An observational study of the planetary boundary layer height at the central nuclear

Transcription

1 An observational study of the planetary boundary layer height at the central nuclear de Almaraz, Spain J.A. Garcia", M.L. Cancillo", J.L. Cano\ C. Vague* Dto. de Fisica, Universidad de Extremadura, Badajoz, Spain ^Dto. de Ciencias Atmosfericas, Universidad Complutense, Madrid, Spain Abstract The evolution of the mixing layer height, both diurnal and nocturnal and the nocturnal planetary boundary layer height has been evaluated at the Central Nuclear de Almaraz, Spain in a winter period. Results show that a surface-based stable boundary layer forms after sunset and grows during the night, and that it is completely eroded by convection only after midday. On the other hand the mixing layer height, which is lower than the nocturnal planetary boundary layer, keeps nearly steady during the night, and then grows due to convection even on a cloudy day 1 Introduction Knowledge of the planetary boundary layer (PBL) height and its time evolution is of major interest in boundary layer modelling and its parameterization. It is also very important in the study of pollutant dispersion because it is used as an external parameter in air quality and dispersion modelling (e.g. Panmier et. al. [6], Sullivan et. al. [8]). The purpose of the present paper is to report on the observational study undertaken at the Central Nuclear de Almaraz, Spain. The objective of the program was a preliminary study of the boundary layer height in winter situations using potential temperatures and bulk Richardson number profiles. The planetary boundary layer height is defined as the vertical extent of the troposphere directly influenced by the presence of the earth's surface responding to surface forcing with a timescale of about one hour or less (Stull)[7]. In environmental meteorology, there is also introduced the mixing layer (ML) height, as the extent of the troposphere where vigorous vertical mixing of pollutants occurs. The PBL thickness varies over the course of the day, from a few hundred meters at dawn to a few kilometers about midday. Not only are changes produced in height but also in its dynamic structure. For example, vertical mixing due to turbulence within the nocturnal boundary layer is much less than during daytime, due to the effects of negative buoyancy associated with nighttime surface temperature inversion.

2 348 Pollution Control and Monitoring During the day the PBL top is usually identified with the base of an elevated inversion or stable layer capping a well-mixed convectively driven boundary layer. In these conditions, this top is also the ML top. In order to evaluate the PBL(ML) height, we shall suppose that the PBL top is reached once the bulk Rich ad son number defined by *< >- f "v "'?";* ' ["raarrj "> has reached the critical value Rib^r 0.5 (see Troen and Mahrt [9], Andre and Mahrt [1]). The height so determined will be denoted h^. The variable 0^ is the virtual potential temperature of the air plumes near the ground, and has been chosen to be the virtual potential temperature of the first level of the observed sounding (5 m), because the ground heating in winter (when the experiment was carried out) is not too important.also, i/max is the maximum wind speed between 0 and }\m- We have taken the maximum value of u, instead of u(h), in order to simulate better the mechanical generation of turbulence in the layer (0, h^). Equation (1), can be written as It can be seen from (2) that the height h, coincides with that obtained from the intersection of the surface adiabatic with the observed virtual potential temperature profile (Gv(hm) = QVS) if the maximum wind speed u^ax is small, i.e., if the generation of mechanical turbulence is negligible. In well-developed convective situations, the influence of the second term in (2) is very small and the election of Ribcr is not critical. On the contrary, in near-neutral situations, h, is quite dependent on the chosen value of Rib^- Another method to identify the ML height is to look at the influence of the surface heating by examining two consecutive virtual potential temperature soundings. During the night or in the case of a surface-based stable boundary layer, neither the PBL height nor the ML height are as well defined as in the convective case. In these situations the buoyancy forces are opposed to the effects of the mechanical turbulence, weakening the mixing processes. However, near the ground, where the shear can be high, there subsists a certain level of turbulence, high enough to keep this layer well mixed, transferring heat from above to the cooled ground and extending this cooling upward, at the same time as pollutants emitted near the ground level are mixed. Surface cooling is also transmitted to the atmosphere through radiative transfer, so that the nocturnal boundary layer depth (NBL), defined as the height up to which cooling of the atmosphere due to the presence of the ground is important, could be much greater than the ML depth. Thus, during the night the planetary boundary layer could be considered as multilayer, with a turbulent layer near the ground and a nonturbulent layer above but with significant cooling. In this simplified picture of the nocturnal planetary boundary layer, we are neglecting effects such as horizontal advection, katabatic winds, high level turbulence associated with the low level wind maximum (LLJ), discontinuous turbulence, etc. The top of the ML is usually defined as the height at which turbulence, measured through the heat flux or the momentum flux, decreases to a small fraction, say

3 Pollution Control and Monitoring 349 5%, of its surface value. High resolution profiles of turbulent parameters covering the range 50 to 500 m are needed to use this definition. Such measurements are only possible at very selected sites or during special boundary layer experiments (e.g. Niewstadt [2]). An easier way to estimate the turbulence level is through the gradient Richardson number Ri as this only requires knowing the temperature and speed profiles. Theoretically, when this number is less than or greater than a critical value (usually between 0.25 and 1), the atmosphere flow becomes turbulent or laminar respectively. Unfortunately its proper evaluation requires a high degree of accuracy in the speed measurements as its square gradient appears at the denominator in the Ri definition. A similar number that does not suffer from this problem is the bulk Richardson number Rib. Again, unfortunately different definitions can be found in the literature, but for continuity with the diurnal cases, we shall use here that given in ( 1). With reference to the nocturnal boundary layer depth, there are several definitions in the scientific literature, among which can be cited: The surface inversion height h,-, proposed by Yu [10] and defined as the height where the temperature gradient reaches its dry adiabatic value. The problem with this definition is that it does not include the thicker layer of significantly stratified air situated above. The height h# to which significant cooling has extended, as judged from the time evolution of the potential temperature profile (Melgarejo and Deardorff) [5]. In this case the problem is that two soundings are needed which could lead to considerable noise. # The height hg where the influence of surface cooling is transmitted, evaluated through there being a significant increase of the stratification (Mahrt et. al.) [3]. This is a graphical method. The problem here is that it sometimes is difficult to determine where the beginning of the stratification is located. * The height h^ where the potential temperature gradient jp exceeds 3.5 W~*K m-i (Andre and Mahrt) [1]. In this paper, the graphical hg, and numerical h^ will be used. 2 Data Data were gathered between February 8th and February 12th at the Central Nuclear de Almaraz, Spain (39 45'N, 5 40'W, 225 m ASL) with an AIR, USA, tethersonde system and a 15 m meteorological mast with a sonic anemometer mounted on its top and more conventional equipment along it. The tethersonde system provides profiles of temperature, pressure, mixing ratio, wind speed, wind direction and height above the ground up to a maximum height of 1000 m. Measurements were taken, when it was possible, 6 times a day at about: 9, 11, 14, 16, 18 and 24 hours, Local Standard Time (UTC + 1). Each ascent takes about 20 minutes. All times indicated in the diagrams and in the text are Local Standard Time (LST) and show the beginning of each sounding. Sunrise and

4 350 Pollution Control and Monitoring sunset took place at 0825LST and 1850LST. During the observational period a frontal system swept the Iberian Peninsula from 10 February toll February with heavy rainfalls over Almaraz during these two days, so that data taken on these days have not been analysed. February 8th was a clear sunny day with low to light winds, February 9th was a cloudy day with eight oktas during the whole day and February 12th was a clear sunny day with moderate to high winds. 3 Results and Discussion 3.1 Day 8 First, the nocturnal boundary layer height will be determined and then the mixing layer height both nocturnal and diurnal. Figure 1 shows virtual potential temperature profiles. In an attempt to identify the top of the layer with significant cooling and calculate hg, the residual layer has been extrapolated to the ground by a dotted line. According to Figure 1, the height hg at 0908LST was m. A similar value was obtained for h^. The inversion strength, measured as the difference of virtual potential temperature between Jig and the ground, was 9 C. At 1108LST hg kept stationary at m, but the inversion intensity had diminished to 3.5 C, having begun, in the layer near the ground, the erosion of the surface inversion with a superadiabatic stratification in the first 50 m. At 1821LST a new NBL has been formed with a height hg of 350 m. This value is uncertain because it is difficult to decide how to extrapolate the residual layer. The cooling depth h# and the stable layer depth h^ give 110 m which seems a more realistic value of the NBL for that time of the day. With reference to the mixing layer height h^, Figure 2 shows the bulk Richardson number profiles, and, for comparison, the Rib = 0.5 isoline. Table 1 shows the values obtained for h^. Table 1: Mixing Layer Height, 8 February time: 0908LST 1108LST 1322LST 1556LST 1821LST height: 20 m 275m 600m 815 m 100m As Figure 2 shows, the ML height does not depend on the critical value chosen for the Rib except for the 1108LST sounding, where, due to a near-neutral layer extending from 100 to 300 m, a small variation in Rib^r gives rise to very different values for h^. It is just in this situation where the height obtained through the critical bulk Richardson number (h^=275 m) differs from the height obtained through the more usual method of extending the surface adiabatic up its intersection with the sounding (h = 50 m). Figure 3 shows all soundings together. Comparing 0908LST and 1108LST profiles, it may be seen that at 1108LST the heating has reached a height of 275 m, which is identical to that of Table 1, what supports the validity of the method. Moreover, comparing the 1322LST and 1108LST profiles, it may be seen that the heating has affected a larger layer than that obtained for h^ in Table 1 at 1322LST. Unfortunately, both profiles

5 Pollution Control and Monitoring 351 are too short to evaluate the influence of this heating. These differences could be justified by heat advection, as a significant wind direction change took place between the two soundings. 3.2 Day 9 Figure 4 shows the profiles of virtual potential temperature; hg values for the NBL are: m at 0921LST and 300 m at 2409LST. At 1057LST, h, is not well defined but the two profiles (0921LST and 1057LST) are quite similar, so one has taken hg = m at this time. At 1953LST, the inversion top could not be reached because the wind was quite strong and the sounding had to be stopped at 300 m for security reasons. With reference to the ML height, Table 2 shows the values calculated for h^. The heights obtained during the day (1057LST, 1355LST, 1953LST) are similar to those obtained comparing its virtual potential temperature profiles. The 1953LST sounding is quite singular due to the high wind. Although the profile was statically stable, the whole layer spanned by the sounding could be considered turbulent, so that it appears in Table 2 with a height h^ greater than 285 Table 2: Mixing Layer Height, 9 February time: Q921LST 1057LST 1355LST 1558LST 1953LST 2429LST height: 125 m 85 m 425 m 570 m > 285 m 190 m 3.3 Day 12 Figure 5 shows virtual potential temperature profiles for 12 February. On this occasion it is quite difficult to calculate the NBL height (hg) because the residual layer is not well defined. In Figure 5 there appear two attempts to calculate it at 0835LST and 2435LST, and the results are 300 and 225 m respectively. With reference to the ML height, Table 3 shows its values for this day. Table 3: Mixing Layer Height, 12 February "time: Q835LST 1Q59LST 1345LST 2Q39LST 2435LST height: 370 m 380 m > 430 m > 340 m > 335 m As may be seen from this Table, the heights obtained at the statically stable situations (0835LST and 2435LST) are much greater than for the previous days. This is due to the mechanical turbulence generated by the strong winds observed. Figure 6 shows the profiles of wind speed. As may be seen at 0835LST, there was a sharp low-level jet at about 300 m. This jet was still present at 1059LST, though it is not so localized. As the morning develops, the jet spreads over

6 352 Pollution Control and Monitoring the whole layer, probably due to convective mixing. With the beginning of the night, the convection ceases and a new jet seems to settle in. Unfortunately the sounding had to be stopped due to the high wind and the jet is not fully resolved. 4 Summary and Conclusions The evolution of the mixing layer and the nocturnal boundary layer height have been investigated by analyzing virtual potential temperature and bulk Richardson number profiles taken at the Central Nuclear de Almaraz in a winter period. Our results show that there exists a surface-based stable layer which forms at sunset and grows during the night as may be seen by comparing hg at 1825LST ( 8 February) and h,, at 0921LST (9 February) (see Figure 7) with maximum heights of meters. On the other hand, comparing values for h^ at these same times as for h,,, the ML height does not grow during the night but instead keeps nearly constant. The surface-based stable layer is completely eroded by convection only after midday, except on 12 February when the erosion took place before, probably due to the high winds observed. It is also noteworthy that a convective boundary layer develops even on a cloudy day as occurs on 9 February (see Figure 4). The maximum height observed for the ML was about 815 m on a sunny day (8 February) and 570 m on a cloudy one (9 February). Figure 7 shows the evolution of the ML and nocturnal boundary layer height for the three clays studied. A cknowledgeinent s We would like to thank Dr. V. L. Mateos and J. Santana for their helping during the measurement period. Thanks are also due to the Central Nuclear de Almaraz for its financial support. References [1] Andre, J.C and L. Mahrt, 'The Nocturnal Surface Inversion and Influence of Clear-Air Radiative Cooling', J. Atrnos. Sci., Vol. 39, pp , [2] Niewstadt, F.T.M., 'The Turbulent Structure of the Stable Nocturnal Boundary Layer', /. Atmos. Sci., Vol 41, pp , [3] Mahrt, L., R.C. Heald, D.H. Lenschow, B.B. Stankov, I. Troen, 'An Observational Study of the Structure of the Nocturnal Boundary Layer', Bound.- Layer Meteor., Vol 17, pp , [4] Mahrt, L., 'Modelling the Depth of the Stable Boundary Layer', Bound.- Layer Meteor., Vol. 21, pp. 3-19, [5] Melgarejo J.W. and J.W. Deardorff, 'Stability Functions for the Boundary- Layer Resistance Laws Based upon Observed Boundary-Layer Heights', J. Atmos. Sci., Vol. 31, pp , 1974.

7 Pollution Control and Monitoring 353 [6] Paumier J. O., S. G. Perry, D.J. Burns, 'CTDMPLUS: A Dispersion Model for Sources near Complex Topography. Part I: Technical Formulations', J. Appl Meteor., Vol. 31, pp , [7] Stull, R.B., An Introduction to Boundary Layer Meteorology, Kluwer Academic Publisher, Dordercht, The Netherlands, [8] Sullivan, T. J. et. al., 'Atmospheric Release Advisory Capability: Real-Time Modeling of Airborne Hazardous Materials', Bull Amer. Meteor. Soc, Vol. 74, pp , [9] Troen, I. and L. Mahrt, 'A Simple Model of the Atmospheric Boundary Layer; Sensitivity to Surface Evaporation', Bound.-Layer Meteo., Vol 37, pp , [10] Yu, T.W., 'Determining Height of the Nocturnal Boundary Layer', J. Appl. Meteor., Vol. 17, pp , 1978.

8 354 Pollution Control and Monitoring Hour: Hour: ,gi 'o> x gi X Hour: Hour: ni Hour: IT Figure 1: Virtual potential temperature profiles for 8 February Height is above ground level. Hour shows the beginning and end of each sounding. Dotted lines are attempts to extrapolate the residual layer.

9 Pollution Control and Monitoring Hour: Hour: OL Rib Hour: Hour: T 600!c cr ' Hour: Rib Rib Figure 2: Bulk Richardson numbers profiles for 8 February Height is above ground level. Hour shows the beginnig and end of each sounding. The isoline Rib = 0.5 has been drawn for comparison. Rib greater than 4 has been reset to 4.

10 356 Pollution Control and Monitoring E 600 _c cr> Figure 3: Virtual potential temperature profiles for 8 February 1993.

11 Pollution Control and Monitoring 357 Hour: Hour: Hour: Hour: Hour: Figure 4: Virtual potential temperature profiles for 9 February Height is above ground level. Hour shows the beginning and end of each sounding. Dotted lines are attempts to extrapolate the residual layer.

12 358 Pollution Control and Monitoring 600 Hour: Hour: I Hour: G.('C) Hour: Hour: G.('C) Figure 5: Virtual Potential Temperature profiles for 12 February Height is above ground level. Hour shows the beginning and end of each sounding. Dotted lines are attempts to extrapolate the residual layer.

13 Pollution Control and Monitoring 359 Hour: Hour: Figure 6: Wind profiles for 12 February 1993.

14 360 Pollution Control and Monitoring 1 UUU o Q 0 i 0» 288 ) o,0,,,,,,,, I3 < ; i8 1 Day Figure 7: Time evolution of the mixing layer heights (o) and nocturnal boundary layer heights ( ) for the three days under study.