Abstract

Two years of Doppler sodar measurements are used to study the time–height structure of the nocturnal boundary layer (NBL), its seasonal variation, and the characteristics of different types of NBL. A total of 220 clear-sky nights during which the inversion layer is clearly visible on a sodar echogram are examined. The NBL depth estimated with sodar data using a wind maxima criterion matches reasonably well with radiosonde-based NBL depth estimates. The NBL exhibits clear seasonal variation with greater depths during the monsoon season. Shallow NBLs are generally observed in winter. The evolution of NBL height shows two distinctly different patterns (called type 1 and type 2), particularly in the second half of the night. Type 1 NBL depth is nearly constant and the wind speed in this type is generally weak and steady throughout the night, while type 2 is characterized by moderate to strong winds with considerable variations in NBL height. The local circulation generated by the complex topography is clearly seen in type 1 throughout the night, whereas it is seen only in the first half of the night in type 2. Type 1 NBLs seem to be more prevalent over Gadanki, India, with nearly 61% of total nights showing type 1 characteristics. Furthermore, type 1 NBL shows large seasonal variability with the majority of type 1 cases in winter. The type 2 cases are mostly observed in monsoon (~60%) followed by summer (39%). The surface meteorological parameters during type 1 and type 2 cases are examined. Differences between type 1 and type 2 NBL patterns are discussed in relation to the surface forcing.

1. Introduction

The development of a stable layer near the earth’s surface during the night by radiative cooling is known as the nocturnal boundary layer (NBL) (or stable boundary layer). Knowledge of the mean wind and turbulence in the NBL over complex terrain is of great importance for dispersion studies. The structure and evolution of the NBL is determined by various processes such as radiative cooling, turbulent mixing, advection, and gravity waves. The nocturnal boundary layer is usually classified according to the thermal stratification, ranging from weak to strong stability stratification (Coulter 1990; Mahrt et al. 1998; Conangla et al. 2008). The strong stable stratification over the land occurs in the presence of light winds and clear-sky conditions, which results intense surface radiative cooling. The other notable feature of a strong stable stratification regime is the presence of weak turbulence, which is largely determined by local effects such as the topography or the terrain heterogeneities (Mahrt et al. 2001; Soler et al. 2002). The weak stable regime over the land occurs with moderate to strong winds producing strong mechanical turbulence and mixing, or with overcast skies that reduce the radiative cooling of the earth’s surface. The classification of these NBL regimes is somewhat difficult (Mahrt et al. 1998). Sometimes, a particular night contains both types of stratification, with periods of weak stratification and strong turbulence as well as periods of light winds with weak turbulence (Conangla et al. 2008).

Identification of NBL top from routine measurements is a classical longstanding problem. Traditionally, it is determined based on turbulence characteristics (Caughey et al. 1979; Wyngaard 1986; Mahrt and Vickers 2006). The depth of the NBL can also be determined by using the low-level wind maxima (Stull 1988; Beyrich and Weill 1993; Murthy et al. 1996). Beyrich and Weill (1993) estimated the NBL height based on digitized profiles of sound detection and ranging (sodar) backscattered echo intensity. They showed that the sodar-derived NBL height matches closely with different NBL depth scales only during certain phases of the night. For instance, the sodar-NBL height matches the lower part of growing surface inversion during the first hours of the night and with the height of the low-level wind maximum later on. Recently, Pichugina and Banta (2010) reviewed different methods for the determination of NBL depth using remote sensing devices. They primarily compared wind-based (wind maxima, minimum, magnitude in the mean wind shear, and first minimum in the curvature profile of mean wind) NBL estimates with turbulence-based (variance) NBL estimations when the vertical structure of low-level jet (LLJ) is in different shapes (“nose,” “flat,” etc.). Overall, the agreement is good between different methods (Pichugina and Banta 2010).

The NBL structure and its evolution have been extensively studied over the past two decades using remote sensing instruments, like sodar and light detection and ranging (lidar), and also with in situ measurements (towers, balloon soundings, aircraft, etc.). Kurzeja et al. (1991) reported the time–height variation of NBL structure over Augusta in different wind regimes using measurements made with a 300-m tower. Coulter (1990) has shown that two nights with similar mean meteorological conditions had very different turbulence characteristics in the NBL, which implies different rates of vertical mixing and diffusion. Most of the studies on NBL linked the structure and evolution of NBL with LLJ in different synoptic and terrain conditions (Blackadar 1957; Banta et al. 2002, 2003; Mahrt 1985, 1999; Mahrt and Vickers 2002; Karipot et al. 2006; Lothan et al. 2008; Abdou et al. 2010). Parker and Raman (1993) observed temperature inversions and wind maxima at two levels and a persistent elevated turbulent layer over complex terrain.

In India, studies on the boundary layer are either based on campaign data or with limited datasets. For example, several experiments were conducted in India, such as the Monsoon Trough Boundary Layer Experiment (MONTBLEX-90), Bay of Bengal Monsoon Experiment (BOBMEX), Land Surface Processes Experiment (LASPEX), and Arabian Sea Monsoon Experiment (ARMEX), to study the structure of the atmospheric boundary layer (ABL) over the Indian region. But most of these studies were confined to the monsoon trough region, Bay of Bengal, and Arabian Sea during the monsoon season. Moreover, most of them are intended to study the daytime boundary layer and its evolution. Only a few studies exist on NBL evolution and structure in the literature (Murthy et al. 1996; Devara et al. 1995). Murthy et al. (1996) investigated the variation of NBL height in a monsoon season, using Doppler sodar observations over Kharagpur, as a part of the MONTBLEX-90 experiment. As can be noticed, none of the above studies discussed different types of NBL and their seasonal variation. For the first time, long-term (2 years) Doppler sodar measurements made over complex hilly terrain are used to characterize the NBL structure in different seasons and also to study different types of NBL, that is, type 1 (strongly stable) and type 2 (weakly stable).

Instruments used in the present study and the data are described in section 2. In section 3, methodology for deriving the NBL height from sodar measurements and seasonal variation of NBL are discussed. It also examines the time–height variation of NBL structure in two types of NBL. The differences in the NBL structure between two types of NBL are discussed in light of variations in low-level jet and surface meteorological parameters (surface forcing). Results are summarized in section 4.

2. Site and data description

The observational site, Gadanki (13.5°N, 79.2°E), India, is located ~80 km away from the east coast of the Indian peninsula. It is situated 370 m above the mean sea level, surrounded by hillocks with varying heights of 200–300 m above ground level (AGL) in a radius of 1 km and 200–700 m AGL in a radius of 10 km (Fig. 1). Different types of ridges and hills are present around Gadanki with varied levels of vegetation. On the eastern side of Gadanki, the topography is very complex with hills varying from 200 to 1000 m. On the other hand, the increase in elevation is more gradual on the western side of Gadanki.

Fig. 1.

The topography around Gadanki. The star symbol indicates the location of the Doppler sodar.

Fig. 1.

The topography around Gadanki. The star symbol indicates the location of the Doppler sodar.

For the present study seasons are divided as follows: premonsoon (summer)—March–May, monsoon—June–September, postmonsoon—October–November, and winter—December–February. The seasonal wind pattern at the 950-hPa level is shown in Fig. 2 for all four seasons. The European Centre for Medium-Range Weather Forecasts (ECMWF)-Interim reanalysis data are used to generate these maps (http://data-portal.ecmwf.int/data/d/interim_daily/). It can be seen from Fig. 2 that the winds are strong (weak) and predominantly westerly (northeasterly) during the monsoon (winter and postmonsoon) season over the study region. The strongest winds are observed during monsoon season, followed by postmonsoon, winter, and premonsoon.

Fig. 2.

Spatial variation of seasonal mean wind at the 950-hPa level during winter, premonsoon, monsoon, and postmonsoon. ECMWF-Interim reanalysis data are used to generate this map. The gray shades indicate the wind speed. The arrows indicate the wind direction. Note that the length of the arrow is normalized to the maximum wind of that season.

Fig. 2.

Spatial variation of seasonal mean wind at the 950-hPa level during winter, premonsoon, monsoon, and postmonsoon. ECMWF-Interim reanalysis data are used to generate this map. The gray shades indicate the wind speed. The arrows indicate the wind direction. Note that the length of the arrow is normalized to the maximum wind of that season.

The Doppler sodar, located at Gadanki, operates at a frequency of 1800 Hz. The pulse length of the sodar used in the present study is 180 m s−1, that is, a range resolution of 30 m. The beam dwell time is ~9 s. Measurements were made in the altitude range of 30–1500 m in three directions (east, north, and vertical). The tilt angle of east and north beams is 16°. The radial wind velocities are computed first and then zonal, meridional, and vertical components of the wind are estimated. The Doppler sodar provides continuous wind information with high temporal resolution (27 s). The accuracy of sodar-derived horizontal wind speed is 0.1 m s−1 and wind direction is 3°. Technical specifications of the sodar system, signal processing, data quality control, and preliminary validation of the system are given in Anandan et al. (2008). In the present study, backscattered signal intensity, wind speed, and wind direction measured by the Doppler sodar during the observation period 2007–08 are used to study the structure and evolution of NBL and its seasonal variation. High-resolution wind measurements are averaged over 10 min to remove any instantaneous spurious value.

We have taken precautions in selecting clear and stable nights for our analysis; that is, we selected those nights that show a clear inversion layer structure on the sodar echogram. The presence of this layer indicates that the atmosphere is stable (Singal et al. 1984, 1986). The correctness of identified clear-sky and stable nights is cross-checked with two independent data records. The first one is visual observations made at the time of balloon launch in the evening [1700 local time (LT)]. Although the timings of sodar observations presented in this paper (mostly during night) and evening visual observations are different, they provide a clue of the atmospheric condition on that day. The other checking is the operation of lidar. A powerful lidar is being operated at the National Atmospheric Research Laboratory (NARL) on all clear-sky nights. The lidar, being an optical remote sensing instrument, operates only in clear-sky and optically thin cloudy conditions. It has been found that the identification of clear-sky nights from sodar echogram matches well with that of other methods discussed above.

The selected nights include a wide range of synoptic conditions covering all seasons of the year, subjected to clear-sky conditions. The sodar-derived winds are categorized as weak, moderate, and strong, following Banta et al. (2002); that is, strong winds: >15 m s−1, moderate winds: 5–15 m s−1 and weak winds: <5 m s−1.

An automatic weather station (AWS) is installed at NARL, Gadanki, to continuously monitor surface meteorological parameters such as temperature, pressure, relative humidity, wind speed, and direction. Data from AWS measurements during the observation period (2007–08) are used to study the variations of the above parameters at the surface. The AWS provides the above parameters at hourly resolution with good accuracy. Temperature and wind measurements obtained from a radiosonde (Meisei–RS01GH) are also used in the present study.

3. Results and discussion

a. Identification of the NBL and its characteristics

The depth of the NBL is one of its most important characteristic properties. The method of estimating the NBL depth using Doppler sodar measurements over uniform terrain is given by Stull (1988) and is verified later by several researchers (Beyrich and Weill 1993; Murthy et al. 1996). According to Stull (1988), the NBL top is the height, where the wind becomes maximum (nocturnal jet height) and a large gradient in temperature exists. We also followed the same procedure for identifying the NBL; that is, NBL is the height, where the wind is at a maximum in the height region of 60–800 m. As Gadanki is surrounded by hills in a complex fashion, it is essential to compare NBL height estimates obtained with the sodar data with those derived with radiosonde. At Gadanki, daily launch of radiosondes is, generally, made at ~1700 LT. Nighttime ascents were made only during special campaigns. Therefore, only a few (10) radio soundings were available for NBL height comparison.

A typical comparison of wind speed profiles obtained with sodar and GPS radiosonde for 2 days (2330 LT on 5 and 6 June 2008) along with radiosonde-derived temperature is shown in Fig. 3. Both sodar and radiosonde wind profiles show a maximum at an altitude of ~500 m, which marks the height of the NBL top. In general, the agreement between sodar- and radiosonde-derived wind profiles is good within the NBL (i.e., below ~500 m). The large difference in wind speed between the sodar and radiosonde at and above ~500 m could be related to the sampling of different environments by these two techniques. Sodar provides overhead measurements, while a balloon drifts away with the wind and therefore the measured wind may represents a different region. However, this difference is in general very small and can be neglected for practical purposes over plain terrain, but could be significant in complex terrain (like Gadanki). This is mainly because of greater small-scale wind variability over complex terrains associated with local circulations and mountain lee waves. The depth of LLJ is also different when obtained with the sodar and sonde. Radiosonde-derived low-level wind maximum does not show as pronounced a peak as sodar measurements at the NBL top; rather it shows a broad layer of thickness 300–500 m. The NBL can also be identified from temperature measurements (Stull 1988). A comparison of NBL height derived with the temperature technique has been made with that derived by wind technique. A well-developed temperature inversion is observed on 6 June 2008 at 500 m, the height where wind maximum is observed. On the other hand, such a clear temperature inversion is not seen on 5 June 2008. This shows that Doppler sodar is a very useful remote sensing instrument to detect the NBL height (using wind measurements alone).

Fig. 3.

Comparison of the wind speed measured by Doppler sodar and GPS radiosonde on two nights [2300 LT (top) 5 and (bottom) 6 Jun 2008]. The radiosonde-derived temperature profile is also included in the figure. The arrow indicates the height of the NBL identified by the Doppler sodar.

Fig. 3.

Comparison of the wind speed measured by Doppler sodar and GPS radiosonde on two nights [2300 LT (top) 5 and (bottom) 6 Jun 2008]. The radiosonde-derived temperature profile is also included in the figure. The arrow indicates the height of the NBL identified by the Doppler sodar.

To examine the seasonal variation of NBL, the NBL height is estimated from 10-min-averaged wind profiles for all clear-sky nights (220 nights). The NBL height varies considerably during the night in all seasons (e.g., Fig. 7). The variation, on some days, is large as 400–500 m. Therefore, the maximum heights of NBL reached on every night are averaged seasonally and these averages are considered as the seasonal mean NBL depths, respectively. Figure 4 shows the seasonal variation of mean NBL (along with standard deviation) during 2007–08. The NBL depth varies considerably between seasons. The mean NBL height is highest in the monsoon season (~490 m) and lowest in the winter (~370 m). Inspection of NBL height for individual nights reveals that the lowest (highest) NBL height observed is 300 (660) m and that it is in the winter (monsoon) season. Similar seasonal variation is observed in both years. The mean NBL is found to be low during winter because of the strong stability in that season.

Fig. 4.

Seasonal variation of the NBL obtained from sodar measurements during 2007–08. The error bars represent the standard deviation.

Fig. 4.

Seasonal variation of the NBL obtained from sodar measurements during 2007–08. The error bars represent the standard deviation.

b. Different types of NBL

As mentioned earlier, there are two types of NBL structures. The evolution and structure of these two types over complex terrain are discussed in this section. To examine how different these two types are and at what time of the night the differences are maximum, we divided the total night into two parts. The first half of the night corresponds to 1800–0000 LT, and the second half of the night corresponds to 0000–0600 LT. It will be shown later that the NBL top shows distinct characteristic features in second half of the night. The two types of NBL are identified based on wind structure. Type 1 is characterized by steady winds and nearly constant NBL height, whereas type 2 is characterized by unsteady winds and decreasing NBL height in the second half of the night. During type 2 nights, the NBL height decreases by >100 m in 3 h between 0200 and 0600 LT. The LLJ is present in both types of NBL, but its magnitude is different in different types. The two types of NBL are somewhat equivalent to the types discussed in Pichugina and Banta (2010). For example, type 1 is equivalent to the LLJ of traditional nose type, while type 2 is equivalent to flat strong LLJ type (see the above reference for definitions and examples for the nose and flat types of LLJ).

1) A case study for type 1 NBL

A typical example showing the structure and evolution of type 1 NBL is shown in Fig. 5. It shows the temporal variation of the wind speed and wind direction during 1–2 June 2008. We have chosen 1–2 June 2008 because it approximately represents the typical NBL structure over the land with clear-sky conditions and light winds from southwest. Wind speed is relatively high up to the height of 600 m in the late evening–early night (before 2000 LT). But later, winds became weak till midnight before they started increasing again. A weak LLJ is formed during 0100–0200 LT at ~300 m. Also, the strong wind speed region is confined to 200–400-m height region in the second half of night. Wind speeds of the order of 4–6 m s−1 are observed during this period. The wind direction shows intriguing variation during night and also with height. Large wind veering is observed with the altitude from southeasterly–easterly in the lower altitudes (<120–150 m) to westerly–northwesterly at higher altitudes (200 m). Throughout the night, the easterly wind component is observed at lower levels, instead of seasonal southwesterly winds. The seasonal (southwesterly–northwesterly) winds are observed at higher altitudes. This is a very stable night, with clear skies and light synoptic flow that favors the development of local circulation patterns (rotors) within the NBL. For example, the easterly component is mainly because of the valley winds from mountains (of height 200–300 m) located on the eastern side of Gadanki (see Fig. 1). Above the height of these mountains, the wind became westerly. When the wind is strong, westerly winds are observed (at ~1900 LT and during 1200–0200 LT) and when it is weak northwesterlies are prominently seen (2000–2100 LT and after 0500 LT) in the height region of 200–500 m.

Fig. 5.

Height–time variation of the wind (top) speed and (bottom) direction during 1–2 Jun 2008 showing the evolution of type 1 NBL.

Fig. 5.

Height–time variation of the wind (top) speed and (bottom) direction during 1–2 Jun 2008 showing the evolution of type 1 NBL.

2) A case study for type 2 NBL

Type 2 NBL is characterized by moderate and strong winds often with the presence of strong LLJs. Figure 6 shows temporal evolution of the wind speed and direction on 8–9 July 2008 depicting the evolution of type 2 NBL. It can be seen from Fig. 6 that the wind speeds are not steady during the night. Large wind speeds (with magnitudes of 6–10 m s−1) appear sporadically during the night, and the depth of these sporadic wind speed regions increases with time. The first sporadic region seen before 2130 LT is confined to the height region of 120–360 m. On the other hand, the last sporadic region, which appeared between 0200 and 0530 LT, is deep (seen between 150 and 900 m), and also the winds are moderate in this region. If we consider the night as consisting of two parts, then the mean wind speed in the first (second) part of the night is ~2 (3) m s−1 at 60 m and ~6 (8) m s−1 at 300 m. Large directional shear is observed between the surface and the top of NBL in the first half of night, while such a large directional shear is not observed during the second half. Just like type 1, the local circulation is seen in the first half with an easterly–southeasterly wind component at lower altitudes turning to southwesterly–northwesterly at higher altitudes. No significant variation in wind direction is seen with the altitude and also with time during the second half of the night. The moderate winds with greater depths and associated large shears generate turbulence. As a result the air is well mixed within the NBL and this mixing diffuses inversions (including near-surface inversions), if any, formed in the first half of the night.

Fig. 6.

As in Fig. 5, but for 8–9 Jul 2008, showing the evolution of the type 2 NBL.

Fig. 6.

As in Fig. 5, but for 8–9 Jul 2008, showing the evolution of the type 2 NBL.

As seen above, differences between the two types of NBL are seen mostly in the second half of the night. Differences are observed in several parameters, such as wind veering, intensification, and NBL variation. To examine these differences in a better way, wind profiles are averaged for 30 min and the time variation of these profiles are shown in Fig. 7 for both types of NBL (i.e., for 2 June 2008 and 9 July 2008). The NBL is identified from 30-min profiles using the wind maxima criterion, and all these points are joined by a solid line to show the variation of NBL during the second half of the night. On 2 June, the NBL height of ~270 m is observed at 0000 LT and increased to ~400 m in 1 h but then remained nearly at the same height (within a variation of 50 m) during the rest of the night. The observed steep increase of NBL is due to the intensification of LLJ. After the LLJ disappears, NBL height remains constant at an altitude of ~400 m for the rest of the night. In contrast, the NBL height increased gradually during 0000–0400 LT 9 July 2008, but decreased again in the early morning period (0400–0600 LT). The NBL height is deep and nearly steady during 0230–0400 LT at >400 m. The NBL height either increased or remained at the same height (before 0400 LT) in the presence of nocturnal LLJ. The weakening of LLJ in the early morning reduces not only the height of NBL but also strong shears and associated turbulence. The height and time of the wind maxima occurrence are consistent with earlier studies over West Africa made with sodar and wind profiler measurements (Abdou et al. 2010; Lothan et al. 2008).

Fig. 7.

Sodar-measured half-hourly averaged wind profiles showing the temporal variation of (top) type 1 and (bottom) type 2 NBL structures in the second half of the night.

Fig. 7.

Sodar-measured half-hourly averaged wind profiles showing the temporal variation of (top) type 1 and (bottom) type 2 NBL structures in the second half of the night.

The NBL height and its evolution depend on the radiative cooling of the surface. Therefore, it is logical to examine the temperature variation at the surface for these two types of NBL. The temperature measurements made by an AWS at 2-m altitude during 1–2 June 2008 and 8–9 July 2008 are used to estimate hourly variation in temperature (or temperature gradients). The variation of temperature gradients as function of time is plotted for the two days in Fig. 8. The NBL height estimated from wind speed maxima is also included in the figure for comparison with surface temperature variation. The temperature gradient is mostly positive (i.e., temperature is decreasing with time) for both cases during the first part of the night. But the temperature variation between the two types of NBL differs during the second half of the night. While it continues to decrease in the type 1 case, it remained nearly constant in the type 2 case. The temperature reduction in the night (between 1900 and 0600 LT) for the type 1 NBL case is nearly twice as large as for the type 2 case (it is 7.5°C for type 1 and 4°C for type 2). In type 1, the wind shear and associated turbulence is weak and the atmosphere within the NBL is very stable. It can be evidenced by the strong local circulation as well as the large temperature reduction at the surface. On the other hand, during the type 2 NBL case, strong winds and large wind shears are observed in the second half of the night. The strong turbulence generated by large wind shear transfers heat flux downward, which keeps the surface temperature nearly constant or at least will not allow the temperature reduction. Warm air advection in the early morning at 0300–0500 LT is observed for both types of NBL structures because of the LLJ activity. Short episodes of temperature increase are observed for both the types during this period may be due to the higher warmer air mixing downward.

Fig. 8.

Temporal variation of the temperature gradient corresponding to type 1 and type 2 NBL. Also plotted is the evolution of NBL during those two nights.

Fig. 8.

Temporal variation of the temperature gradient corresponding to type 1 and type 2 NBL. Also plotted is the evolution of NBL during those two nights.

3) A case study for the formation of type 1 and type 2 NBLs on consecutive nights

Type 1 and type 2 NBL structures can form on consecutive nights. Figure 9 shows a case study showing different types of NBL formation during 16–18 July 2007. Type 1 NBL is seen on 16–17 July 2007 with light to moderate winds and not much variation in the height of NBL during that period. The NBL depth is observed in the height region of 500–600 m. During this period (nights), moderate wind speeds of 6–8 m s−1 are observed at the top of the NBL. The NBL is directly transformed to mixed layer without any reduction in height. The type 2 NBL observed during 17–18 July 2007 night also shows moderate winds with varying magnitude. The NBL depth decreased drastically during early morning from ~570 m at 0400 LT to 400 m at 0600 LT.

Fig. 9.

Height–time variation of the wind speed during 16–18 Jul 2007, showing the evolution of different types of NBLs on consecutive nights (NBL height is indicated by the black thick line).

Fig. 9.

Height–time variation of the wind speed during 16–18 Jul 2007, showing the evolution of different types of NBLs on consecutive nights (NBL height is indicated by the black thick line).

The surface meteorological parameters during 16–18 July 2007 (Fig. 10) are examined to understand the linkages between the evolution of different types of NBL and surface meteorological parameters. The surface temperature shows gradual reduction during 16–17 July 2007 at night when type 1 NBL is present. On the other hand, as also seen in Fig. 10, the reduction in temperature during 17–18 July 2007 (representing type 2 NBL) is sharp in the first half of the night, but remained nearly constant in the second half. The surface wind speed is nearly zero for type 1 NBL, whereas 1–2 m s−1 is observed when type 2 NBL is present. Not only wind speed but also wind direction at the surface shows large variations during the night when type 2 NBL is present.

Fig. 10.

(top to bottom) Temporal variation of temperature at the 2-m level, wind speed and direction, and relative humidity during 16–18 Jul 2007 depicting the variation of surface meteorological parameters when type 1 and type 2 NBLs are present.

Fig. 10.

(top to bottom) Temporal variation of temperature at the 2-m level, wind speed and direction, and relative humidity during 16–18 Jul 2007 depicting the variation of surface meteorological parameters when type 1 and type 2 NBLs are present.

To study whether the increase in wind speed during 17–18 July 2007 (type 2) is a localized phenomenon or synoptic feature, the ECMWF-Interim wind at the 950-hPa level is examined. Figure 11 shows synoptic wind pattern at 2330 LT 16 July 2007, 0530 and 2330 LT 17 July 2007, and 0530 LT 18 July 2007. The bottommost panel shows the spatial variation of the wind speed difference between type 1 and type 2 nights (wind speed on type 2 night − wind speed on type 1 night) at two synoptic hours (i.e., at 2330 and 0530 LT). It is evident from the figure that there is not much change in the wind direction between the two nights, but the wind speed is higher during type 2 night at 0530 LT. Nevertheless, the wind speed is nearly same during both types of NBL at 2330. The above case studies clearly show that the wind speed and its temporal variation is nearly same for both types of NBL in the first half of the night, but they are different in the second half of the night. The following section examines this issue in a statistical sense; that is, using data from all clear-sky nights.

Fig. 11.

Synoptic wind pattern at the 950-hPa level at (top left) 2330 LT 16 Jul 2007, (top right) 0530 LT 17 Jul 2007, (middle left) 2330 LT 17 Jul 2007, and (middle right) 0530 LT 18 Jul 2007. (bottom) The spatial variation of wind speed difference between type 2 and type 1 nights. Positive (negative) wind speed difference indicates that the wind speed is larger (smaller) on type 2 nights than on type 1 nights.

Fig. 11.

Synoptic wind pattern at the 950-hPa level at (top left) 2330 LT 16 Jul 2007, (top right) 0530 LT 17 Jul 2007, (middle left) 2330 LT 17 Jul 2007, and (middle right) 0530 LT 18 Jul 2007. (bottom) The spatial variation of wind speed difference between type 2 and type 1 nights. Positive (negative) wind speed difference indicates that the wind speed is larger (smaller) on type 2 nights than on type 1 nights.

c. Statistics on different types of NBL occurrence

The total data are divided into type 1 and type 2 following their characteristics discussed in section 3b. Out of 220 nights, 134 nights show type 1 NBL characteristics and the remaining nights show type 2 characteristics. Table 1 lists number of nights that show type 1 and type 2 characteristics in different seasons. It is evident from Table 1 that, on average, the occurrence of type 1 NBL is more prevalent over Gadanki. The type 1 NBL shows pronounced seasonal variation with maximum occurrence percentage in the winter season (~82%) followed by the postmonsoon season (71%). The occurrence percentage of type 1 NBL is small in the monsoon season (40%), whereas type 2 NBL is more frequently observed in the monsoon season (60%) followed by summer (39%). From the above statistics, it is clear that type 1 NBL forms mostly during the cold season and type 2 NBL during the warm season. Furthermore, an earlier study on climatology of winds reveals that the low-level winds are stronger in the monsoon season than in any other season (Fig. 2). As discussed earlier, strong winds generate intense turbulence and in turn modulate the NBL structure similar to type 2.

Table 1.

Total number N of cases observed for each season during the observation period.

Total number N of cases observed for each season during the observation period.
Total number N of cases observed for each season during the observation period.

To examine the mean vertical structure of wind and the evolution of NBL during type 1 and type 2 NBL nights, composites of normalized wind speed and NBL height are constructed (Figs. 12 and 13, respectively). In Fig. 12, the wind speed is plotted with reference to NBL height, instead of height axis. First, the NBL height is estimated from each profile and the vertical axis is changed with reference to NBL height; that is, NBL height corresponds to 0 and the positive (negative) values on the y axis indicate how high (low) they are above (below) the NBL height. Since, the characteristics of two types of NBL are different only in the second half of the night, the data during the second half are only considered to construct the normalized composite wind profiles. As seen in Fig. 2, the normalized winds show clear seasonal variation on both types of NBL nights. The seasonal variation is also similar for both types of NBL with strong winds during monsoon and weak winds during winter. Nevertheless, wind magnitudes are different from type 1 to type 2 NBL. In all seasons, winds during type 2 nights are stronger than during type 1 nights. The vertical wind structure is also different in both types. As seen in case studies, type 1 shows a sharp wind maximum at NBL height while type 2 NBL shows a relatively broader wind maximum.

Fig. 12.

Vertical profiles of normalized (with reference to NBL height) seasonal wind speed for (top) type 1 and (bottom) type 2 cases, depicting the wind variation as a function of season and also NBL height.

Fig. 12.

Vertical profiles of normalized (with reference to NBL height) seasonal wind speed for (top) type 1 and (bottom) type 2 cases, depicting the wind variation as a function of season and also NBL height.

Fig. 13.

Temporal evolution of seasonal mean NBL during (top) type 1 and (bottom) type 2 nights for four seasons. The standard deviation is represented with error bars.

Fig. 13.

Temporal evolution of seasonal mean NBL during (top) type 1 and (bottom) type 2 nights for four seasons. The standard deviation is represented with error bars.

The evolution of the seasonal mean NBL height is different during type 1 and type 2 NBL nights (Fig. 13). The height of the NBL increases gradually in the first half of the night in a similar way during type 1 and type 2 nights, but the height variation is different in the second half for different types of NBL. The NBL height remained nearly constant during type 1 nights, but decreased gradually with time in the later part of second half (i.e., during 0300–0600 LT) during type 2 nights. The height variation is observed to be greater than 100 m in 3 h. A similar kind of variation is observed in all seasons. But the NBL height showed clear seasonal variation, similar to Fig. 4, at all hours during the night. The NBL is found to be high during monsoon and low during winter for both types of NBL. The height difference of NBL between monsoon and winter is found to be 100–200 m. On the other hand, not much difference is seen in the NBL height in premonsoon and postmonsoon (height differences are statistically insignificant), particularly during the second half of the night.

As seen above, the NBL height is mostly dictated by the radiative cooling at the surface and by the strength of LLJ and associated turbulence. To understand the role of surface forcing, meteorological parameters of all nights at the surface are considered. As seen in case studies, the difference between type 1 and type 2 NBL is observed in the second half of the night, where temperature is nearly constant and wind speed and direction show large variations when type 2 NBL is present. To study these features, the nights and corresponding surface meteorological data are grouped as type 1 and type 2. First, the hourly differences are computed for all meteorological parameters. These differences are plotted in the form of contour frequency by altitude diagram (CFAD). Figure 14 shows CFADs for hourly temperature difference, hourly wind difference, and hourly wind direction difference corresponding to type 1 and type 2 NBL cases. Also shown on each panel is the mean difference profile for each parameter superposed on CFAD. In addition, mean hourly differences for both type 1 and type 2 are given in one plot for easy comparison (bottom panel).

Fig. 14.

CFADs for (left) hourly temperature differences, (center) hourly wind speed differences, and (right) hourly wind direction differences corresponding to (top) type 1 and (middle) type 2 NBL cases. The overlaid black line with symbols is the mean hourly difference profile of the corresponding meteorological parameter. (bottom) The mean hourly difference profiles for type 1 and type 2 for easy comparison.

Fig. 14.

CFADs for (left) hourly temperature differences, (center) hourly wind speed differences, and (right) hourly wind direction differences corresponding to (top) type 1 and (middle) type 2 NBL cases. The overlaid black line with symbols is the mean hourly difference profile of the corresponding meteorological parameter. (bottom) The mean hourly difference profiles for type 1 and type 2 for easy comparison.

The mean (and also mode) of the distribution for temperature corresponding to type 1 NBL shows a gradual decrease in temperature with time, whereas the mean temperature difference curve for type 2 shows a sharp gradient in the early night (before 2000 LT) but later remained nearly constant with time. Also, considerable data for type 2 show no temperature difference between 2000 and 0400 LT. As discussed above, the downward heat flux by turbulence is mainly responsible for the constant temperature in the second half of the night. The magnitudes of temperature difference during type 2 NBL nights are smaller than those observed during type 1 NBL nights. The hourly differences in wind speed and wind direction are small for type 1 NBL, while they are relatively large for type 2 cases. The mean values of wind speed and wind direction differences for type 2 also show large variations from hour to hour. Because of the constant wind speed for type 1 case, cooler air is advected into the region. This suggests that type 1 formation is associated with cooler air advection. Most of the type 1 cases are observed in winter and postmonsoon season (section 3a). On the other hand, type 2 NBLs are mostly seen in premonsoon and monsoon, when surface temperature is relatively high. During these seasons, nocturnal LLJs are generally formed and they in turn generate strong shears and turbulence. Observed large hourly wind differences are mainly due to the formation of nocturnal LLJ, which intensifies in the second half of the night as seen in case studies. Earlier studies elsewhere also reported intensification of winds in the second half of the night (nocturnal LLJ) and presence of strong shears below the LLJ peak (Abdou et al. 2010).

4. Conclusions

For the first time, long-term Doppler sodar observations over complex terrain are used to study the time–height structure of the NBL, the seasonal variation of the NBL, and different types of NBL structures. We identify 220 clear-sky nights from a 2-year-long dataset comprising different seasons. Two distinctly different NBL structures are noticed in the second half of the night. Type 1 cases not only have light winds, but also low NBL height and low wind speed maxima. Type 2 cases have windy nights with high NBL heights and high wind speed maxima. Important results obtained from the present study are described in brief below.

  1. The NBL height exhibits a clear seasonal variation with deeper NBL (seasonal mean NBL height—490 m) in the monsoon season and shallower NBL (seasonal mean NBL height—370 m) in the winter season.

  2. Type 1 NBL is characterized by the steady and weak–moderate winds with nearly steady NBL top (not much variation in the second half of the night). The combination of weak winds and clear-sky conditions results in a strongly stable stratification regime. The type 2 NBL is identified as a weakly stable regime composed of unsteady and moderate to strong winds, strong wind shears, and associated mechanical turbulence. The NBL height also varies with time, particularly in the early morning. The nocturnal LLJ is present in both types, albeit with different magnitudes. Major differences in many parameters (e.g., wind, shear, turbulence, local circulation) between type 1 and type 2 are observed in the second half of the night.

  3. The surface temperature, wind speed, and wind direction exhibit distinctly different variations with time in type 1 and type 2 cases. The temperature in the type 1 case decreases gradually with time during the whole night, whereas it decreases in the first half of the night, but remained constant in the second half when type 2 NBL is present. The downward heat flux by turbulence is mainly responsible for the constant temperature in the second half of the night. Wind speed and wind shear variations at the surface with time are small for type 1 cases, whereas they are relatively large for type 2 cases.

  4. The majority of the nights show type 1 NBL structure over this complex topographical location. Type 1 and type 2 NBLs exhibit different seasonal cycles. While type 1 is more prevalent in the winter season, type 2 is frequently seen in the monsoon season.

Acknowledgments

We are very thankful to the ECMWF data center for providing the ERA-Interim dataset for the analysis (http://data-portal.ecmwf.int/data/d/interim_daily/) and to both anonymous reviewers for their constructive comments and suggestions, which improved the quality of the manuscript.

REFERENCES

REFERENCES
Abdou
,
K.
,
D. J.
Parker
,
B.
Brooks
,
N.
Kalthoff
, and
T.
Lebel
,
2010
:
The diurnal cycle of lower boundary-layer wind in the West African monsoon
.
Quart. J. Roy. Meteor. Soc.
,
136
,
66
76
.
Anandan
,
V. K.
,
M.
Shravan Kumar
, and
I.
Srinivasa Rao
,
2008
:
First results of experimental tests of newly developed NARL phased array Doppler sodar
.
J. Atmos. Oceanic Technol.
,
25
,
1778
1784
.
Banta
,
R. M.
,
R. K.
Newsom
,
J. K.
Lundquist
,
Y. L.
Pichugina
,
R. L.
Coulter
, and
L.
Mahrt
,
2002
:
Nocturnal low-level jet characteristics over Kansas during CASES-99
.
Bound.-Layer Meteor.
,
105
,
221
252
.
Banta
,
R. M.
,
Y. L.
Pichugina
, and
R. K.
Newsom
,
2003
:
Relationship between low-level jet properties and turbulent kinetic energy in the nocturnal stable boundary layer
.
J. Atmos. Sci.
,
60
,
2549
2555
.
Beyrich
,
F.
, and
A.
Weill
,
1993
:
Some aspects of determining the stable boundary layer depth from sodar data
.
Bound.-Layer Meteor.
,
63
,
97
116
.
Blackadar
,
A. K.
,
1957
:
Boundary layer wind maxima and their significance for the growth of nocturnal inversions
.
Bull. Amer. Meteor. Soc.
,
38
,
283
290
.
Caughey
,
S. J.
,
J. C.
Wyngaard
, and
J. C.
Kaimal
,
1979
:
Turbulence in the evolving stable boundary layer
.
J. Atmos. Sci.
,
36
,
1041
1052
.
Conangla
,
L.
,
J.
Cuxart
, and
M. R.
Soler
,
2008
:
Characterisation of the nocturnal boundary layer at a site in northern Spain
.
Bound.-Layer Meteor.
,
128
,
255
276
.
Coulter
,
R. L.
,
1990
:
A case study of turbulence in the stable nocturnal boundary layer
.
Bound.-Layer Meteor.
,
52
,
75
91
.
Devara
,
P. C. S.
,
R. P.
Ernest
,
B. S.
Murthy
,
G.
Pandithurai
,
S.
Sharma
, and
K. G.
Vernekar
,
1995
:
Intercomparison of nocturnal lower-atmospheric structure observed with lidar and sodar techniques at Pune, India
.
J. Appl. Meteor.
,
34
,
1375
1383
.
Karipot
,
A.
,
M. Y.
Leclerc
,
G.
Zhang
,
T.
Martin
,
G.
Starr
,
D.
Hollinger
,
J. H.
McCaughey
, and
G. R.
Hendrey
,
2006
:
Nocturnal CO2 exchange over a tall forest canopy associated with intermittent low-level jet activity
.
Theor. Appl. Climatol.
,
85
,
243
248
.
Kurzeja
,
R. J.
,
S.
Berman
, and
A. H.
Weber
,
1991
:
A climatological study of the nocturnal planetary boundary layer
.
Bound.-Layer Meteor.
,
54
,
105
128
.
Lothan
,
M.
,
F.
Said
,
F.
Lohou
, and
B.
Campistron
,
2008
:
Observation of the diurnal cycle in the low troposphere of West Africa
.
Mon. Wea. Rev.
,
136
,
3477
3500
.
Mahrt
,
L.
,
1985
:
Vertical structure and turbulence in the very stable boundary layer
.
J. Atmos. Sci.
,
42
,
2701
2711
.
Mahrt
,
L.
,
1999
:
Stratified atmospheric boundary layers
.
Bound.-Layer Meteor.
,
90
,
375
396
.
Mahrt
,
L.
, and
D.
Vickers
,
2002
:
Contrasting vertical structures of nocturnal boundary layers
.
Bound.-Layer Meteor.
,
105
,
351
363
.
Mahrt
,
L.
, and
D.
Vickers
,
2006
:
Extremely weak mixing in stable conditions
.
Bound.-Layer Meteor.
,
119
,
19
39
.
Mahrt
,
L.
,
J.
Sun
,
W.
Blumen
,
T.
Delany
, and
S.
Oncley
,
1998
:
Nocturnal boundary-layer regimes
.
Bound.-Layer Meteor.
,
88
,
255
278
.
Mahrt
,
L.
,
D.
Vickers
,
R.
Nakamura
,
M. R.
Soler
,
J. L.
Sun
,
S.
Burns
, and
D. H.
Lenschow
,
2001
:
Shallow drainage flows
.
Bound.-Layer Meteor.
,
101
,
243
260
.
Murthy
,
B. S.
,
T.
Dharmaraj
, and
K. G.
Vernekar
,
1996
:
Sodar observations of the nocturnal boundary layer at Kharagpur, India
.
Bound.-Layer Meteor.
,
81
,
201
209
.
Parker
,
M. J.
, and
S.
Raman
,
1993
:
A case study of the nocturnal boundary layer over a complex terrain
.
Bound.-Layer Meteor.
,
66
,
303
324
.
Pichugina
,
Y. L.
, and
R. M.
Banta
,
2010
:
Stable boundary layer depth from high-resolution measurements of the mean wind profile
.
J. Appl. Meteor. Climatol.
,
49
,
20
35
.
Singal
,
S. P.
,
B. S.
Gera
, and
S. K.
Aggarwal
,
1984
:
Nowcasting by acoustic remote sensing: Experiences with the system established at the National Physical Laboratory, New Delhi
.
J. Sci. Ind. Res.
,
43
,
469
.
Singal
,
S. P.
,
B. S.
Gera
,
D. R.
Pahwa
, and
S. K.
Aggarwal
,
1986
:
Studies of surface-based shear echo structures
.
Atmos. Res.
,
20
,
125
131
.
Soler
,
M. R.
,
C.
Infante
,
P.
Buenestado
, and
L.
Mahrt
,
2002
:
Observations of nocturnal drainage flow in a shallow gully
.
Bound.-Layer Meteor.
,
105
,
253
273
.
Stull
,
R. B.
,
1988
:
An Introduction to Boundary Layer Meteorology. Kluwer Academic Publishers, 666 pp
.
Wyngaard
,
J. C.
,
1986
:
Measurement physics. Probing the Atmospheric Boundary Layer, D. H. Lenschow, Ed., Amer. Meteor. Soc., 5–18
.