Observations of the planetary boundary layer over equatorial Indonesia with an L band clear-air Doppler radar:

Transcription

1 Radio Science, Volume 30, Number 4, Pages , July-August 1995 Observations of the planetary boundary layer over equatorial Indonesia with an L band clear-air Doppler radar: Initial results Hiroyuki Hashiguchi, Shoichiro Fukao, Toshitaka Tsuda, and Manabu D. Yamanaka Radio Atmospheric Science Center, Kyoto University, Kyoto, Japan Daniel L. Tobing, Tien Sribimawati, and Sri Woro B. Harijono Agency for the Assessment and Application of Technology, Jakarta Pusat, Indonesia Harsono Wiryosumarto National Institute of Aeronautics and Space, Jakarta Timur, Indonesia Abstract. An L band ( MHz) boundary layer radar (BLR) has been in continuous successful operation in Serpong, Indonesia (6.4øS, 106.7øE), since November The performance of the BLR with respect to the observation height range and the wind measurement reliability has been examined on the basis of simultaneous meteorological observations. We have found that there are two types of strong echo structures appearing systematically in the equatorial planetary boundary layer with diurnal variations on clear days. The first type is the striking appearance of a strong echo layer ascending from below 300 m (in the morning) to above 3-5 km (in the afternoon), which is identified with a diurnal variation of the top of the mixing planetary boundary layer. As expected, it is higher in the Indonesian equatorial region than in midlatitudes. Another type is layered echoes appearing at 2-3 km heights from nighttime to morning, which seem to be coincident with humidity gaps. 1. Introduction Recent progress in radar profiling techniques has been extended to the planetary boundary layer (PBL) [Ecklund et al., 1988, 1990; May and Wilczak, 1993]. We have developed an L band transportable system called the Kyoto University boundary layer radar (BLR) [Hashiguchi et al., 1995; S. Fukao et al., An L-band clear-air Doppler radar for the atmospheric boundary layer, submitted to Journal of the Meteorological Society of Japan, 1995, hereinafter referred to as submission, 1995]. The high-resolution (both in time and in height) and the reliability of the three-dimensional wind velocity vector observed with the BLR are expected to improve our knowledge and interpretation of the PBL [Kropfii, 1990], since classical techniques such as balloons, aircraft, and anemometric towers could Copyright 1995 by the American Geophysical Union. Paper number 95RS /95/95R S $08.00 not observe the PBL in detail [e.g., Stull, 1988; Garratt, 1992]. It is now widely recognized that the dynamics of the PBL in the Indonesian equatorial region (the maritime continent) are very important to our understanding of global climate control mechanisms associated with the maritime continent surrounded by high surface temperature ocean. For example, PBL dynamics govern the generation of convection, and the large year-to-year variations of the cloud convection over the maritime continent are closely related to the behavior of the E1 Nifiosouthern oscillations (ENSO) [e.g., Nitta et al., 1992]. Tall cumulonimbus clouds often penetrate the tropopause and transport various minor constituents in the troposphere deeply into the stratosphere [Holton, 1984]. They also excite various atmospheric waves, such as gravity waves, Kelvin waves, and other long-period oscillations [Tsuda et al., 1992, 1994a, b]. However, these studies are based on too coarse (or limited numbers of) observations with satellites or radiosondes to clarify 1043

2 1044 HASHIGUCHI ET AL.: PLANETARY BOUNDARY LAYER OBSERVATIONS.....: o ' I years (as of October 1994) and are planni to extend them at least until the end of In the following sections we first give a brief description of the BLR system. A description of the performance of the BLR observations in Serpong is ' given in 3. Observation evidence conce ing the vertical extent and diurnal variations of the PBL is described in section 4. In section 5 we describe conclusions at the present stage, including future targets. 100 in N Serpøugo I kl PUSPIPTEIC Bo oro Figure 1. Map of the observation site of the boundary layer radar (BLR) in Indonesia. Regions higher than approximately 450 m above the mean sea level are shaded. essential features of PBL dynamics over the maritime continent around Indonesia. We started international cooperative observations of the PBL over Indonesia using the BLR in November 1992 at PUSPIPTEK (National Center for Research, Science and Technology) (6.4øS, 106.7øE; 50 m above sea level) in Serpong, West Java, which is located in the southwest suburbs of Jakarta (see Figure 1) [Tsuda et al., 1995]. The observations are conducted by BPPT (Agency for the Assessment and Application of Technology) and LAPAN (National Institute of Aeronautics and Space) on the Indonesian side and by RASC (Radio Atmospheric Science Center) of Kyoto University on the Japanese side. We have been continuing radar observations without serious problems for 2 2. Table The 1, Kyoto and details University have been Boundary desc Layer by S. Radar (BLR) The basic parameters of the BLR are summarized Fukao et al. (submission, 1995). The B LR is a small and transportable radar operating at a frequency of MHz (L band) with a peak transmitter power of 1 kw. It is designed to receive backscattered echoes from refractive index fluctuations, which are mainly generated by fluctuations of humidity and atmospheric stability profiles associated with atmospheric turbulence. Figure 2 shows the physical configuration of the B LR in Indonesia. Three parabolic antennas are pointed into the vertical and two oblique directions aligned to the east and north at a zenith angle of 15 ø. The BLR provides vertical profiles of three components of the wind velocity vector, turbulence parameters (on clear days), and raindrop characteristics (on rainy days) in the lower troposphere, including the PBL, with time and height resolutions of about 1 min and 100 m, respectively. In addition, temperature profiles can be obtained by means of the RASS (radio acoustic sounding system) technique (T. Tsuda et al., pre- Table 1. Principal Specifications of the Kyoto University Boundary Layer Radar Parameter Operating frequency Antenna Aperture Beam width Beam directions Transmitter Peak power Average power Bandwidth Pulse length Interpulse period Value MHz (L band) Three parabolic antennas 3.1 m2 (2 m in diameter) 7.6 ø (half power) Fixed into three directions Three solid state amplifiers I kw (maximum) 20 W (duty ratio 2%) (maximum) 4 MHz 0.67, 1.0, 2.0/as (variable) 50, 100, 200/as (variable) Data from S. Fukao et al. (submission, 1995).

3 ... HASHIGUCHI ET AL.: PLANETARY BOUNDARY LAYER OBSERVATIONS 1045.;:?.... ;:... ;.... '.. :.:. Figure 2. View of the BLR installed at PUSPIPTEK in Serpong near Jakarta, Indonesia. All the radar equipment except for three parabolic antennas is installed in the rear container. print, 1994) although they are not described in this paper. The B LR was first installed at the MU (middle and upper atmosphere) Observatory located among small mountains in Shigaraki, Japan (34.85øN, øE; 385 m above sea level), in December 1991, and continuous observations were conducted during May and August 1992 [Hashiguchi et al., 1995]. After that, it was installed at PUSPIPTEK in Serpong near Jakarta, Indonesia, in October 1992 and has been continuously operating since November 9, With control by an on-line computer, unmanned operation of the BLR is automatic, except for the exchange of data storage tapes every 4 or 5 days. In the present study the echoes from each of 64 observed heights ( km) are integrated coherently for 3.2 ms to form each sampled value, and a 128-point time series of these values is stored. A 128-point (0.4-s average) Doppler spectrum is then calculated for each height by using a fast Fourier transform. Beam direction is instantly changed after taking the Doppler spectrum. Finally 32 Doppler spectra are averaged for each height. We can obtain height profiles br 3 directions in about 50 s, including the data transfer time (---10 s) to the computer. 3. Observations in Serpong, Indonesia Observations with the BLR in Serpong were conducted for about 2 years without any serious problems, except for relatively short stops mainly due to commercial electric power line problems. At the radar site, fundamental meteorological parameters such as surface winds, temperature, humidity, precipitation, and solar and net infrared radiations were also monitored with standard instruments. In addition to routine observations, we carried out two intensive campaigns in the dry (October 8-15, 1993) and rainy (February 15-22, 1994) seasons. In each campaign approximately 50 rawinsondes were launched at 3-hour intervals at the radar site [Tsuda et al., 1995]. In this section we describe the performance of the B LR in Serpong, mainly concerning the observation height range and the wind measurement reliability Observation Height Range of the BLR The lowest height of BLR observations (throughout this paper the height from the ground is used) is determined to be m from the delay between signal transmission and reception. However it is significantly affected by both ground clutter and atmospheric echoing properties [ Hashiguchi et al.,

4 1046 HASHIGUCHI ET AL.: PLANETARY BOUNDARY LAYER OBSERVATIONS precipitation particles (raindrops or drenched ice crystals), so that the height coverage is greatly increased. Precipitation echo is extended occasionally up to near the tropopause height under conditions of deep convection (C. R. Williams et al., preprint, 1994), although the data are sampled up to only 6.4 km height in routine observations. It is noted, strictly speaking, that since reflectivityweighted fall speed is measured in precipitation, it needs to be subtracted in estimations of horizontal winds Ratio of Effective Samples (%) Figure 3. Height coverage of effective samples of BLR echoes obtained during 1 hour, calculated for November 1992 through February 1994 in Serpong, Indonesia (solid curve), and April through August 1992 in Shigaraki, Japan (dashed curve). 1995; S. Fukao et al., submission, 1995]. Figure 3 shows the ratio of effective samples to all samples over 1 hour with a height resolution of 100 m. Wind velocities below 600 m were difficult to obtain (probability <80%) in Shigaraki, whereas in Serpong wind velocities can be obtained from 300 m, which is the system limit of the BLR. This is because the ground clutter level is reduced in Serpong, since we selected the radar site without any surrounding mountains or tall buildings, lowered the antenna height, and constructed a clutter prevention fence around the antennas. The maximum height of BLR observations depends on echo intensities from the atmosphere. In the clear atmosphere without precipitation, the echo intensity increases with humidity, temperature, and their vertical gradients. This is why the height coverage was normally limited to below 2.2 km (with probability ->80%) in Shigaraki, while it increases to 3.2 km in Serpong, owing to higher humidity and temperature. It is noted that the highest height in Shigaraki was a little lower than 3 km even under conditions with the highest humidity and temperature in August. For the same reason, the height coverage shows large variations depending on local time. On clear days the height coverage is highest in midafternoon and lowest during the nighttime. The L band radar is very sensitive to 3.2. Wind Measurement Reliability The BLR detects the mean radial Doppler velocity inside a sampling volume given by the beam width of 7.6 ø for one beam and the vertical resolution of 100 m. The vertical velocity determined by using one vertical beam corresponds to the mean value for a volume of roughly 130 m wide and 100 m thickness at 1 km height. Since the zonal and meridional velocities are determined by using vertical and 15ø-oblique beams, these velocity components are the mean values for a volume of about 400 m width and 100 m thickness at 1 km height under an assumption that the wind fields are homogeneous in this volume. The estimated velocities can be regarded as instantaneous values in time taken at 1 o 15 LT on 10 Ock lo Velocity (ms-1) Figure 4. Profiles of zonal (solid curves) and meridional (dashed curves) wind velocities, which were obtained simultaneously with the BLR (thick curves; averaged over 15 min) and rawinsondes (thin curves; VAISALA, RS80-15N) at about 1500 LT on October 10, 1993.

5 HASHIGUCHI ET AL.: PLANETARY BOUNDARY LAYER OBSERVATIONS Oct i, i, i, i. i, i, i, i, i, i...!iiiiili!11iii -, i I Hour Local Time Figure 5. Diurnal variation of root-mean-square differences between the BLR and the rawinsondes for (a) zonal and (b) meridional winds obtained for 51 simultaneous observations during the period October 8-15, the sampling interval ( 1 min), although some averaging is required to obtain reliable values. We have compared wind velocities obtained simultaneously with the B LR and rawinsondes (VAISALA, RS80-15N; launched at the radar site every 3 hours) during the period October 8-15, Considering that a rawinsonde took about 15 min to pass through the observation h6ight range of the BLR, we have averaged the BLR data over 15 min after the launch time of the rawinsondes. As shown in Figure 4, both wind profiles coincided fairly well without any significant biases, and the BLR detected smaller-scale variations compared with those of the rawinsondes. The inherent accu- racy for the BLR wind measurements is better than 1 m s -1 (S. Fukao et al., submission, 1995). Figure 5 shows the diurnal variation of root-mean-square (RMS) differences between the two techniques for zonal (Figure 5a) and meridional (Figure 5b) components. The RMS differences were better than 3 m s -1 almost all of the time, but at 0300 LT large differences, which are possible contaminations due to birds and/or insects, discussed further below, were obtained. It is noted that the B LR can obtain

6 1048 HASHIGUCHI ET AL.: PLANETARY BOUNDARY LAYER OBSERVATIONS Oct 1993 [ I [ I [ I i I i I 'l I i I I I, I i I [ I i 4- ',., 1 A.:'-I ' -[ " '/":,, J' ".... I I '... ":: - :. t'::'"... ' -' '- ' ' " ' - '- '"':: -"... ' - - ' : :- : : : :: - :;0, -. ':,.....?-.:..... " ', - :..:: ;---" " -,::. :' " : Z: :? L :- ' -:-:.' '." " -11. " :: "' '-/ ': '.'.'2¾: ' :':. - : "- '"--., 'X..., ' " " - 4,, _... -go ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' lfi Ho 10 -I : Oct 1993 C T e Figure 6. Time-height cross section of the equivalent radar reflectivity factor observed with the BLR averaged every 30 min for the eastward beam (azimuth = 90 ø, zenith = 15 ø) in Serpong during the period October 10-12, Circles with numbers indicate echo types (see text for details). three-dimensional wind profiles (including vertical velocity only when there is no precipitation) with higher vertical and temporal resolutions than rawinsondes. In addition, we sometimes observed strong echoes below the type 2 echo layer of Figure 6, in particular in the nighttime. We examined every sample of these echoes and found a discontinuity in height and time in the shape of the Doppler frequency spectrum. We consider that such echoes may be contaminations caused by birds, bats, or insects [e.g., Browning and Atlas, 1966; Lhermitte, 1966; R6ttger and Larsen, 1990; J. M. Wilczak et al., preprint, 1994]. In order to remove this contamination we fitted a Gaussian function to the data neglecting any peak that was not continuous in height and time. We will investigate in detail these echoes and appropriate reduction procedures in the future. 4. Observations of the Equatorial Planetary Boundary Layer In this section we demonstrate only the most striking results obtained so far with the BLR in Serpong, Indonesia. These are the observational evidence for the vertical extent and diurnal varia- tions of the equatorial PBL, which are quite different from those observed in Shigaraki, Japan, by Hashiguchi et al. [1995] Diurnal Variations of Equatorial PBL on Clear Days Figure 6 shows the time-height cross section of echo intensity (equivalent radar reflectivity factor) observed with the B LR in Serpong during the period October 10-12, Although smaller scale variations were complicated, two types of strong echo regions were found. First, every morning at about 0800 LT (Indonesian Western Standard Time equals universal time plus 7 hours), a significant

7 HASHIGUCHI ET AL.' PLANETARY BOUNDARY LAYER OBSERVATIONS 1049, Oct 1993 õ [ I, I I I I I I I I I I I I I I I I I./..... :, ".,. ' '--,',, - x,,,,,,,,.,-,.... ' ' I " 'l '-'... ' 'I 0... I '... I '" I... ' I ' I x''' ' I lll ' I... I ' " '1 ' I ' li,ll l... I- 10 I 1 -I- c T e - 10 (ms -t 0 Horn* q Oct 1993 F re 7. T me-he h cross section of he zon -me d on w nds vem ed eve 1 hour observed w h he BL du n he pe od October 1 12, 1. The ows p o ed he bottom re obtained w h standard nemome er (OGASAWA A, WS-A ) bom 10 m bove round he md r s e. The vector dkecfion s upward for no hw rd w nd, h w rd for n e s w rd w nd. echo layer (type 1) appeared at the lowest observation height (300 m) and gradually ascended up to 3-5 km height until about 1600 LT. Such a layer structure is smeared in the evening and is identified with the top of the PBL (or the mixing layer) in the following subsections. Secondly, another type of thin layered echo (type 2; we named this type "necklace echo") was observed at 2-3 km heights during LT, descending about 300 rn at around 0600 LT every morning. These 3 days were typical clear days (without any precipitation at the ground) in the dry season, and observations on other clear days showed similar diurnal variations of echo power intensity. On cloudy days such features were weak or disappeared. On rainy days strong echoes caused by rainfall appeared, but they are entirely different and distinguished from the typical behavior of clear-air echoes mentioned above. Figures 7 and 8 show variations of the three components of wind velocity observed with the BLR. We found that the wind direction and speed changed drastically near the type 1 (ascending) echo layer in the daytime and also near the type 2 necklace echo layer in the nighttime. Below the type 1 echo layer (or inside the mixing layer) the horizontal wind was generally weak and the vertical velocity was strong and highly variable. Below the necklace echo layer, both the horizontal and vertical velocities were variable, although the vertical velocity fluctuations were relatively weaker than those below the type 1 echo layer. Above these two types of layers, an easterly wind was dominant throughout the observational period (at least in this season). The downward vertical velocities were observed below 1 km around 1200 LT throughout the observational period. It is possible that this was very light precipitation which cannot be measured by the rain gauge on the ground. In Figure 7 surface winds (about 10 m above the ground) monitored at the radar site with a standard anemometer (OGASAwARA, WS-A54) are also

8 1050 HASHIGUCHI ET AL.' PLANETARY BOUNDARY LAYER OBSERVATIONS Oct I1.5 (ms Hour I- 10 -I. 11 I 12 -I Oct 1993 Local Time Figure 8. Time-height cross section of the vertical velocity averaged every 10 min observed with the BLR during the period October 10-12, E Oct 1993, I I, I, I, I, I, I, I, I, I, I, 1.O ',. 08 CI06 0 ' "" 00 0 ' v Hour * , 12 t Oct , I, I, I, I, I, I, J, t, I, I, I, /? (b) 8o Hour ' , I Oct 1993 Local Time Figure 9. Time variations of (a) solar radiation (EKO, M$-42) and (b) temperature (VAIsALA, HMP-133Y) avera;ed eve l0 min at 1. m above the ;round of the radar site dufin; the period October l 12, 1993.

9 HASHIGUCHI ET AL.' PLANETARY BOUNDARY LAYER OBSERVATIONS li Virtual Potential Temperature e v (K) Figure 10. Profiles of the virtual potential temperature using rawinsondes launched at the radar site about every 3 hours during the period October 10-12, Each profile is shifted by 1 K/hour according to differences of the launch time from the first profile launched at 0000 LT on October 10. Arrows indicate the top of the strong mixing layer (or daytime PBL, seen as almost constant virtual potential temperature layer). Circles with numbers indicate the heights of strong echoes and their respective types observed with the BLR (see text for details). I 400 plotted. Although variations in the wind were less intense compared with those from the BLR, the wind direction tends to have northerly and southerly directions, in daytime and nighttime, respectively, throughout the observational period. We consider that these diurnal variations of surface winds correspond to the sea-land breeze circulation, since the observatory is located about 40 km south of the Java coastline, and there are no mountains between observatory and coast. The sea breeze clearly extended to about 1 km height and there was the hint of a return flow aloft. However, the land breeze was not clear in the BLR wind velocities and was, as expected, relatively shallow Features of Radiation, Temperature, and Humidity We found from pyranometer (EKO, MS-42) data (Figure 9) that after sunrise at around 0600 LT, solar radiation started to increase rapidly on the ground from 0700 LT until around 1200 LT. Apparently there were not many clouds before noon. In the afternoons of October 11 and 12, clouds appeared frequently during the period LT, getting less frequent after that until sunset at 1800 LT. On October 10, relatively thick clouds covered the site almost continuously during the period LT, although such a difference is not clear in the echo power and wind variations. The surface temperature (VAISALA, HMP-133Y) variations were consistent with those features of radiation (or cloudiness) and showed typical diurnal characteristics on a clear day (maximum after local noon and minimum before local sunrise). Figure 10 shows profiles of the virtual potential temperature observed with rawinsondes launched at the radar site at approximately 3-hour intervals. We found that the type 1 echo layer appeared near the top of the mixing layer (indicated by arrows)

10 HASHIGUCHI ET AL.' PLANETARY BOUNDARY LAYER OBSERVATIONS Oct ::::::::::::::::::::::::::::::::::::: ====================================== ::::::::::::::::::::::::::... ::::::::::::::::::::::::::::::::: ::::::.: ::::::::::::::::. ::::::::::::::::::::::::::::::.::::::::::: -:::::::::::::::: ================================= ::::::::::::::::::::::::::::::::: =============================== ================================= ::::::::::::::::::::::... :-:::::::.- ::::.::: ::::::- ß..... :::o:.,:.:--- ( ),, Hour I 10 I Oct 1993 Local Time Figure 11. Time-height cross section of the relative humidity observed with rawinsondes during the period October 10-12, defined' by an almost constant virtual potential temperature. In considering the echoing mechanisms of the radio waves, we suppose that strong turbulence and vertical velocity fluctuations inside the mixing layer and the striking gap in the vertical distributions of humidity and temperature contribute to the generation of the strong echo layer observed with the BLR [Angevine et al., 1994; Hashiguchi et al., 1995]. Figure 11 shows the time-height cross section of relative humidity observed with rawinsondes. The type 2 (necklace) echo layer at 2-3 km heights corresponded to the sharp decrease of humidity with height in nighttime. Such humidity gaps were also found as weak gaps of virtual potential temperature in Figure 10. This echo layer was similar to that observed with the 915-MHz B LR at the height of the trade wind inversion in Hawaii, United States [Rogers et al., 1993]. The type 2 echo layer associ- ated with the humidity gap was clear in Serpong in nighttime, while the echo layer was observed in Hawaii during the whole day Discussion The observational features of the type 1 echo layer in this study were quite similar to those observed with the B LR and another 915-MHz radar in midlatitudes on clear days [Ecklund et al., 1990; Angevine et al., 1994; Hashiguchi et al., 1995]. The latter features were also consistent to the diurnal variations in a clear daytime PBL simulated by a numerical model [ Yamada and Mellor, 1975]. However, the top height of such a layer in the midlatitudes was limited up to 1-2 km, whereas that observed here in the equatorial region reached 3-5 km. According to Ekman's classical theory, the thickness of the PBL is estimated by (K/f) 1/2, where K is eddy diffusivity and f is the Coriolis parameter. If K is constant, the ratio of the thickness in Serpong (6øS) to that in Shigaraki (35øN) becomes (sin 35ø/sin 6 o) u This prediction seems consistent with our observational results, although the interpretation of the PBL in the equatorial region is not so simple [c.f., Firestone and Albrecht, 1986]. K can

11 HASHIGUCHI ET AL.' PLANETARY BOUNDARY LAYER OBSERVATIONS 1053 be estimated from the echo power (spectral width or intensity) after removing effects of background wind, humidity, and temperature, but it is beyond the scope of this paper. 5. Conclusion We have been successfully conducting observations with the B LR in Serpong in equatorial Indonesia since November The good performance of the BLR has been examined by simultaneous meteorological observations. Even in the initial results presented here, the BLR has proved that there exist obvious diurnal variations of the strongly mixing PBL on clear days in the equatorial ! '1 - _ I_ region, which arc partly lmlla[' to, OUt nluc[1 thicker than, those in midlatitudes. Some features of the equatorial PBL described here have not been found yet in midlatitudes. These prove that the highresolution observations with the B LR must play an important role in various studies of the equatorial lower atmosphere. In subsequent papers we will compare the B LR data with cloud distributions observed with a meteorological satellite and prevailing large-scale winds such as the Indian Ocean monsoon and the Pacific Ocean trade wind. More quantitative analysis such as examination of power spectra and transport amounts should be done based on the B LR data. Acknowledgments. We acknowledge S. Kato of Japan-Indonesia Science and Technology Forum (JIF) for his ceaseless encouragement of this observation campaign. We are grateful to T. Sato of Department of Electrical Engineering II, Kyoto University, M. Yamamoto, T. Nakamura, and T. Adachi of RASC, Kyoto University, for their useful suggestions and comments. We thank G. Fraser (on leave from the University of Canterbury, New Zealand) and W. O. J. Brown of RASC and the two reviewers (W. L. Ecklund and P. T. May) for their careful reading of the original manuscript with constructive comments. Thanks are also due to A. Nurza- man of LAPAN and T. Shimomai and N. Yoshino of RASC for their cooperation for the dry season campaign. We thank all the operators who are maintaining the Serpong radar observatory, and colleagues of B PPT, LAPAN, and RASC for their collaborations. The first author (Hashiguchi) is supported by a grant (1289) of the Japan Society for the Promotion of Science (JSPS) under the Fellowships for Japanese Junior Scientists. The present study was financially supported by Grants-in- Aids ( , 04NP0201, 05NP0201) of the Japanese Ministry of Education, Science and Culture. References Angevine, W. M., A. B. White, and S. K. Avery, Boundary-layer depth and entrainment zone characterization with a boundary-layer profiler, Boundary Layer Meteorol., 68, , Browning, K. A., and D. Atlas, Velocity characteristics of some clear-air dot angels, J. Atmos. Sci., 23, , Ecklund, W. L., D. A. Carter, and B. B. Balsley, A UHF wind profiler for the boundary layer: Brief description and initial results, J. Atmos. Oceanic Technol., 5, , Ecklund, W. L., D. A. Carter, B. B. Balsley, P. E. Currier, J. L. Green, B. L. Weber, and K. S. Gage, Field tests of a lower tropospheric wind profiler, Radio Sci., 25, , Firestone, J. K., and B. A. Albrecht, The structure of the atmospheric boundary layer in the central equatorial Pacific during January and February of FGGE, Mon. Weather Rev., 114, , Garratt, J. R., The Atmospheric Boundary Layer, 316 pp., Cambridge University Press, New York, Hashiguchi, H., M.D. Yamanaka, T. Tsuda, M. Yamamoto, T. Nakamura, T. Adachi, S. Fukao, T. Sato, and D. L. Tobing, Diurnal variations of the planetary boundary layer observed with an L-band clear-air Doppler radar, Boundary Layer Meteorol., in press, Holton, J. R., Troposphere-stratosphere exchange of trace constituents: The water vapor puzzle, in Dynamics of the Middle Atmosphere, edited by J. R. Holton and T. Matsuno, pp , D. Reidel, Norwell, Mass., Kropfii, R. A., The atmospheric boundary layer: Panel report, in Radar in Meteorology: Battan Memorial and 40th Anniversary Radar Meteorology Conference, edited by D. Atlas, pp , American Meteorological Society, Boston, Mass., Lhermitte, R. M., Probing air motion by Doppler analysis of radar clear air returns, J. Atmos. Sci., 23, , May, P. T., and J. M. Wilczak, Diurnal and seasonal variations of boundary layer structure observed with a radar wind profiler and RASS, Mon. Weather Rev., 121, , Nitta, T., T. Mizuno, and K. Takahashi, Multi-scale convective systems during the initial phase of the 1986/87 El Nifio, J. Meteorol. Soc. Jpn., 70, , Rogers, R. R., W. L. Ecklund, D. A. Carter, K. S. Gage, and S. A. Ethier, Research applications of a boundarylayer wind profiler, Bull. Am. Meteorol. Soc., 74, , R6ttger, J., and M. F. Larsen, UHF/VHF radar techniques for atmospheric research and wind profiler ap-

12 1054 HASHIGUCHI ET AL.: PLANETARY BOUNDARY LAYER OBSERVATIONS plications, in Radar in Meteorology: Battan Memorial and 40th Anniversary Radar Meteorology Conference, edited by D. Arias, pp , American Meteorological Society, Boston, Mass., Stull, R. B., An Introduction to Boundary Layer Meteorology, 666 pp., Kluwer Academic, Norwell, Mass., Tsuda, T., Y. Murayama, H. Wiryosumarto, S. Kato, S. W. B. Harijono, S. Fukao, M. Karmini, C. M. Mangan, S. Saraspriya, and A. Suripto, A preliminary report of radiosonde observations of the equatorial atmosphere dynamics over Indonesia, J. Geomagn. Geoelectr., 44, , Tsuda, T., Y. Murayama, H. Wiryosumarto, S. W. B. Harijono, and S. Kato, Radiosonde observations of equatorial atmosphere dynamics over Indonesia, 1, Equatorial waves and diurnal tides, J. Geophys. Res., 99,10,491-10,505, 1994a. Tsuda, T., Y. Murayama, H. Wiryosumarto, S. W. B. Harijono, and S. Kato, Radiosonde observations of equatorial atmosphere dynamics over Indonesia, 2, Characteristics of gravity waves, J. Geophys. Res., 99, 10,507-10,516, 1994b. Tsuda, T., et al., A preliminary report on observations of equatorial atmosphere dynamics in Indonesia with radars and radiosondes, J. Meteorol. $oc. Jpn., in press, Yamada, T., and G. Mellor, A simulation of the Wangara atmospheric boundary layer data, J. Atmos. $ci., 32, , S. Fukao, H. Hashiguchi, T. Tsuda, and M.D. Yamanaka, Radio Atmospheric Science Center, Kyoto University, Uji, Kyoto 611, Japan. ( Hasiguti@ kurasc.kyoto-u.ac.jp) S. W. B. Harijono, T. Sribimawati, and D. L. Tobing, Agency for the Assessment and Application of Technology, Jakarta Pusat 10340, Indonesia. H. Wiryosumarto, National Institute of Aeronautics and Space, Jakarta Timur 13220, Indonesia. (Received May 13, 1994; revised November 10, 1994; accepted November 30, 1994.)