Asia-Pacific J. Atmos. Sci.

, 49(5), 587-601, 2013

DOI:10.1007/s13143-013-0052-8

Assessment of Turbulent Kinetic Energy Budget and Boundary Layer

Characteristics during Pre-Monsoon Thunderstorm Season over Ranchi
Bhishma Tyagi1 and A. N. V. Satyanarayana2
1
Department of Meteorology, Stockholm University, Stockholm, Sweden
2
Centre for Oceans, Rivers, Atmosphere and Land Sciences, Indian Institute of Technology Kharagpur, West Bengal, India

(Manuscript received 3 August 2012; revised 27 February 2013; accepted 28 March 2013)
© The Korean Meteorological Society and Springer 2013

Abstract: In the present work turbulent kinetic energy (TKE) budget the flux–gradient relationship in the surface layer to the
and boundary layer characteristics are studied for an Indian tropical stability parameter, z/L, over homogeneous surfaces, where z is
station Ranchi (23o25N, 85o26E), situated over Chota Nagpur plateau. the height above the ground scale and L is the Obukhov length
The pre-monsoon months (March-May) data for years 2008-2010
(Ha et al., 2007). The relations for turbulent fluctuations are
has been used in the present study which is the period of severe
thunderstorm over the North East India. TKE budget terms, dissipa- site and season specific. Keeping in mind the importance of
tion rates, and normalized standard deviations of wind and turbulent fluctuations in surface layer, various researches
temperature along with skewness of temperature have been analyzed proposed relations between parameters of moisture and heat
to find out characteristic difference between days of thunderstorm transport (Wyngaard and Cote, 1971; McBean, 1971; Panofsky
and days of clear weather. Present study brought out significant et al., 1977; Wilczeck, 1984; Kader and Yaglom, 1990; Roth,
variations in the turbulence transportation between the days of 1993; Kaimal and Finnigan, 1994; Hogstrom and Bergstrom,
thunderstorm activity to that of fair weather days. Site and season
1996; Andreas et al., 1998; Bian et al., 2003; Li et al., 2006).
specific relationships normalized standard deviations of wind and
temperature with atmospheric stability during pre-monsoon thunder- A measure of turbulence in the atmosphere is turbulent
storm and non thunderstorm days over Ranchi are proposed. One of kinetic energy (TKE) and is directly related to the transport of
the important outcomes of the study is proposing site specific momentum, heat, and moisture through the boundary layer
relationships between TKE dissipation rates with respect to atmo- (Zhang and Park, 1999). TKE budget has comprised various
spheric stability. These results are useful in validating the mesoscale terms which are related to physical processes i.e., associates the
simulations of thunderstorm activity. local storage of turbulence to the shear production, buoyancy
production, dissipation, and the transport processes, and having
Key words: Turbulent kinetic energy, dissipation rates, standard
deviations of wind and temperature numerous applications in both empirical and computational
modeling in boundary-layer meteorology (Han et al., 2000;
Holzäpfel and Robins 2004; Frech, 2007). TKE has been
1. Introduction studied during various field campaigns based on MOST
(Businger et al., 1971; Wyngaard and Coté 1971; McBean and
The exchange of momentum, heat and moisture between
Miyake, 1972; Pino et al., 2003; Li et al., 2008). Bister and
Earth’s surface and atmosphere happens in atmospheric surface
Emanuel (1998), and Businger and Businger (2001) reported
layer (ASL) where the frictional drag force of the Earth and
that the viscous dissipation of TKE in the surface layer of
turbulent forces are predominant (Hegde et al., 2010). Turbu-
storms is important to storm evolution. Dimensionless dissi-
lent processes are essential for atmosphere-surface interactions
pation rate of the TKE is one of the important non-dimensional
and transportations of these parameters in the surface layer
forms to emerge in atmospheric boundary layer (Kaimal and
(Filho et al., 2008). Correct representations of turbulent
Finnigan, 1994; Yahaya et al., 2003). Dimensionless TKE
fluctuations in this layer is highly important for understanding
dissipation rates are being analysed by researchers over the
of the local climatic variation and proper modeling of turbulent
years in various parts of the world to quantify the TKE dissi-
exchange processes. Usually this representation has been ex-
pation (e.g., Zhang and Park, 1999; Hartogenesis and De Bruin,
pressed by relating the non-dimensional variances of velocity
2005; Frech, 2007). Droegemeier et al. (1987) investigated the
and temperature with some relevant similarity parameters of
dynamics of thunderstorm outflow using a two dimensional
ASL pertinent to the local area (Singha and Sadr, 2012).
numerical model and reported the importance of turbulent
Monin–Obukhov similarity theory (MOST) is used to relate
mixing. Bowen (1995) has observed reduced horizontal and
vertical turbulence due to thunderstorm outflow. Weinbeck
(1997) found that validity of averaging and other statistical
Corresponding Author: A. N. V. Satyanarayana, Centre for Oceans,
Rivers, Atmosphere and Land Sciences Indian Institute of Tech- tests are limited during thunderstorm winds and lateral and
nology Kharagpur, Kharagpur-721 302, West Bengal, India. longitudinal turbulence intensities did not increase with
E-mail: anvsatya@coral.iitkgp.ernet.in increasing wind speed, but appeared to be a function of con-
588 ASIA-PACIFIC JOURNAL OF ATMOSPHERIC SCIENCES

vective activity. as part of the Department of Science and Technology, New

Such kinds of field experiments to study atmospheric bound- Delhi, sponsored project “Observational study of land surface
ary layer turbulence are rare in Indian subcontinent and the atmosphere interaction in the monsoon trough along its active
deficiency of data over tropical region has been a severe con- eastern end”. The site is having scanty dry patchy grass near
straint on the study of turbulence structure. Turbulence charac- the instrumentation area, and having clear fetch (> 1 Km)
teristics in the atmospheric surface layer are studied by few towards North West to South directions. During NTD wind is
researchers, e.g., Sivaramakrishnan et al. (1992), Yadav et al. predominantly from North to North Westerly, while on TD
(1996), Krishnan and Kunhikrishnan (2002), Ramana et al. winds are mainly from South-South Westerly quadrant, with
(2004), Dharamaraj et al. (2009). TKE budget terms variations some influence from North Westerly and South Easterly as
during southwest monsoon period has been studied at two well. The location of tower, details of employed sensors and
tropical Indian stations, namely Varanasi and Kharagpur, and site map can be obtained from Tyagi et al. (2012).
variation of the budget parameters with stability has shown Data sets of pre-monsoon months of 2008, 2009, and 2010
differences between active and break phases of monsoon at Ranchi have been used for the present study. We have used
(Viswanadham et al., 1997a, 1997b; Srivastava and Parth slow response data (SRD) of 1 Hz as well as fast response data
Sarthi, 2002). (FRD) of 10 Hz (sonic data) of atmospheric variables in the
However, detailed studies of TKE budget variations and present study for thunderstorm days (TD) and non thunder-
turbulence characteristics in relation to thunderstorm events storm days (NTD). Criterion to define a TD is based on
are absent over NE Indian region, where the pre-monsoon occurrence of thunderstorm event at any time of that particular
(March-May) thunderstorms occur frequently. The severe thun- day (Rodriguez et al., 2010). The chosen TD and NTD cases
derstorms during pre-monsoon season in eastern and northe- are listed in Table 1 along with timing of thunderstorm event,
astern part of India have significant socio-economy impact as associated rainfall with event, and total rainfall in that day.
72% of tornadoes are associated with these thunderstorms Data sets have been subjected to quality check before going in
which cause enormous losses of lives and damage of property. to analysis. The data sets, failing to meet any of these quality
The hails associated with these severe thunderstorms cause checks are not considered in to further analysis.
huge damage to crops; in particular to mango production and The method of Vickers and Mahrt (1997) has been followed
summer paddy production. Though, few attempts have been for the spike removal. It computes the mean and standard
made to understand the turbulence mechanism during the deviation for a series of moving window, which moves one
period of these thunderstorm but most of these are in form of point at a time through the series. Any individual value falling
case studies and are limited due to lack of observations. The within ± 3 standard deviation from the mean is retained and
improvement in prediction of these thunderstorms is also han- the rest are replaced by mean ± 3 standard deviations (Vis-
dicapped due to lack of mesoscale observations and insuffi- wanadham et al., 1997). Double rotation (DR) scheme and tilt
cient understanding (STORM Science Plan, 2005). Present angle corrections are incorporated in data (Wilcjak et al.,
study focuses on understanding of intensity of turbulence (both 2001). Linear detrending has been done to remove any possible
mechanically and thermally induced); and contribution of trends present in the data. The purpose of detrending is to
various TKE budget parameters culminating in the total kinetic separate the active turbulent transport (eddy flux) from slower,
energy, and surface layer characteristics which plays an im- deterministic atmospheric motions and instrument drift (Mon-
portant role in the energy exchange from atmospheric surface crieff et al., 1998). For linear detrending, the line of best fit
layer in to higher layers under different atmospheric stability over the period has been found and subtract instead of
conditions over Ranchi (23o25N, 85o26E). The study site is subtracting the mean from the signal in a period. For example,
having its importance for pre-monsoon thunderstorm as some in covariance calculations, the fluctuations are obtained by
of the thunderstorms are being generated over Chota Nagpur subtracting a signal from a realization mean x from an
plateau and move towards North West subsequently (STORM instantaneous mean Xt, i.e., x′t = xt – Xt
science plan, 2005). While for linear detrending, the mean is calculated by linear
regression line Xt = St + I over the period T. Where S is
2. Site description, data and quality checks regression slope and I is intercept (Rannik and Vesala, 1999).
The data sets during rain hour are rejected because of
The study site for present study is Ranchi, India. The soil is possible effects of water droplets on transducer heads of sonic
of ultisol type, having sandy loam texture which is a mixture anemometers significantly influencing the data quality (Thomas
of sand (60%), silt (8.7%), and clay (31.3%) (Gupta and and Foken, 2007). Spectrum of velocity, temperature and
Gajbhiye, 2002). A 32 m tower is established at main campus humidity fluctuations in the atmosphere decay with respect to
of Birla Institute of Technology, Ranchi, with slow response the frequency with the Kolmogorov slope of −5/3 in the
sensors (1 Hz) for measurement of Air temperature, humidity, inertial sub-range. The data have been subjected to spectral
wind speed and direction at six levels (1, 2, 4, 8, 16 and 32 m analysis to check the applicability of data for use in flux
respectively), and fast response measurement of wind and tem- computations. Only those data sets which coincide with the
perature (10 Hz) at 10 m height by using a sonic anemometer universal laws of surface layer have been considered (Kolmo-
30 November 2013 Bhishma Tyagi and A. N. V. Satyanarayana 589

Table 1. Information about TD and NTD cases and associated rainfall minute interval). The covariance of w and x for each small
at Ranchi during pre-monsoon season of 2008-2010. interval (M = 1 to 6) has been computed, and the average value
Non-Thunder- has been noted. This value will be compared with the
Thunderstorm Days
storm Days covariance determined for the whole interval. The time series
S. No. Local Time of Rainfall during in steady state if the difference between both covariances is
Date Thunderstorm Thunderstorm Date less than 30 % (Foken and Wichra, 1996).
Event (h) (mm)
1 3 April 2008 1420-1503 1.1 7 April 2008 3. Methodology
2 13 April 2008 1408-1511 0.1 8 April 2008
3 15 April 2008 1448-1532 0.1 11 April 2008 Assuming a steady state and horizontally homogeneous con-
4 4 May 2008 1209-1408 0.3 21 April 2008 ditions and choosing a coordinate system aligned to mean
5 18 May 2008 1408-1624 1.8 25 April 2008 wind, the turbulent kinetic energy (TKE) budget equation can
6 19 May 2008 1512-1612, 6.9, 5.4 28 April 2008 be written as (e.g., Wyngaard and Cote, 1971; McBean and
2015-2100 Elliott, 1975),
7 20 May 2008 1833-1903 7.4 1 May 2008
∂ ( w′e ) 1 ( w′p′ )
---- ( w′θv ′) –w′u′ ∂u
------ – ---------------- –--- ∂-----------------
g
8 21 May 2008 2020-2138 8.2 14 May 2008 - –ε = 0 (1)
9 22 May 2008 1403-1527 41.9 15 May 2008 θv ∂z ∂z ρ ∂z
10 25 May 2008 1427-1551 No Data 16 May 2008 where u', v', w' are zonal, meridional and vertical compo-
11 31 May 2008 1411-1459 No Data 17 May 2008 nents of wind fluctuations; u is the mean wind speed; p′, θv ′
12 11 May 2009 1236-1442 4.8 27 May 2008 are turbulent parts of pressure and virtual potential tempera-
13 12 May 2009 1353-1551 23.1 1 April 2009 ture; g is acceleration due to gravity, z is log mean height, ρ
14 15 May 2009 1314-1709 0.7 2 April 2009 and θv are mean density and mean virtual potential tem-
15 17 May 2009 1408-1549 0.3 13 April 2009 perature, e is turbulent kinetic energy per unit mass and ε is
16 18 May 2009 1357-1437 No Data 14 April 2009 dissipation, w′θv ′, u′w′, w′e and w′p′ are vertical kinematic
17 20 May 2009 1333-1438 6.8 15 April 2009 eddy fluxes of heat, momentum, energy and pressure, respec-
18 28 May 2009 1644-1800 9.8 18 April 2009 tively. From the left hand side of Equation (1), first term is
19 26 April 2010 1401-1509 No Data 27 April 2009 buoyant production term which acts as a source (sink) term
20 4 May 2010 1601-1724 1.4 8 May 2009 during day (night) time. The second term is shear production
21 5 May 2010 1554-1652 0.2 10 May 2009 term, which is a source term always. Third term is divergence
22 15 May 2010 1434-1554, No Data 19 May 2009 of vertical TKE flux, which can be a source or sink term
1743-1855 depending on whether there is a flux convergence or diver-
23 17 May 2010 1551-1721 No Data 3 April 2010 gence. Fourth term is a pressure correlation term that describes
24 20 May 2010 1512-1636 No Data 4 April 2010 the redistribution of TKE by pressure perturbations. The fifth
25 26 May 2010 1526-1629 0.2 5 April 2010 and last term is viscous dissipation of TKE, which is always a
26 - - - 11 April 2010 sink term. e has been computed as:
27 - - - 12 April 2010
28 - - - 19 April 2010 2 2 2
e = 0.5 ( u′ + v′ + w′ ) (2)
29 - - - 25 April 2010
30 - - - 11 May 2010 During the experiment, the fast response turbulence mea-
31 - - - 12 May 2010 surements using sonic anemometer are available at 10 m
height only, making it not possible to estimate the flux diver-
gorov, 1941; Kaimal and Finnigan, 1994). During convective gence following the methodology suggested by Wyngaard and
conditions, there arise a question of existence of steady state Cote (1971). We make estimates of the magnitudes of flux
and applicability of MOST. The steady state test proposed by divergence following McBean and Elliott (1975), by calcu-
Foken and Wichra (1996) has been employed to ensure the lating it as a function of friction velocity (u*), Monin-Obhukhov
applicability of MOST. For the present study, the data sets Length (L) and slope (m) between w′e ⁄ u 3* and stability
failing to meet steady state test are rejected before the analysis. parameter z/L,
The test compares the statistical parameters determined for the
averaging period and for short intervals within the data period. i.e., Flux Divergence = u 3* m ⁄ L (3)
For example, the time series for the determination of the
covariance of the measured signals w (vertical wind) and x McBean and Elliot (1975) suggest that Monin-Obukhov
(horizontal wind component or scalar) of about 30 minutes similarity theory may not apply to these fluxes.
duration will be divided into M = 6 intervals of about 5 Dissipation values are calculated from the high-frequency
minutes. N is the number of measuring points of the short end of spectra obtained by available fast response turbulence
interval (N = 3000 for 10 Hz scanning frequency and a 5 measurements of sonic anemometer. The following equations
590 ASIA-PACIFIC JOURNAL OF ATMOSPHERIC SCIENCES

are used in calculating dissipation (e.g., Kaimal et al., 1972). An attempt has been made to study the relation between the
skewness of temperature and ζ during TD and NTD. Skewness
3⁄2 3⁄2
ε = [ nSu, v, w( n ) ] ( 2πn ⁄ U )A (4) is a measure of the asymmetry of the probability distribution of
a real-valued random variable. The skewness of a sample
where n is natural frequency (s−1), S is spectral density, u, v, variable ai can be represented as
w are zonal, meridional and vertical velocity components, U is N
1- ( a – a )3
1- ---
σa3 N ∑
mean wind speed and A is a universal constant for the inertial Sa = ---- i (9)
sub-range (chosen 0.55 for u and 0.73 for v and w, following i=1

Roth et al. (2006)). Pressure transport term of the budget N

1- a
N i∑
equation of Equation (1) is estimated as a residual, which is where a = --- i (10)
likely to contain any possible errors in all the remaining terms =1

of TKE budget equation. N

1- ( a – a )2
N i∑
The Non-dimensional form of dissipation rate (φε) of TKE is and σa2 = --- i (11)
assumed to follow MOST and therefore can be expressed as a =1

function of the stability parameter, z/L (Yahaya et al., 2003), 2

a and σ in general referred to as the mean and variance of
i.e.,
the variable ai.
kzε ⎛ z ⎞ The averaging period for mean and fluxes of data sets has to
φ ε = ------- = f⎝ ---⎠ (5)
u*3 L be chosen for these calculations, and choices of this time
period varies in literature with a most common value of 30
where k is von Karman constant, u* is frictional velocity (m
minutes. Vickers and Mahrt (2003) has proposed that this time
s−1). There have been discussions about value of von Karman
period may vary from 30 minutes if all the non turbulent
constant (e.g., Hogstrom, 1996; Frenzen and Vogel, 2001;
motions have been removed from the data set by removing the
Foken, 2006; Andreas et al., 2006; Trinh, 2012). It is found for
cospectral gap, and this removable time period typically
the present study that Karman’s constant is confined to a
depends on the particular data set. In the present study, we have
narrow range around the canonical value of 0.4, and hence for
not attempted the multiresolution analysis method to determine
the present study we have taken the value 0.4.
the cospectral gap and averaging period for mean and fluxes of
According to MOST, the non-dimensionalised standard de-
data sets used in the present study is taken as 30 minutes.
viation of u, v and w components of wind normalized by
friction velocity (u*) are universal functions of the stability
parameter (z/L) in horizontally homogeneous and steady flow
4. Results
(Stull, 1988). Friction velocity is a fundamental surface layer
a. Variation of mean TKE and budget terms
velocity scaling parameter and is given by
--1-
2 2 4 Mean TKE ( e ) has been computed using Equation (2) for
u* = ( ( u′w′ ) + ( v′w′ ) ) (6)
all TD and NTD cases. It is noticed that diurnal variation of e
where the over bar and the prime indicate time average and
deviation from the mean value, respectively. The scaling
parameter of temperature (T*) and length scaling are given by

T* = w′T′ ⁄ u* (7)

and
3
L = –u* ⁄ [ k( g ⁄ T ) ( w′T′) (8)

where u', v', w' and T' indicates the fluctuation in wind com-
ponents and temperature obtained by subtracting the mean
value from the instantaneous value of each parameter. T* is
friction temperature in Kelvin, u*, frictional velocity, k, von
Karman constant, g is acceleration due to gravity and w is
vertical velocity, respectively. Similarity functions of surface
layer turbulence such as σu / u*, σv / u*, σw / u*, and σT / T* as
measured by sonic anemometer system has been computed
and relations with respect to z/L (here after referred as ζ) are Fig. 1. Mean TKE Variation for 13 April 2008 (TD) and 25 April
derived separately for TD and NTD. 2008 (NTD) at Ranchi.
30 November 2013 Bhishma Tyagi and A. N. V. Satyanarayana 591

is different during TD and NTD. Diurnal variation of mean 1800 to 2330 LST) averages of all TKE budget terms during
TKE at Ranchi during 13 April 2008 (TD) and 25 April 2008 each of the TD and NTD cases as well as respective cumulative
(NTD) are shown in Fig. 1 Mean TKE values are high in TD averages are presented in Tables 2 and 3.
to that of NTD. This pattern is observed in all TD and NTD On 13 April 2008, thunderstorm event happened at site
cases at Ranchi. Gap in the curve for 13 April 2008 is due to during 1408-1511 LST. The variation and significance of each
data removal by the quality checks. Even in night time, value term of Equation (1) has been explained for the present study.
of TKE remains higher on TD, which shows presence of
highly turbulent atmosphere on TD compare to NTD. Similar (1) Shear term
kind of variation has been noticed with higher values of TKE More shear production was observed on TD from early
during individual TD cases to that of NTD. morning to afternoon 1300 LST, up to the thunderstorm
The contribution of different terms of TKE budget during all initiation (Fig. 2a). It is seen from the wind speed values that
cases of TD as well as NTD has been analyzed at the site. until 0600 h the winds were in range of 2 to 4.5 m s−1 and then
Diurnal variation of TKE budget parameters on one TD case picked up to the order of 4.5 to 8.5 m s−1, and 15.95 m s−1
(13 April 2008) and one NTD case (3 April 2008) at Ranchi during the thunderstorm occurrence. The wind speed comes
are depicted in Fig. 2, and general variations observed on all down after thunderstorm event. This would be attributed to the
other TD and NTD cases are also discussed. The daily, daytime higher values of shear production from 0600 LST onwards.
(0600 to 1800 hrs LST) and night time (0000 to 0600 LST and Table 2 and Table 3 depicts that daytime shear production is

Fig. 2. Diurnal variation of TKE Budget parameters at Ranchi for TD (13 April 2008), and NTD (07 April 2008) for (a) Shear production, (b)
Buoyancy, (c) Divergence, (d) Dissipation, and (e) Pressure transport term.
592 ASIA-PACIFIC JOURNAL OF ATMOSPHERIC SCIENCES

Table 2. Turbulent kinetic energy budget terms during non-thunderstorm days (NTD) at Ranchi. Number of half hour runs is given in brackets.
∂ ( w′e )
–w′u′∂u –--- ∂-----------------
g ( w′θ ′) 1 ( w′p′)
------ ---- v – ---------------- −ε -
∂z θv ∂z (× 10 m2s−3)
−5 ρ ∂z
Date
(× 10−5 m2s−3) (× 10−5 m2s−3) (× 10−5 m2s−3) (× 10−5 m2s−3)
Daily Day Night Daily Day Night Daily Day Night Daily Day Night Daily Day Night
7 April 312.6 413.7 211.5 117.8 243.4 −7.8 −67.2 −108.7 −25.6 −900.0 −1502.5 −297.4 536.8 954.2 312.6
08 (27) (11) (16) (27) (11) (16) (27) (11) (16) (27) (11) (16) (27) (11) (16)
25 April 811.8 1510.7 112.8 162.9 363.3 −37.6 −87.9 −198.9 23.1 −287.2 −527.8 −46.6 −599.5 −1147.4 811.8
08 (25) (17) (8) (25) (17) (8) (25) (17) (8) (25) (17) (8) (25) (17) (8)
13 April 311.4 603.0 19.8 188.3 382.6 −6.0 −171.0 −348.2 6.2 −674.4 −1321.9 −26.9 345.6 684.4 6.9
09 (14) (10) (4) (14) (10) (4) (14) (10) (4) (14) (10) (4) (14) (10) (4)
19 May 68.5 102.1 34.9 140.7 283.9 −2.5 −158.6 −320.1 2.9 −480.7 −595.2 −366.2 430.1 529.2 330.9
09 (10) (6) (4) (10) (6) (4) (10) (6) (4) (10) (6) (4) (10) (6) (4)
3 April 85.8 110.6 61.0 113.9 245.2 −17.4 −80.6 −173.4 12.1 −274.0 −417.5 −130.6 155.0 235.1 74.8
10 (23) (11) (12) (23) (11) (12) (23) (11) (12) (23) (11) (12) (23) (11) (12)
4 April 135.6 229.4 41.9 118.6 258.4 −21.1 −51.7 −112.5 9.1 −467.6 −765.0 −170.1 265.0 389.8 140.2
10 (22) (10) (12) (22) (10) (12) (22) (10) (12) (22) (10) (12) (22) (10) (12)
12 April 149.3 279.0 19.6 245.7 501.0 −9.5 −44.6 −91.0 1.7 −483.0 −919.7 −46.3 132.6 230.6 34.6
10 (20) (12) (8) (20) (12) (8) (20) (12) (8) (20) (12) (8) (20) (12) (8)
11 May 73.5 113.8 33.2 128.9 265.5 −7.8 −119.4 −246.3 7.4 −232.8 −422.8 −42.8 149.9 289.8 9.9
10 (21) (13) (8) (21) (13) (8) (21) (13) (8) (21) (13) (8) (21) (13) (8)
Average 243.6 420.3 66.8 152.1 317.9 −13.7 −97.6 −199.9 4.6 −475.0 −809.1 −140.9 176.9 270.7 215.2

Table 3. Turbulent kinetic energy budget terms during thunderstorm days (TD) at Ranchi. Number of half hour runs is given in brackets.
∂ ( w′e )
–w′u′∂u –--- ∂-----------------
g ( w′θ ′ ) 1 ( w′p′ )
------ ---- v – ---------------- −ε -
∂z θv ∂z (× 10 m2s−3)
−5 ρ ∂z
Date
(× 10−5 m2s−3) (× 10−5 m2s−3) (× 10−5 m2s−3) (× 10−5 m2s−3)
Daily Day Night Daily Day Night Daily Day Night Daily Day Night Daily Day Night
3 April 921.6 376.9 1466.4 129.5 290.1 −31.1 −143.6 −322.5 35.4 −1314.2 −809.7 −1818.6 406.6 465.3 347.9
08 (21) (11) (10) (21) (11) (10) (21) (11) (10) (21) (11) (10) (21) (11) (10)
13 April 512.3 701.9 322.5 199.9 438.3 −38.5 −160.9 −352.8 30.9 −723.6 −997.1 −450.3 172.4 209.5 135.3
08 (30) (15) (15) (15) (15) (15) (15) (15) (15) (15) (15) (15) (15) (15) (15)
18 May 545.9 1028.5 63.2 154.1 324.1 −15.9 −120.1 −252.8 12.7 −1100.0 −1938.6 −261.5 520.1 838.7 201.5
08 (26) (15) (11) (26) (15) (11) (26) (15) (11) (26) (15) (11) (26) (15) (11)
12 May 527.4 933.1 121.7 149.7 301.6 −2.1 −227.9 −459.3 3.5 −1586.8 −2256.5 −917.2 1137.6 1481.1 794.1
09 (13) (6) (7) (13) (6) (7) (13) (6) (7) (13) (6) (7) (13) (6) (7)
20 May 403.7 737.3 70.2 46.2 85.7 6.6 −24.6 −45.7 −3.6 −877.7 −1318.6 −436.8 452.4 541.2 363.7
09 (12) (6) (6) (12) (6) (6) (12) (6) (6) (12) (6) (6) (12) (6) (6)
4 May 698.9 674.1 723.9 161.5 337.8 −14.7 −194.5 −406.5 17.5 −1128.1 −1011.8 −1244.4 462.1 406.7 517.5
10 (36) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18) (18)
15 May 762.9 882.8 643.1 173.9 390.4 −42.4 −184.5 −413.5 44.5 −1183.4 −1348.4 −1018.5 430.9 488.6 373.3
10 (32) (20) (12) (32) (20) (12) (32) (20) (12) (32) (20) (12) (32) (20) (12)
17 May 938.5 1453.6 423.4 183.3 330.4 36.5 −81.4 −146.9 −15.6 −1981.2 −2523.1 −1439.2 940.6 886.2 995.1
10 (30) (15) (15) (15) (15) (15) (15) (15) (15) (15) (15) (15) (15) (15) (15)
20 May 434.7 604.1 265.3 174.3 366.2 −17.6 −304.9 −640.4 30.5 −765.7 −1040.5 −491.1 461.7 710.6 212.8
10 (41) (21) (20) (41) (21) (20) (41) (21) (20) (41) (21) (20) (41) (21) (20)
Average 638.4 821.4 455.5 152.5 318.3 −13.2 −160.3 −337.8 17.3 −1184.5 −1471.6 −897.5 553.8 669.8 437.9

more than the nighttime. As expected, the shear term is the shown in Fig. 2b. Day time buoyancy term is positive (Table 2
main source term of the TKE budget. and Table 3), which shows active thermal convection up in to
atmosphere, whereas negative night time buoyancy values can
(2) Buoyancy term be explained on the basis of static stability present in atmo-
In the early morning hours up to 0900 LST, buoyancy term sphere, which tends to suppress or consume TKE. The daily
is noticed of same order for both TD and NTD. After 0900 average value reveals higher buoyancy generation during TD
LST, higher magnitude was noticed on TD to that of NTD, as to that of NTD. Positive contribution of buoyancy generation
30 November 2013 Bhishma Tyagi and A. N. V. Satyanarayana 593

is seen on both contrasting days but more during TD. (6) Year wise variation of Budget terms
The analysis of all the terms of the TKE budget equation
(3) Flux divergence term reveals that shear term is source during both daytime as well as
Table 2 and Table 3 elucidates that flux divergence term is nighttime. Higher values of nighttime shear production terms
negative during daytime and positive during nighttime. Mag- are observed in 2008 compared to 2009 and 2010 (Table 2, and
nitude of flux divergence term is higher on TD in comparison Table 3). Buoyancy term is acting as source during daytime
to NTD for both daytime as well as nighttime (Fig. 2c). This and as sink during nighttime for all the days of 2008, but
means more upward transportation of TKE in the TD cases to acting as source for one TD case each in both daytime as well
that of NTD. During these days, buoyancy term is also having as nighttime in 2009 and 2010. Flux divergence term is sink
small nighttime value. It is noticed that the flux divergence and during daytime and source during nighttime, except for the 7
buoyancy term varies oppositely during daytime and nighttime April 2008 (NTD), 20 May 2009 (TD), and 17 May 2009 (TD)
during both TD as well as NTD cases. The diurnal variability where it was sink in night time as well. Dissipation is a sink
during the TD is seen according to the intensity of thunder- term during both daytime and nighttime. Higher dissipation is
storm activity as in the case of shear production. The flux observed during NTD of 2008 compare to 2009 and 2010.
divergence term is a sink during daytime and a source during Pressure transportation terms are source during both daytime
nighttime. Under stable conditions, classic MOST (Wyngaard and nighttime during both TD and NTD. Higher shear pro-
and Cote, 1971) states this term is negligible (= 0). However duction, and buoyancy generation is noticed during TD than
for present study this term is a source during nighttime which that of NTD. The variation of shear production and flux
contradicts classic MOST. McBean and Elliot (1975) observed divergence terms are in strong relation with the intensity of the
that the flux divergence term of vertical components showed a thunderstorm activity. Magnitude of buoyancy term is higher
consistent increase with instability while horizontal compo- than flux divergence term with opposite sign during TD and
nents have no effects on the TKE budget. They further suggest NTD. Dissipation term is compensating all other terms of TKE
that MOST may not apply to these fluxes. For present study as budgeting for the present study; while maximum time’s pre-
well, the fluxes during stable conditions are deviating from ssure transport term is a source term.
MOST.
b. Kinetic energy dissipation rates (φε)
(4) Dissipation term
Higher dissipation is noticed on TD up to occurrence of It is suggested by some researchers (Businger et al., 1971;
thunderstorm event (Fig. 2d), but after the event, dissipation Wyngaard and Cote, 1971; Kaimal and Finnigan, 1994) that in
values came down on TD and comparatively higher for NTD the nearly neutral conditions mechanical production term
during same time. Dissipation acts opposite to shear produc- equals the dissipation one, i.e., φε is equal to 1. However, some
tion as seen in the other studies elsewhere (Wyngaard and researchers found that φε is not always equal to 1 in nearly
Cote, 1971), and it remains always a sink term. Higher shear neutral conditions, and it may be > 1 or < 1 (e.g., Hogstrom,
production noticed during different timings in TD and NTD 1990; Hogstrom, 1992; Frenzen and Vogel, 1992; Albertson et
leads to more generation of TKE, and this higher TKE pro- al., 1997; Frenzen and Vogel, 2001; Li et al., 2008). Different
duction in turn leads to higher dissipation. In both TD and φε relations with stability for different sites has been proposed
NTD, the magnitude of average dissipation term is almost 6 by various researchers e.g., Stull (1988), Thiermann and Grassl
times to that of buoyancy term. Dissipation of TKE can be a (1992) (hereafter referred as TG), Kaimal and Finnigan (1994),
significant source of heat in storms, and it increases the Edson and Fairall (1998) (hereafter referred as EF), Pahlow et
efficiency of the storm (Businger and Businger, 2001). al. (2001), Yahaya et al. (2003). It was proposed by Yahaya et
al. (2003) that during stable conditions (night time), when the
(5) Pressure transport term ground is colder than the air φε can adopt the z-less stratifi-
Pressure transport term has been estimated as a residual in cation behavior. For the present study normalized dissipation
Equation (1) by considering steady state and horizontally rates of vertical wind {φε (w)} with respect to ζ, for stable and
homogeneous conditions. Thus it is difficult to interpret result unstable conditions during TD and NTD are estimated (based
of this term as it may possibly contain the errors of all the on Equation 4). Bin quartiles are computed for equal log
other terms of the budget equation. Pressure transport term spaced bins from data points to fit the empirical relationships
with higher magnitude is seen during entire day of TD than on in both TD and NTD. It has been found that earlier proposed
NTD. It is observed that pressure transport term is positive relationships for unstable or stable conditions are not satisfying
during daytime as well as nighttime for both TD and NTD on present study at both the sites, and hence there is a need for
chosen case (Fig. 2e), as well as in other days (Table 2, Table new relations in present study for TD and NTD. Relations are
3) except 25 April 2008 (TD) case, where it is negative. different for TD and NTD at the site, which is showing
Pressure transport term is greater than flux divergence term as different dissipation during TD to that of NTD cases. Inter-
found by McBean and Miyake (1972). estingly for the present study in near neutral conditions for
NTD φε is equal to 1, while it is < 1 for TD. Relations of
594 ASIA-PACIFIC JOURNAL OF ATMOSPHERIC SCIENCES

Fig. 3. Dimensionless dissipation rate of TKE as a function of ζ at Ranchi for the unstable condition (Fig. 3a), and stable condition (Fig. 3b).

dissipation rates are as follows: rates can be attributed to the higher dissipation during unstable
During NTD: conditions of TD.

φε(ζ) = (1 − 27.5ζ)−1 − 0.82ζ, ζ≤0 (12) c. Variation of budget parameters with stability

φε(ζ) = 1 + 0.8ζ, 0 ≤ ζ ≤ 10 (13) Based on ζ, the period of the study is divided into three
stability groups: near neutral (−0.09 < ζ < 0.09), unstable (ζ <
During TD: −0.09) and stable (0.09 < ζ) during TD as well as NTD cases.
Table 4 provides the information of the stability ranges chosen,
φε(ζ) = (0.55 − 19.5ζ)−0.57 − 0.5ζ, ζ≤0 (14) TKE budget parameters during TD and NTD separated
individually into different stability groups; and their averages
φε(ζ) = 0.52 + 1.3ζ, 0 ≤ ζ ≤ 10 (15) computed under each stability group. All the TD and NTD
cases (of 2008, 2009 and 2010) have been taken in to account
These empirical relations, along with relations of EF and by dividing them in to stability groups (Table 4).
TG, φε (w) with respect to ζ, Bin quartile points for unstable
and stable case during TD and NTD are shown in Fig. 3. The (1) Unstable Case
dissipation rates of TD are lower than that of NTD in the Mean stability is showing that NTD cases are more unstable
unstable region of −ζ > 3, but higher than that of NTD for −ζ (−0.621) compare to TD cases (−0.392). Shear production,
< 3. These results are showing different behavior of two buoyancy, and pressure transport terms acts as source terms,
dissipation rates in the free convection sub layer (−ζ > 2) as while flux divergence and dissipation acts as sink terms in both
proposed by Albertson et al. (1997). However, in the stable TD and NTD. Flux divergence and dissipation together com-
region, the dissipation rates for NTD are higher than TD. The pensates the other terms of the budget equation. Magnitude of
effect of highly convective atmosphere and higher production terms in TD is higher than that of NTD during unstable con-

Table 4. Variation of TKE Budget parameters at Ranchi during thunderstorm (TD) and non thunderstorm days (NTD) with atmospheric stability.
∂ ( w′e )
–w′u′∂u –--- ∂-----------------
g ( w′θ ′) 1 ( w′p′)
------ ---- v – ---------------- −ε
-
Stability Mean Stability ∂z θv ∂z ρ ∂z
(× 10−5 m2s−3)
(z/L) (× 10−5 m2s−3) (× 10−5 m2s−3) (× 10−5 m2s−3) (× 10−5 m2s−3)
TD NTD TD NTD TD NTD TD NTD TD NTD TD NTD
Unstable
−0.392 −0.621 608.75 409.76 442.68 310.13 −481.82 −199.43 −1150.70 −801.45 581.10 280.99
(< −0.09)
Near neutral
0.001 −0.001 1065.19 789.63 12.05 100.51 0.78 −44.79 −1837.34 −430.66 759.31 −414.69
(−0.09 − +0.09)
Stable
0.813 1.154 115.68 62.13 −26.67 −24.92 25.85 14.38 −188.52 −108.01 73.69 56.42
(> +0.09)
30 November 2013 Bhishma Tyagi and A. N. V. Satyanarayana 595

σ
ditions. More buoyancy production during TD to that of NTD -----u = 2.3( 1 – 3ζ )1 ⁄ 3 −10 < ζ ≤ 0 (16)
can be explainable based on availability of sufficient moisture u*
σv
in the lower layers of the atmosphere and rapid increase of ----- = 2.18 ( 1 – 5ζ )1 ⁄ 3 −10 < ζ ≤ 0 (17)
convection in terms of sensible heat flux. Higher wind speeds u*
σw
during day time in TD can be attributed for higher values of ------ = 1.3( 1 – 1.8ζ )1 ⁄ 3 −10 < ζ ≤ 0 (18)
shear production. Higher transportation of energy up in to the u*
atmosphere has been observed during TD to that of NTD by For TD during stable stratification at Ranchi:
flux divergence. Interpretation of pressure transport term is
σ
difficult as it is estimated as residual which contains errors of -----u = 2.3( 1 + 3ζ )1 ⁄ 3 0 < ζ ≤ 10 (19)
the other terms. It is observed that during unstable conditions, u*
σv
shear production, buoyancy, and dissipation term values for ----- = 2.18 ( 1 + 2.5ζ )1 ⁄ 3 0 < ζ ≤ 10 (20)
TD are higher than that of NTD, which shows presence of u*
σw
highly convective atmosphere present during TD cases. ------ = 1.3( 1 + 0.8ζ )1 ⁄ 3 0 < ζ ≤ 10 (21)
u*
(2) Near Neutral Case For NTD during unstable stratification at Ranchi:
Overcast sky and strong winds results in more shear pro-
σ
duction in TD cases to that of NTD. Mean stability is having -----u = 2.25( 1 – 2ζ )1 ⁄ 3 −10 < ζ ≤ 0 (22)
positive value for TD (0.001), and negative for NTD (−0.001). u*
σv
Shear production and buoyancy terms are source terms for ----- = 2.1( 1 – 2ζ )1 ⁄ 3 −10 < ζ ≤ 0 (23)
both TD and NTD, and dissipation is a sink for both TD and u*
σw
NTD. But flux divergence and pressure transport terms are ------ = 1.42 ( 1 – ζ)1 ⁄ 3 −10 < ζ ≤ 0 (24)
source terms for TD and sink terms for NTD. Interestingly, the u*
shear production is higher in TD to that of NTD, but buoyancy For NTD during stable stratification at Ranchi:
and flux divergence are showing higher values during NTD.
σ
-----u = 2.25( 1 + 3ζ)1 ⁄ 3 0 < ζ ≤ 10 (25)
(3) Stable Case u*
σv
For stable region also, mean stability values are higher during ----- = 2.1( 1 + 1.15ζ )1 ⁄ 3 0 < ζ ≤ 10 (26)
NTD (1.154) to that of TD (0.813). All terms of TKE budget u*
σw
are having higher magnitudes in TD to that of NTD. Buoyancy ------ = 1.42( 1 + 0.4ζ )1 ⁄ 3 0 < ζ ≤ 10 (27)
is almost same for both TD and NTD (with −26.67 for TD and u*
−24.92 for NTD), and is negative (acts as sink term). In stable Unstable region is less scatter for NTD and points are
case, buoyancy and dissipation together compensates all other following a definite range, which is widely scatter in stable
terms. region, while in TD, both unstable and stable regions are
having widely scattered values. In the higher instability region
d. The non-dimensionalised standard deviation of wind com- relations for σu ⁄ u* (Fig. 4) are significantly different for TD
ponents and NTD (with higher values for TD), but converging to same
during near neutral conditions. During stable stratification, TD
The existing relations for wind components in literature are and NTD are parallel to each other keeping TD on higher side
found not suitable for present study. The reason for this may be relations. For σv ⁄ u* (Fig. 5) also new relations have been pro-
the difference in data sets which for present study are based on posed for the present study. The values of σw ⁄ u* in unstable
pre-monsoon thunderstorm activity season, while earlier studies region shows similar behavior as that of σu ⁄ u* (Fig. 6a). It is
were not focused on the thunderstorm season or similar atmo- seen that σw ⁄ u* increases with increase in instability and obeys
spheric conditions. Hence new relations have been developed. 1/3 power law which is in general agreement with the published
Relations close to the present study are also plotted along with literature (Panofsky et al., 1977; Kedar and Yaglom, 1990;
the new developed relations. These relationships are being Sivaramakrishnan et al., 1992; Krishnan and Kunikrishnan,
fitted by considering 50% bin quartile values for NTD and TD. 2002; Bian et al., 2003). For the stable region, values of σw ⁄ u*
Horizontal fluctuations of winds are following 1/3 power law, are higher for NTD to that of TD, and again TD values are high
in both unstable and stable stratification irrespective of TD and for ζ > 1.
NTD. Plots of standard deviation of horizontal, meridional,
and vertical wind components, along with the 50% bin e. The non-dimensionalised standard deviation of temperature
quartiles, and fitted new relations for TD and NTD at Ranchi and skewness of temperature
has been shown in Figs. 4-8. The relations for TD and NTD
are found to be different in both unstable and stable stratifi- The relationship of σT ⁄ T* vs. ζ have been plotted for NTD
cation which are as follows: and TD shown in Fig. 7. Present study fit has been done by
For TD during unstable stratification at Ranchi: considering 50% bin quartile values for NTD and TD along
596 ASIA-PACIFIC JOURNAL OF ATMOSPHERIC SCIENCES

Fig. 4. Normalized standard deviation of horizontal velocity as a function of ζ at Ranchi for (a) unstable stratification, (b) stable stratification.

Fig. 5. Normalized standard deviation of meridional velocity as a function of ζ at Ranchi for (a) unstable stratification, (b) stable stratification.

Fig. 6. Normalized standard deviation of vertical velocity as a function of ζ at Ranchi for (a) unstable stratification, (b) stable stratification.
30 November 2013 Bhishma Tyagi and A. N. V. Satyanarayana 597

Fig. 7. Normalized standard deviation of temperature as a function of ζ at Ranchi for (a) unstable stratification, (b) stable stratification.

Fig. 8. Skewness of temperature as a function of ζ at Ranchi for (a) unstable stratification, (b) stable stratification.

σT
with noted literature relationships. The σT ⁄ T* relation is ----- = 3.2( 1 + 0.2ζ )–1 0 < ζ≤ 10 (31)
T*
following −1/3 power law in unstable condition, and in stable
case, the relationship is varying with σT ⁄ T* decrease with It can be noticed from Fig. 7, that the relations for TD and
increasing ζ (as follows by Bian et al., 2003). The relation NTD are varying in line to each other which can be justified
between ζ and σT ⁄ T* in unstable and stable stratification for with small differences in buoyancy term at the site as well.
NTD and TD are: However, in near neutral range of unstable stratification, the
For TD at Ranchi: temperature deviations are more dispersed in TD and the
relations are significantly different. The σT ⁄ T* value for near
σT
----- = ( ζ )–1 ⁄ 3 −10 < ζ≤ 0 (28) neutral conditions is found to be higher than σu, v, w ⁄ u* . This
T*
kind of higher values for σT ⁄ T* has been reported earlier by
σT
----- = 3.5 ( 1 + 0.47ζ )–1 0 < ζ≤ 10 (29) Dharamraj et al. (2009), and Krishna and Kunhikrishnan (2002)
T*
as well. The reason to such increase in normalized standard
For NTD at Ranchi: deviation of temperature can be explained by considering the
relation of temperature scale to friction velocity and heat flux.
σT
----- = 2.25 ( 1 – 7.25ζ )–1 ⁄ 3 −10 < ζ≤ 0 (30) As we approach towards neutral range, the value of heat flux
T*
598 ASIA-PACIFIC JOURNAL OF ATMOSPHERIC SCIENCES

decreases which makes higher values of σT ⁄ T* (Dharamraj et

al., 2009).
New relations for skewness of temperature (Skew T) with ζ
during TD and NTD have been shown in Fig. 8. Relations for
TD and NTD are as follows:
TD Cases at Ranchi:

SkewT = 1.18 + 0.25ln(−ζ) for −10 < ζ ≤ 0 (32)

SkewT = −0.12 for 0 ≤ ζ < 1 (33)

NTD Cases at Ranchi:

SkewT = 1.1 + 0.25ln(−ζ) for −10 < ζ ≤ 0 (34)

SkewT = −0.18 for 0 ≤ ζ < 1 (35)

These relations are different from the earlier proposed rela-

tions by Tillman (1972), Businger et al. (1971), Antonia et al.
(1983), Andreas et al. (1998) and Ohtaki (1985), but close to
Andreas et al. (1998). Skew-T is constant in the highly stable
regime during both TD (0.12) and NTD (0.18). In the unstable Fig. 9. Non-Dimensional wind shear as a function of ζ at Ranchi
regime of TD the constant value of 0.82 (shown as dotted line) during TD and NTD.
is noticed where as the value is 1.2 during NTD situation.
Table 5 is showing the average values of all the turbulence and relationship of φM vs. ζ have been plotted for NTD and TD
parameters for different stability regions. Mean values of ζ are shown in Fig. 9. The relations proposed by Businger et al.
found to be higher in magnitude for unstable regimes of TD, (1971) and Dyer (1974) have also been plotted for the present
and for stable regimes of NTD. Values of all the turbulence data. It is found that both TD and NTD follow the Businger-
parameters are found to be higher in TD to that of NTD in all Dyer curve in unstable region. However, in the stable region
three stabilities. These values are close to the range of reported the nature of curve is similar to Businger-Dyer curve but with
results in the near neutral conditions e.g., Andreas et al. a different slope. Thus for stable region a new fit has been
(1998), Bian et al. (2003), Dharamraj et al. (2009), and obeys done by considering 50% bin quartile values for NTD and TD
MOST. For TD both unstable and stable values are widely along with Businger and Dyer relations.
scattered compared to NTD; which is showing anisotropic For stable conditions after z/L > 1, the wind shear starts to
behavior of turbulence during TD. It is noticed that TD and level off and the φM relation deviates from Businger-Dyer re-
NTD values are different to each other, with higher TD values lations during stable conditions which follow the non station-
during normalized standard deviation for horizontal velocity arity of the prevailing conditions. Such differences of φM vs. ζ
and temperature. relationships from simplified expression of Businger-Dyer are
also reported by Hicks (1976), Howell and Sun (1999), and
f. Dimensionless wind shear Hartogensis and DeBruin (2005). It is found that common
assumption of height independent flux required by Monin-
The dimensionless wind shear (φM ) has also been analysed Obukhov similarity theory may not be applicable at night

Table 5. Average values of surface layer turbulence parameters for unstable, neutral, and stable stratification during TD and NTD cases at Ranchi
for the present study.
Number of σu σv σ σT
Stability ζ ----- ----- -----w- ----- Skew-T
Points u* u* u* u*
unstable (< −0.09) 220 −0.479 3.064 3.178 1.499 −1.621 0.844
neutral
−0.007 −0.336
TD

141 2.691 2.349 1.390 0.264

(−0.09 < z/L < 0.09)
stable (> 0.09) 127 0.456 3.249 3.013 1.464 3.268 0.079
unstable (< −0.09) 257 −0.463 2.861 2.899 1.464 −1.654 0.718
NTD

neutral
82 −0.022 2.512 2.464 1.382 −0.173 0.017
(−0.09 < z/L < 0.09)
stable (> 0.09) 141 0.967 3.077 2.700 1.525 3.300 0.112
30 November 2013 Bhishma Tyagi and A. N. V. Satyanarayana 599

(Howell and Sun, 1999). The possible reason for such behavior Acknowledgments. Authors would like to thank Dr N.C.
was thought to be site-specific considerations such as advec- Mahanti and Manoj Kumar, BIT Mesra, for providing the tower
tion (Hicks, 1976). data obtained as a part of Department of Science & Technology,
The relation between ζ and φM in unstable and stable Govt. of India, New Delhi for providing funding the research
stratification for NTD and TD are: project to conduct “Observational study of land surface atmo-
For TD and NTD at Ranchi: sphere interaction in the monsoon trough along its active
eastern end”. We express our thanks to Mr. D. Pradhan, DDGM,
φM = (1 − 15ζ)−1/4 −10 < ζ≤ 0 (36) India Meteorological Department, Kolkata, for providing DWR
information about thunderstorms. Authors are thankful to an-
φM = 1 + 1.2ζ 0 < ζ≤ 10 (37) onymous reviewers who thoroughly check the manuscript and
constructively suggested in improvement of present manuscript.
5. Summary
Edited by: Jimy Dudhia
In the present study, high values of mean TKE during TD
have been found, which are showing highly turbulent atmo- REFERENCES
sphere on TD compare to NTD. TKE budget terms are
revealing the fact of high turbulence generation and dissipation Albertson J. D., M. B. Parlange, G. Kiely, and W. E. Eichinger, 1997: The
on TD compare to NTD. Highly unstable atmosphere was average dissipation rate of turbulent kinetic energy in the neutral and
noticed during TD while during night time, stability is higher unstable atmospheric surface layer. J. Geophys. Res., 102, 13423-
13432, doi:10.1029/96JD03346.
on NTD. Higher shear production, and higher dissipation with Andreas E. L., and Coauthors, 2006: Evaluation of the von Kármán
high values of buoyancy and flux divergence are noticed prior constant in the atmospheric surface layer. J. Fluid Mech., 559, 117-149.
to thunderstorm event in TD, which are becoming smaller to ______, R. J. Hill, J. R. Gose, D. I. Moore, W. D. Otto, and A. D. Sarma,
that of NTD after the event. These results are supporting the 1998: Statistics of surface layer turbulence over terrain with metre-scale
presence of favorable conditions for the occurrence of thunder- heterogeneity. Bound.-Layer Meteor., 86, 379-408.
storm during TD. Antonia R. A., S. Rajagopalan, and A. J. Chambers, 1983: Conditional
It is observed that dissipation rates are higher on NTD in sampling of turbulence in the atmospheric surface layer. J. Appl.
Meteor. Climatol., 22, 69-78.
highly unstable and stable regions, but during the near neutral Bian L., X. Xu, L. Lu, Z. Gao, M. Zhou, and H. Liu, 2003: Analysis of
conditions dissipation rates are higher during TD. Variation of turbulence parameters in the Near-Surface Layer at Qamdo of the
TKE budget terms with respect to stability has shown that the Southeastern Tibetan Pleateau. Adv. Atmos. Sci., 20(3), 369-378.
values are higher for TD to that of NTD for all terms in Bister M., and K. A. Emanuel, 1998: Dissipative heating and hurricane
unstable and neutral region. However in stable region buoyancy intensity. Meteor. Atmos. Phys., 65, 233-240.
values of NTD are higher to TD in Ranchi. Bowen B. M., 1995: Example of reduced turbulence during thunder-
Turbulent fluctuations of wind and temperature were storm outflow. J. Appl. Meteorol., 35, 1028-1032.
Businger J. A., J. C. Wyngaard, Y. Izumi, and E. F. Bradley, 1971: Flux
measured in the atmospheric surface layer and statistics of the
profile relationships in the atmospheric surface layer. J. Atmos. Sci., 28,
fluctuations both in the time and frequency domains were 181-189.
investigated from the point of view of the MOST. New site Businger S., and J. A. Businger, 2001: Viscous dissipation of turbulence
specific relations for TD and NTD are proposed. Wind kinetic energy in storms. J. Atmos. Sci., 58, 3793-3796.
components are following 1/3 power law, while temperature is Dharamaraj T., G. R. Chintalu, and P. E. Raj, 2009: Turbulence charac-
following −1/3 power law in unstable region, and −1 power teristics in the atmospheric surface layer during summer monsoon of
law for the stable region. 1997 over a semi-arid location in India. Meteor. Atmos. Phys., 104, 113-
123.
The values are more scattered in unstable region of TD Droegemeier K. K., and R. B. Wilhelmson, 1987: Numerical simulation of
compare to NTD, which may be effect of mesoscale convec- thunderstorm outflow dynamics. Part I: Outflow sensitivity experi-
tive activity i.e., thunderstorms at the site which are affecting ments and turbulence dynamics. J. Atmos. Sci., 44, 1180-1210.
the turbulent fluctuations, and hence surface layer charac- Dyer A. J., 1974: A review of flux-profile relationships. Bound.-Layer
teristics during TD. Relations for σu, v, w ⁄ u* at Ranchi are Meteor., 7, 363-372.
showing higher values for TD than that of NTD in unstable Edson J. B., and C. W. Fairall, 1998: Similarity relationships in the marine
stratification. For stable region, values are higher in TD for atmospheric surface layer for terms in the TKE and scalar variance
budgets. J. Atmos. Sci., 55, 2311-2328.
σu ⁄ u* and σv ⁄ u* but less than NTD for σw ⁄ u* . Convection is
Filho E. P. M., L. D. A. Sa´, H. A. Karam, R. C. S. Alvala´, A. Souza, and
mainly forced and not free as there is no asymptotic free M. M. R. Pereira, 2008: Atmospheric surface layer characteristics of
convection regime present. The temperature deviations are turbulence above the Pantanal wetland regarding the similarity theory.
more dispersed in TD case rather than NTD for unstable Agric. Forest Meteor., 148, 883-892
stratification, while the range of points is equally distributed Foken T., 2006: 50 Years of the Monin-Obukhov similarity theory. Bound.-
for both TD and NTD in stable stratification. Average values Layer Meteor., 119, 431-447.
of temperature deviations are always higher magnitude wise ______, and B. Wichra, 1996: Tools for quality assessment of surface-
based flux measurements. Agric. Forest Meteor., 78, 83-105.
during TD than that of NTD. Frech M., 2007: Estimating the turbulent energy dissipation rate in an
600 ASIA-PACIFIC JOURNAL OF ATMOSPHERIC SCIENCES

airport environment. Bound.- Layer Meteor., 123, 385-393. 129, 115-136.

Frenzen P., and C. A. Vogel, 1992: The turbulent kinetic energy budget in McBean G. A., 1971: The variations of the statistics of wind, temperature
the atmospheric surface layer: a review and an experimental and humidity fluctuations with stability. Bound.-Layer Meteor., 1, 438-
reexamination in the field. Bound.-Layer Meteor., 60, 49-76. 457.
______, and ______, 2001: Further studies of atmospheric turbulence in ______, and J. A. Elliott, 1975: The vertical transports of kinetic energy by
layers near the surface: scaling the TKE budget above the roughness turbulence and pressure in the boundary layer. J. Atmos. Sci., 32, 753-
sublayer. Bound.-Layer Meteor., 99, 173-206. 766.
Gupta S., and V. T. Gajbhiye, 2002: Effect of concentration, moisture, and ______, and M. Miyake, 1972: Turbulent transfer mechanism in the
soil type on the dissipation of flufenacet from soil. Chemosphere, 47, atmospheric surface layer. Quart. J. Roy. Meteor. Soc., 98, 383-398.
901-906. Moncrieff J., R. Clement, J. Finnigan, and T. Meyers, 1998: Averaging,
Ha K. J., Y. K. Hyun, H. M. Oh, K. E. Kim, and L. Mahrt, 2012: Evaluation detrending, and filtering of eddy covariance time series. In: Lee X,
of boundary layer similarity theory for stable conditions in CASES-99. Massman W, Law B. (eds.) Handbook of Micrometeorology, Kluwer
Mon. Wea. Rev., 135, 3474-3483, doi: 10.1175/MWR3488.1. Academic Publishers, 7-31.
Han J. g., S. P. Arya, S. Shan, and Y. L. Lin, 2000: An estimation of Ohtaki E., 1985: On the similarity in atmospheric fluctuations of carbon
turbulent kinetic energy and energy dissipation rate based on dioxide, water vapour and temperature over vegetated fields. Bound.-
atmospheric boundary layer similarity theory, a report submitted to Layer Meteor, 32, 25-37.
national aeronautics and space administration langley research center Pahlow M, M. B. Parlange, and F. Port´e-Agel, 2001: On the Monin-
prepared for langley research center hampton, Virginia 23681-2199 Obukhov similarity in the stable atmospheric boundary layer. Bound.-
NASA/CR-2000-210298, 25 pp. Layer Meteor., 99, 225-248.
Hartogensis O. K., and H. A. R. DeBruin, 2005: Monin-Obukhov similarity Panofsky H. A., D. H. Tennekes, and J. C. Wyngaard, 1977: The
functions of the structure parameter of temperature and turbulent kinetic characteristics of turbulent velocity components in the surface layer
energy dissipation rate in the stable boundary layer. Bound.- Layer under convective conditions. Bound.-Layer Meteor., 11, 355-361.
Meteor., 116, 253-276. Pino D., J. Vila`-guerau de arellano, and P. G. Duynkerke, 2003: The
Hegde A. K., R. Venkateshan, C. V. Srinivas, and K. M. Balakrishna, 2010: contribution of shear to the evolution of a convective boundary layer. J.
Estimation of atmospheric surface layer parameters and numerical Atmos. Sci., 60, 1913-1926.
simulation using MM5 at Mangalore, West Coast of India. Meteor. Rodriguez C. A. M., Rosmeri P. da Rocha, and R. Bombardi, 2010: On the
Atmos. Phys., 107,161-172 development of summer thunderstorms in the city of Sao Paulo: Mean
Hicks B. B., 1976: Wind profile relationships from the ‘Wangara’ meteorological characteristics and pollution effect. Atmos. Res., 96,
experiment. Quart. J. Roy. Meteor. Soc., 102, 535-551. 477-488.
Högström U., 1990: Analysis of turbulence structures in the surface layer Ramana M. V., P. Krishnan, and P. K. Kunhikrishnan, 2004: Surface
with a modified similarity formulation for near neutral conditions. J. boundary layer characteristics over a tropical inland station: seasonal
Atmos. Sci., 47, 1949-1972. features. Bound.-Layer Meteor., 111, 153-175.
______, 1992: Further evidence of 'inactive' turbulence in the near neutral Rannik U., and T. Vesala, 1999: Autoregressive filtering versus linear
atmospheric surface layer. In: 10th symp. on turbul. and diffusion, detrending in estimation of fluxes by the eddy covariance method.
Portland, OR, Ameran Meteorology Society, Boston, MA, 188-191 pp. Bound.-Layer Meteor., 91, 259-280.
______, 1996: Review of some basic characteristics of the atmospheric Roth M., 1993: turbulence transfer relationships over an urban surface II:
surface layer. Bound.- Layer Meteor., 78, 215-246. Integral statistics. Quart. J. Roy. Meteor. Soc., 19, 1105-1120.
______, and H. Bergstrom, 1996: Organized turbulence structures in the ______, J. A. Salmond, and A. N. V. Satyanarayana, 2006: Methodological
near-neutral atmospheric surface layer. J. Atmos. Sci., 53, 2452-2464. considerations regarding the measurement of turbulent fluxes in the
Holzäpfel F, and R. E. Robins, 2004: Probabilistic two-phase aircraft urban roughness sublayer: The role of scintillometery. Bound.-Layer
wake-vortex model: application and assessment. Journal of Aircraft, 41, Meteor., 121, 351-375.
1117-1126. Singha A., and R. Sadr, 2012: Characteristics of surface layer turbulence in
Howell J. F., and J. Sun, 1999: Surface layer fluxes in stable conditions. coastal area of Qatar. Environ. Fluid Mech., doi:10.1007/s10652-012-
Bound.-Layer Meteor., 90, 495-520. 9242-7.
Kader B. A., and A. M. Yaglom, 1990: Mean fields and fluctuation Sivaramakrishnan S., S. Sangeeta, and K. G. Vernekar, 1992: Characteristics
moments in unstably stratified turbulent boundary layers. J. Fluid of turbulent fluxes of sensible heat and momentum in the surface
Mech., 212, 637-662. boundary layer during the Indian summer monsoon. Bound.- Layer
Kaimal J. C., J. C. Wyngaard, Y. Izumi, and O. R. Coté, 1972: Spectral Meteor., 60, 95-108.
characteristics of surface layer turbulence. Quart. J. Roy. Meteor. Soc., Srivastava M. K., and P. Parth Sarthi, 2002: Turbulent kinetic energy in the
98, 563-589. atmospheric surface layer during the summer monsoon. Meteor. Appl.,
______, and J. J. Finnigan, 1994: Atmospheric boundary layer flows: their 9, 239-246.
structure and measurement. Oxford University Press, New York, 289 STORM Science Plan, 2005: Severe Thunderstorms- Observations and
pp. Regional Modeling (STORM) Programme, Science Plan, Dept. of
Kolmogorov A. N., 1941: Energy dissipation in locally isotropic tur- Science and Technology, Govt. of India, 118 pp.
bulence. Doklady AN SSSR, 32, 19-20. Stull R. B., 1988: An introduction to boundary layer meteorology. Kluwer
Krishnan P., and P. K. Kunhikrishnan, 2002: Some characteristics of Academic Publishers, 670 pp.
atmospheric surface layer over a tropical inland region during south- Thiermann V., and H. Grassl, 1992: The measurement of turbulent surface
west monsoon period. Atmos. Res., 62, 111-124. layer fluxes by use of bichromatic scintillation. Bound.-Layer Meteor.,
Li M., Y. Ma, W. Ma, Z. Hu, H. Ishikawa, Su Zhongbo, and F. Sun, 2006: 58, 367-389.
Analysis of turbulence characteristics over the northern Tibetan plateau Thomas C., and T. Foken, 2007: Flux contribution of coherent structures
area. Adv. Atmos. Sci., 23(4), 579-585. and its implications for the exchange of energy and matter in a tall
Li X., N. Zimmerman, and M. Princevac, 2008: Local Imbalance of spruce canopy. Bound.-Layer Meteor., 123, 317-337.
Turbulent Kinetic Energy in the Surface Layer. Bound.- Layer Meteor., Tillman A. S., 1972: The indirect determination of stability, heat and
30 November 2013 Bhishma Tyagi and A. N. V. Satyanarayana 601

momentum fluxes in the atmospheric boundary layer from simple scalar outflow, a thesis in atmospheric science. Master of Science Thesis
variables during dry unstable conditions. J. Appl. Meteorol., 11, 783- Submitted to the Graduate Faculty of Texas Tech University, 107 pp.
792. Wilcjack J. M., 1984: Large-scale eddies in the unstably stratified
Trinh K. T., 2012: On the Karman constant. In the Archives of arXiv.org, atmospheric surface layer, Part I: Velocity and temperature structure. J.
Cornell University. http://arxiv.org/ftp/arxiv/papers/1007/1007.0605.pdf Atmos. Sci., 41(24), 3537-3550.
Tyagi B., A. N. V. Satyanarayana, M. Kumar, and N. C. Mahanti, 2012: ______, S. P. Oncley, and S. A. Stage, 2001: Sonic anemometers tilt
Surface energy and radiation budget over a tropical station: An corrections algorithms. Bound.-Layer Meteor., 99, 127-150.
observational study. Asia-Pacific J. Atmos. Sci., 48(4), 411-421, DOI: Wyngaard J. C., and O. R. Cote, 1971: The budget of turbulent kinetic
10.1007/s13143-012-0037-z. energy and temperature variance in the atmospheric surface layer. J.
Vickers D., and L. Mahrt, 1996: Quality control and flux sampling Atmos. Sci., 28, 199-201.
problems for tower and aircraft data. J. Atmos. Oceanic Technol., 14, Yadav A. K., S. Raman, and M. Sharan, 1996: Surface turbulence spectra
512-526. and dissipation rates during low wind in tropics. Bound.-Layer Meteor.,
______, and ______, 2003: The cospectral gap and turbulent flux 79, 205-223.
calculations. J. Atmos. Oceanic Technol., 20, 660- 672. Yahaya S., J. P. Frangi, and D. C. Richard, 2003: Turbulent characteristics
Viswanathan D. V., A. N. V. Satyanarayna, S. Mishra, and P. Partha Sarthi, of a semiarid atmospheric surface layer from cup anemometers - effects
1997a: Turbulent kinetic energy budget parameter over Varanasi from of soil tillage treatment (Northern Spain). Ann. Geophys., 21, 2119-
MONTBLEX-90. Proc. Indian Natl. Sci. Acad., 63A, 403-412. 2131.
______, ______, M. K. Srivastava, and S. Mishra, 1997b: Surface layer Zhang H, and S. Park, 1999: Dissipation rates of turbulent kinetic energy
turbulent kinetic energy budget over Kharagpur. J. Scientific Res., 47, and temperature and humidity variances over different surfaces. Atmos.
11-20. ISSN: 0447-9483. Res., 50, 37-51.
Weinbeck S. W., 1997: Surface layer characteristics of thunderstorm