Analysis of the Nocturnal Development

1
Analysis of the Nocturnal Development
Dispersion of Smoke Puffs in the Atmosphere using
lidar

• -By
• Abhijeeth Linganagari, Dr. Ron Calhoun,
• Dr. H.J.S Fernando
• Department of Mechanical Aerospace Engineering
• ARIZONA STATE UNIVERSITY

2
OVERVIEW

• Motivation
• Introduction
• Review
• Observation
• Theory
• Results
• Conclusion Future Work

3
OVERVIEW

• Motivation
• Introduction
• Review
• Observation
• Theory
• Results
• Conclusion Future Work

4
MOTIVATION

• Perceive and comprehend the nocturnal development
  and dispersion of smoke puffs in the atmosphere.
• Puff coalescence, disintegration, puff
  meandering modeling of puff shape and size
  etc.
• Occurrence of nocturnal jets in the boundary
  layer.

5
OVERVIEW

• Motivation
• Introduction
• Review
• Observation
• Theory
• Results
• Conclusion Future Work

6
INTRODUCTION

• An atmospheric dispersion study was conducted in
  Oklahoma City, Oklahoma, in July 2003 during the
  Joint Urban study.
• Fireworks display on July 4th, 2003 (1000pm
  1040pm)
• The ASU lidar was located at the south-southeast
  of the CBD and ARL lidar was located towards the
  east. (Coherent Doppler Lidars were used by ASU
  and ARL)

7

• The exact locations of the lidars are given
  below
• ASU lidar - 35?26.330N, 97?29.533W
• ARL lidar - 35?28.387N, 97?30.265W
• The ASU lidar performed PPI (Plan Position
  Indicator) scans at elevation angles of 2.5 1
  degrees.
• The ARL lidar performed RHI (Range Height
  Indicator) scans at azimuthal angles of 270, 245,
  241, 209 degrees.

8
(No Transcript)
9
(No Transcript)
10
ASU Lidar
11
ASU Lidar Characteristics
12
(No Transcript)
13
(No Transcript)
14
SCANNING MODES
15
(No Transcript)
16
OVERVIEW

• Motivation
• Introduction
• Review
• Observation
• Theory
• Results
• Conclusion Future Work

17
REVIEW

• Stack plume model (e.g., Hanna et al. 1982),
• Turbulence from measured dispersion parameters
  (Nappo 1981).
• Urban dispersion studies (Biltoft 2002, for
  emergency response Van Aalst 1990).
• Remote sensing techniques were used to observe
  the shape, size, or optical changes in plumes
  over time. (Stoughton Miller 2002, Savov et al
  2002).

18

• Horizontal dispersion of non-buoyant plumes by
  fanning (Arya, 1999) in stratified boundary
  layers.
• Observations show the nocturnal boundary layer to
  be quite dynamic which makes plume fanning
  occurrence unusual. (Poulos et al., 2002)
• Measurement of vertical dispersion (Hiscox , et.
  al 2005 and 2006).

19
Areas of our focus

• Horizontal and Vertical dispersion parameters.
• Track the motion, shape, and size of puffs in the
  nocturnal boundary layer.
• Puff coalescence and disintegration.

20
OVERVIEW

• Motivation
• Introduction
• Review
• Observation
• Theory
• Results
• Conclusion Future Work

21
OBSERVATION

• Successful prediction and dealing of puffs
• An accurate dispersion model to yield the puff
  dispersion and spread rate.
• Predict shape and size of puffs,
• Various phenomena occurring due to interactions
  of puffs in the atmosphere.
• Knowledge of the local Topography and
  meteorological data.

22
(No Transcript)
23
OVERVIEW

• Motivation
• Introduction
• Review
• Observation
• Theory
• Results
• Conclusion Future Work

24
THEORY

• The two main mechanisms for dispersion
• Turbulent Diffusion causes mixing within the
  puff.
• Mean Shear responsible for stretching and
  shearing of the puff.
• Both these mechanisms may have been active on
  the night of July 4th , 2003.

25

• Assume vertical and horizontal crosswind conc.
  distribution in the puff to be Gaussian.
• 
  
  (1)
• where Ce and Cm are the edge and maximum
  concentration values, s - std. deviation of the
  puff concentration, ? - distance between the
  location of maximum concentration to the location
  of the edge contour.

26

• The concentration distribution in a puff is
  assumed proportional to the backscatter values
  (reasonable, given homogeneity of puff particle
  shape and constituency) obtained from the lidar
  scans.
• Ce a Be and Cm a Bm
  (2)
• where Be and Bm are the edge contour and
  maximum backscatter values of the puff.

27

• We obtain the backscatter values from the
  following relations
• 
  (3)
• where ?Bp - and define ?Bx and ?Bz as
  the amplitude of the Gaussian peaks in X and Z
  cross wind directions.
• B(x)x and B(z)z give the Backscatter values for
  the Gaussian curve fit in their respective
  directions.

28

• Eq(3) can be further simplified to get the
  following relations
• 
  (4)
• ?x and ?z are the distances from the point with
  maximum concentration. sx and sz signify the
  dispersion parameter to be calculated.

29

• sx is determined by minimizing the mean square
  error between estimated backscatter values from
  the lidar data, Bdata(x), model predictions
  B(x)x , similarly for sz
• 
  (5)
• where Nx and Nz are the number of lags used.

30
OVERVIEW

• Motivation
• Introduction
• Review
• Observation
• Theory
• Results
• Conclusion Future Work

31
RESULTS

• An estimate of the relative dispersion in the X
  and Z crosswind directions can be calculated from
  the ratio sx/sz.
• Backscatter values are plotted along the lines
  where PPI and RHI scans intersect and also at
  other cross sections through the puff.
• Gaussian curves are fitted to find sx and sz
  from these plots.

32

• Fig 1. Representative Lidar PPI scan (ASU)
  showing the location at which data is considered.

33

• Fig 2. Plot of sx (m) vs Mean square error
  (m-2sr-2) along the X cross wind direction at
  location C.

34

• Fig 3. Actual Backscatter values (m-1sr-1)
  plotted in comparison with the Gaussian fit along
  the horizontal cross-wind (X) direction at C.
• We obtain sx 147.75 m.

35

• Fig 4. Plot of sz (m) vs Mean square error
  (m-2sr-2) along the Z cross wind direction at
  location C.

36

• Fig 5. Actual Backscatter values (m-1sr-1)
  plotted in comparison with the Gaussian fit along
  the vertical cross-wind (Z) direction at C
  (from the Centre of Mass of the Puff, X -1980m
  Y 3800m)
• We obtain sz 33.93m.

37

• This gives us an idea of the relative dispersion
  estimates in both the directions, sx/sz 4.35.
• Thus horizontal cross-wind dispersion is about
  4.35 times that of the vertical cross-wind
  dispersion.
• We observe the fact that s (std. deviation of
  the puff concentration) is a function of time (t)
  and the distance along the puff in the direction
  of flow (Y).
• s s (y,t)
• So for a particular cross-section (Y) and time
  (t), s is a constant.

38
PUFF EVOLUTION

• The different stages in the evolution of the puff
  can be studied using the lidar.
• Track the puffs at different locations as they
  advect and observe shape changes and meandering
  in the path.
• Various phenomena like puff coalescence and
  disintegration are also noticed.

39
(No Transcript)
40
(No Transcript)
41
(No Transcript)
42
(No Transcript)
43
(No Transcript)
44
(No Transcript)
45
(No Transcript)
46
(No Transcript)
47
(No Transcript)
48

• Fig 6. Representative Lidar PPI scan (ASU)
  showing the location at which data is considered

49

• Fig 7. Plot of sx 1 (m) and sx 2 (m) vs Mean
  square error (m-2sr-2) along the X cross wind
  direction at location A

50

• Fig 8. Actual Backscatter values (m-1sr-1)
  plotted in comparison with the Gaussian fit in
  the horizontal cross-wind (X) direction at A.
• We obtain sx1 49m and sx2 127m.

51

• Fig 9. Plot of sx 1 (m) and sx 2 (m) vs Mean
  square error (m-2sr-2) along the X cross wind
  direction at location B.

52

• Fig 10. Actual Backscatter values (m-1sr-1)
  plotted in comparison with the Gaussian fit in
  the horizontal cross-wind (X) direction at B.
• We obtain sx1 134m and sx2 69m.

53
Pasquill Grifford Stability Class
54

• Fig 11. Std. deviations of the horizontal
  dispersion as a function of distance from the
  source according to Pasquill-Gifford.

55

• Fig 12. Std. deviations of the vertical
  dispersion as a function of distance from the
  source according to Pasquill-Gifford.

56
NOCTURNAL JETS

• The data provides evidence of nocturnal jet
  formation along with the development of a
  statically stable layer near the ground, with
  stability decreasing smoothly towards the neutral
  with height.
• The wind velocity profiles are time averaged over
  a 10 minute period (1030pm 1040pm local time)
  and plotted against the height above the ground
  at various distances from the lidar towards the
  CBD.

57

• Fig. Plot of wind velocity vs. height at
  different distances from the ARL lidar in the
  x-crosswind direction (for RHI scan with
  azimuthal angle 208o).

58
OVERVIEW

• Motivation
• Introduction
• Review
• Observation
• Theory
• Results
• Conclusion Future Work

59
CONCLUSIONS FUTURE WORK

• We have demonstrated the use of lidar in various
  aspects
• Tracking the evolution of the puff (coalescence
  and shape changes)
• Dispersion estimates in both the horizontal and
  vertical cross-wind directions
• Characterization of the wind field in the
  nocturnal boundary layer.
• The efficiency of this study can be increased by
  improving the scanning strategy of the lidars
  used.

60
SCANNING TECHNIQUE

• PPI scans with elevation angles increasing in
  steps of 0.5o from 1o to 5o so as to scan the
  region above the tall buildings in the CBD to
  higher up in the atmosphere
• A set of intersecting RHI scans with azimuthal
  angles varying in steps of 5o to cover the region
  of interest.

61

• Understand the variation of sx and sz with time
  and wind direction.
• Study the spread rate of the puff.
• An idea about puff meandering.

62
ACKNOWLEGMENT

• The support of the Army Research Office grant
  W911NF-04-1-0146, Project Officer W. Bach is
  gratefully acknowledged.
• Collaboration with the Army Research Laboratory
  for their lidar data was critical for the above
  analysis.

63
THANK YOU