c: 77 cases),

ocity due to surface roughness

ber of intense rainfall cases at each station to that of all intenseations “SNJ”, “IKB”, SBY”, and “TKY” in (a) indicate the placend Tokyo Railway Station, respectively. The background figureuilding floors in every 50 m x 50 m areas, and wind vectors atystem) stations averaged for respective cases.

nd zd (b) in the Tokyo Wards area. Areas with large z0 and zd

The seventh International Conference on Urban Climate, 29 June - 3 July 2009, Yokohama, Japan

(a)

z0

Figure 2. Distribution of calculated roughness lengthmethod

250(b)

250(a)

100

150

200

Hei

ght (

m)

(b)

100

150

200

Hei

ght (

m)

(a)

0 2 4 6 8 10

50

Wind Speed (m/s)0 2 4 6 8 10

0

50

Wind Speed (m/s)

Figure 3. Example of estim(a) and (b) indicate the estimated vertical profiles of horiz

are observed around Sinjuku, Ikebukuro, and the cgovernment offices are located. To evaluate the effecthe distribution of ascending wind velocity in the lowe

using calculated z0 and zd values assuming a wind vWest-Sinjuku (b), and West-Sinjuku (a) where both ofdifference in horizontal wind velocity between (a) and (bfrom convergence of horizontal wind between (a) and (b)

the distribution of ascending wind velocity in the loweThe distribution of horizontal wind velocity at an arbFig. 1(b)-(d) is evaluated by applying the logarithmivelocity of 10 m/s at 250 m (Fig.3). In the case of the>0.2 m/s appears near both Shinjuku and Ikebukuroappears near Shinjuku, although that around IkebukMoreover, the ascending wind velocity near Kasumiganortherly winds (Fig. 4c) as compared with southerly wThe distribution of estimated ascending wind velocity vthat of intense rainfall frequency for the respectivehigh-rise buildings is a possible factor for localized hig

(b)

zd

h z0 (a), and zero-plain displacement zd (b) by Raupach’s

250(c)

250(d)

100

150

200

Hei

ght (

m)

(c)

100

150

200

Hei

ght (

m)

(d)

0 0 2 4 6 8 100

50

Wind Speed Difference (m/s)0 0.1 0.2 0.3

0

50

Vertical Wind Speed (m/s)

mation of ascending wind velocityzontal (southerly) wind velocity by applying the logarithmic rule

entral Tokyo (vicinity of Tokyo Railway Station) where thect of surface roughness on the occurrence of intense rainfall,r layer is estimated from the convergence of horizontal wind

velocity of 10 m/s at 250 m for 1 km windward (southward) ofz0 and zd indicate the largest, respectively. (c) indicates the

b). (d) shows the estimated ascending wind velocity calculated).

r layer is estimated from the convergence of horizontal wind.bitrary height for the three wind directions corresponding toc rule using calculated z0 and zd values assuming a wind

e southerly wind direction, a large ascending wind velocity of(Fig. 4a). A large ascending wind velocity of about 0.2 m/s

kuro is smaller in the case of the easterly winds (Fig. 4b).aseki (near the Tokyo Railway Station) is larger in the case ofwinds (Fig. 4a).varies according to the wind direction, which corresponds towind directions. Thus, large surface roughness created by

gh-frequency areas of intense rainfall in urban areas.

The seventh International Conference on Urban Climate, 29 June - 3 July 2009, Yokohama, Japan

(a)

(c)

Figure 4. Distribution of estimated ascending wind velocnortherly wind (c)In Fig. 4, areas with descending wind are suppressed.

4. CONCLUDING REMARKS

The localized high-frequency areas of summertime inleeward of high-rise buildings areas (e.g. Shinjuku,based on the calculated aerodynamic parameters (suggests that large surface roughness created byfrequency areas of intense rainfall in urban areas.

ReferencesReferencesChangnon, S.A., 1978. Urban effects on severe local

578-586.Grimmond, C.S.B and Oke, T.R., 1999. Aerodynamic

form. Journal of the Applied Meteorology, 38, 1262-Raupach, M.R., 1992. Drag and drag partition on rougRaupach, M.R., 1994. Simplified expressions for ve

function of canopy height and area index. BoundaryRaupach, M.R., 1995. Corrigenda. Boundary-layer MeShepherd, J.M., H. Pierce and A.J. Negri, A.J., 2002

from spaceborne rain radar on the TRMM satellite.

(b)

city at 200 m height for southerly wind (a), easterly wind (b), and

ntense rainfall in the Tokyo Metropolitan Area appear to theIkebukuro). A simple estimation of ascending wind velocity(z0 and zd) representing 1 km2 areas by using DSM datahigh-rise buildings is a possible factor for localized high-

storms at St. Louis. Journal of the Applied Meteorology. 17.

c properties of urban areas derived from analysis of surface-1289.gh surfaces. Boundary-layer Meteorology, 60, 375-395.egetation roughness length and zero-plan displacement asy-layer Meteorology, 71, 211-216.eteorology, 76, 303-304.2. Rainfall modification by major urban areas： ObservationsJournal of Applied Meteorology. 41, 689-701.

The seventh International Conference on Urban Climate, 29 June - 3 July 2009, Yokohama, Japan