S3-6 Lidar Observation of Atmospheric Processes in Hong Kong

Andrew Y. S. Cheng(1) , Heping Liu(2) and Johnny C. L. Chan(1)(2)

(1)Department of Physics and Materials Science, City University of Hong Kong, Hong Kong, China(2)Centre for Coastal Pollution and Conservation, City University of Hong Kong, Hong Kong, China

Phone: (852)2788-7810, Fax: (852) 2788-7830, Email: apaysc@cityu.edu.hk

1 Introduction.

The planetary boundary layer (PBL) is definedas the part of the troposphere that is directlyinfluenced by the earth's surface, and respondsto the surface forcing (Stull, 1988). The PBL ischaracterized by a high aerosol and water vaporconcentration. In comparison, a relatively lessaerosol and moisture concentrations are found inthe free atmosphere, leading to a sharp contrastin lidar backscatter signal and making it possibleto determine the PBL height. In addition,different aerosol and moist concentrations can beanalyzed to show different energy and masstransport processes. Cooper and Eichinger(1994) have studied the urban PBL structureswith a scanning lidar and radiosondeobservations, presenting the detailed patterns ofthe nocturnal and daytime PBL structures, andshowing the energy and mass exchangecharacteristics.The Lidar system in the City University ofHongKong has been developed for the atmosphericstudy over Hong Kong area. Some results havebeen presented including lidar measurement ofwind velocity by tracking a single cloudmovement, the study of the thermal internalboundary layer (TIBL) growth under sea-breezecirculation, and PBL structures during cold frontpassage. In this paper, these results have beensummarized to show the research activities inHong Kong. Finally, the future researchobjectives are also discussed.

2. Lidar System

The lidar at the City University ofHong Kong isa ground-based, backscattered system, consistingof a frequency-doubled Nd:YAG laser operatedat 532 nm wavelength and lOHz repetition rate.The pulse duration is 10ns and the rangeresolution can be up to 3m. The laser beam islaunched to the atmosphere in the directioncontrolled by a PC computer-controlled scanner.The backscatter signal by aerosol is collected bya Newtonian telescope with 1.5m focal length.The signal intensity is proportional to thevolumetric backscattering coefficient of theaerosol particles, atmospheric attenuation, andrange correction (Cooper and Eichinger, 1994).Assuming a relationship between the volumetric

backscattering coefficient and attenuationcoefficients of the atmosphere, we can obtainrelative extinction coefficient along the path.Generally, the extinction coefficient isassociated with the aerosol particleconcentration, size, and shape. Therefore, lidar-derived profile can produce a detailed picture ofthe PBL structures and characteristics. Inparticular, due to sparse radiosonde data, it isvery difficult to analyze the PBL evolution.However, Lidar can provide us temporalcontinuous measurements. In addition, spatialscanning can also show us the two-dimensionalPBL structures.The lidar system is located on the top of a nine-story building which is about 57m high abovethe sea level, and within the Kowloon city centreof Hong Kong. The site is about 3km northwestof the Hong Kong Observatory (HKO) whereother meteorological data including radiosondedata are obtained (See Figure 1). Somecomparisons between the lidar and radiosondedata have been made to show their goodagreement in determination of the mixing layerheight (Mok et al. 1998).

Fig. 1 Map of the Kowloon City ofHong Kong

3 Results

(1). Lidar Measurement of Wind Velocity

The principle of lidar measurement of windvelocity is realized by tracking a single cloudmovement in a given time period. Generalconsideration, the velocity of cloudmeasurement can indicate wind velocity. Leunget al (1995) have presented their measurementresults including the change of wind direction

67

and velocity with time and the comparisonresults of lidar and balloon wind measurementsin Hong Kong (See Figure 2). A good agreementhas been obtained to show the applicableperspective of lidar measurements of windvelocity. The error is mainly dependent upon theuncertainty of cloud displacement measurementdue to the change of the cloud properties withtime, but decreasing the measurement period cancontrol this uncertainty.

Wind direction2 3 0

2 1 0

1 9 0

1 7 0

*

* * / *l ̂ ^ ^ ^ ^ ^ ^ ^ ^ l B K fl ^ H

* **

* #

1 5 0T 3

* / ** / *

1 3 0CQ

I ̂ ^ ^ ^ K ^ ^ ^ ^ ^ M

o5K ?K jfit Vfi 5*C

1 10 #

9 0

7 0

* / * *

^ m okI I I I I I I I

50

50 70 90 110 130 150 170 190 210 230

Lidar (deg.)

Wind Speed (m/s)

Fig. 2 A comparison of lidar and balloon windmeasurement in Hong Kong (after Leung et al.,1995)

(2). TIBL Growth under Sea-breeze Circulation

Slab models provide a physical approach to theprediction of TIBL growth, and have been usedwidely in coastal air pollution dispersionbecause of their simplicity. The slab modelshave successfully predicted growth of the TIBLin many coastal regions. In order to void the

singularity problem with near-neutral onshoreflow, Luhar (1998) present an analytical slabmodel that can be used for neutral to stronglystable onshore flows. It has been tested toperformance well in describing the growth of theTIBL in both the neutral and stable stratificationusing the Kwinana Coastal Fumigation Study(Luhar et al., 1998). However, the model needsto be tested by lots of measurements in differentcoastal sites especially with complex coastal lineand complex terrain. Also, it should be testedagainst the neutral onshore flow which is alwaysencountered in Hong Kong (Liu et al., 1999a).However, it is very difficult to measure TIBLheight growth with the inland distance bylaunching radiosondes in many sites, while lidarcan realized it by two dimensional scanningalong the wind direction to obtain TIBL heightsin different inland distances.

à"5 600-1

sz_J

4000 8000 1 200Inland fetch (m)

Fig. 3. The comparison of the TIBL heightgrowth of the slab model prediction and lidarobservation in Hong Kong on 1400 (LT) 31March 1999. (Solid line: model calculation;Triangle: lidar measurement).

Figure 3 presents the comparison results of theslab model and lidar measurements of the TIBLheights at 1400LT 31 March 1999. It clearlyshows that the TIBL height grows rapidly nearthe coastal line. As the downwind distanceincreases, the slow TIBL height growth is found.In general, Figure 3 indicates that the slab modelestimates of the TIBL height growth are smallerthan those by lidar measurements probably dueto the error of onshore flow stabilities atdifferent heights that are obtained from theradiosonde data in HKO. Meanwhile, we foundthat TIBL heights are non-spatially uniform dueto non-uniform heating of underlying surfacesand the effect of the complex terrain withnumerous hills, which can not be identified bythe slab model. Furthermore, It is expected thatthe TIBL might be three-dimensional structureover Hong Kong area due to its peninsula-liketerrain, which will be investigated in the future.

68

Under this condition, this complex structurewould require a three-dimensional numericalmodel (Cheng et al., 1999) and high spatialresolution measurements by lidar and othertechniques.

(3). PBL Structure

a. A case study of nocturnal mixing layer

Most cities are anthropogenic sources of heatand pollution. Large coverage of buildings andstreets can be found in the Kowloon City ofHong Kong, making both the surface albedoesand heat capacity larger than that in the ruralarea. Therefore, the air in the surface layer of theurban area is warmer than its surrounding, whichgives rise to urban heat island (Stull, 1988; Oke,1987). Figure 4 shows the temperaturesmeasured in HKO (urban) and Lau Fu Shan(rural). The temperature in the urban area isalways greater than that in the rural area.

12

0 6 12 18 2Time (LT)

Fig. 4 Diurnal variation of the temperatures inthe urban (solid line) and the rural areas (dashedline) in Hong Kong on 5 January 1998.

Lidar measurement has indicated. the mixinglayer height reaches its maximum of around1100m at around 1600LT with the maximumground temperature at that time. After sunset, theturbulence begins to decay due to no fluxtransport from the surface. With the cooling ofthe ground surface, stably nocturnal boundarylayer forms that can be found from theradiosonde data obtained in HKO. From Figure5, we can see that the stable boundary layerheight can be estimated about 150m with theresidual layer aloft up to 1000m where thecapping inversion layer is located. At around1800LT, the ground surface temperature

suddenly decreases slowly, and there is nochange at 2000LT and 2100LT. After 2200LT,the temperature begins to rise up to 17.6°C at2300LT and 17.8°C at 2400LT. While the

temperatures in its rural keep decreasing from16.7°C at 1600LT down to 13.4°C at 2400 LT.

The temperature in the urban is 4.4°C higherthan its rural area at 2400LT. At the same time, ashallow convective mixing layer was observedafter 2 100LT from the lidar data.

3500 -

3000 -=

2500 -=

2000 -=

à"§? 1500X

1000

500

0 I IT I I 1 1

285 290 295 300 30

Virtual potential temperature (K)

Fig. 5. The radiosonde measurement of the PBLstructure at 2000LT of 5 January 1998.

1500

1200

E 900

sz2 600

300

0.0000 0.0002 0.0004 0.0006 0.0008

Extinction coefficient (relative unit)

Fig. 6. Lidar measurement of the nocturnalconvective mixing layer at 2200LT of 5 January1998.

Figure 6 shows an example of the lidarobservation at 2200LT. The convective mixinglayer height can be obviously observed around700m. Such a high nocturnal mixing layer isunusually observed in other urban sites. Becauseof the shallow stable boundary layer (around

69

150m), the rapid growth can be found in theresidual layer once the convective boundarylayer erodes and reaches the top of the stableboundary layer. At that time, the fumigation isoccasionally found to be similar to the morningmixing layer erosion because the pollutant in theresidual layer is transported down to the surface.

b. PBL structures during a clod front passage

The PBL structures are not only influenced bylocal turbulent transports, but also by advectiveactivities, subsidence and rising. In winter, thecold air activities have occasional affect on thePBL structures in Hong Kong. The evolutionprocesses of the PBL have been studied during acold front passage by integrating both thestandard meteorological data and the lidarobservations on 4 January and 5 January 1998. Ithas given a better understand the heat and masstransport over complex terrain and under theconditions of dry cold advective and warmmoistair advective (Liu et al, 1999b). In the differentstages of the cold front passage, the PBLstructures change significantly due to subsidenceand rising associated with the cold air advectionand the local urban heat island climate. Becauseof the sparse radiosonde data, the complex PBLstructures and their evolution could be analyzedmore reliably by integrating standardmeteorological data and lidar data, rather than bytraditional techniques alone.

4 Summary

The complex atmospheric processes occurring inthe PBL leads to the different aerosol and watervapor concentration profiles. Lidar system canbe used for determining the PBL structure bycoupling with other meteorological data. Itscontinuous measurement can provide us muchuseful information such as wind speed, winddirection, the TIBL height growth, the PBLstructures under different weather conditions.Our study will focus the vertical structures undersea- and land- breeze conditions, threedimensional TIBL structure and urban PBLcharacteristics.

5 Acknowledgements

Wewould like to acknowledge the contributionsof Felix Fong for his help in data processing.This work is partly supported by CERG Grant(Grant 9040396), the Strategic Research Grantsof the City University of Hong Kong (Grants7000429 and 7000646) and the DirectionAllocation Grant of City University of HongKong (Grants 7100002 and 7100064).

References

Cheng Y. S. Andrew, Heping Liu, Johnny C. L.Chan, and Felix Fong. (1999). Numericalsimulation and lidar measurements of theinternal boundary layer evolution overHong Kong urban area. To be published inthe Proceedings of the 15th InternationalCongress of Biometeorology &International Conference on UrbanClimatology, Sydney, Australia, 8-12November 1999.

Cooper, D. I. and Eichinger W. E. (1994).Structure of the atmosphere in an urbanplanetary boundary layer from lidar andradiosonde observations, J. G. R. 99 (Dl l),22937-22948.

Kiemle, C, Kaestner, M., and Ehret, G. (1995),The convective boundary layer structurefrom lidar and radiosonde measurementsduring the EFEDA'91 campaign, J. Atmos.& Oceanic Tech 12, 771-782.

Leung, K. M., Lin, T. J., and Chan, J. C. L.(1995) Lidar measurement of WindVelocity by Tracking a Single CloudMovement, Chinese Journal of Lasers,22(1 1), 842-846.

Liu Heping, Andrew Y. S. Cheng, Johnny C. L.Chan and Felix Fong. (1999a). Internalboundary layer growth over Hong Kongurban area under sea breeze conditions:lidar measurements and model estimates.To be published in the Proceedings of15thInternational Congress of Biometeorology& International Conference on UrbanClimatology, Sydney, Australia, 8-12November 1999.

Liu Heping, Johnny C. L. Chan and Andrew Y.S. Cheng (1999b). A case study of theplanetary boundary layer structures duringa cold front passage in Hong Kong, To besubmitted to Boundary Layer Meteorology.

Luhar A. K. (1998). An analytical slab model forthe Growth of the coastal thermal internalboundary layer under near-neutral onshoreflow conditions, Boundary LayerMeteorology, 88, 103-120.

Mok, T. M., K. M., Leung, J. C. L. Chan, A. H.

P. Ho, and C. N. Ng, (1998). Lidar

observation of the nocturnal Inversiontransition in the lower atmosphere overHong Kong, SPIE, Optical RemoterSensing for Industry and EnvironmentalMonitoring, Beijing, September 1998.

Stull, R. B. (1988). An Introduction to BoundaryLayer Meteorology, Academic, San Diego,California.

70