On the optics of

1
On the optics of  light-scattering particles at
the mesopause from Lyman-? to the infra-red
Georg Witt Stockholm University
LPMR-8, Fairbanks,AK, August 20-23 2007
2
This contribution is an overview in a nutshell of
the current stand of optical-spectroscopic
research related to Noctilucent Clouds (in some
parts of the world also known as Polar
Mesospheric Clouds )
and an attempt to identify some problems and
questions which remain to be resolved.
with apologies for some of the text repeating
my words in 2001 in Monterey, then in the future
tense, but now in past tense
3
2007- a year of anniversaries

• Some 125 years have passed since Noctilucent
  Clouds have been properly identified as
  light-scattering particles in the 82-km height
  range.
• It is also 100 years ago that Gustav Mie and
  Pieter Debye independently presented the exact
  solution of light scattering by spherical
  objects. codified 50 years ago by H C van de
  Hulst in his classical book (vande Hulst,1957)

For many years, NLC remained a curiosity
restricted to the Northern Arctic that mainly
attracted astronomers interest and speculations.
Until, 25 years ago, B Fogle of GI placed them
in the Southern Hemisphere....
However, 50 years ago, Grishin in the USSR
published the first optical spectrum of NLC. At
about the same time, realising the detrimental
effect of atmospheric extinction, Witt used the
dual-wavelength polarisation as indicator of the
particle size. Then, almost exactly 40 years ago,
the first UV-VIS pair of rocket polarimeters
penetrated an NLC layer showing evidence that the
NLC layer is indeed sharply delimited from below
and the particles are indeed in the size range
foreseen. Also the layer was found sharply
delimited from below as expected for a
condensation process.
4
The recent rapid development of experimental
methods, computing power and not the least
Internet has led to an enormous leap forward in
our understanding of the atmosphere
notably, the inter-dependence of the Arctic
Mesosphere Region and the rest rest of the
SYSTEM and, not the least, the complicated
interplay of dynamics with gas-phase and
heterogeneous chemistry
5

What is it we really want to know ?
and of course the dynamic conditions required
for generating, maintaining and modulating an
aerosol layer at the mesopause

• Temporal and spatial variability
• Composition/Habit
• Shape/Orientation
• Size range
• Chemical interactions

6
Optical information desired

• Composition/Habit
• Structured optical absorption
• Shape/Orientation
• Polarisation state of scattered light
• Size range
• Spectral and angular scattering parameters

7
The Mueller matrix
the code of optical information
Stokes vector of the illuminating beam
Scattering matrix of target - ensemble of
randomly oriented particles
8
The matrix elements are ensemble averages over a
large population of particles. The contribution
of each kind (n) of particle is
This type of problem does not admit a unique
solution and a-priori assumptions are needed.
This in particular applies to the functional form
of the size distribution f (r) where the
choice has to comply with the mode of formation
of the aerosol.
The desired information can be derived from a set
of carefully chosen parameters and measuring
conditions. (Scattering angle and wavelength
combination). For NLC the expected particle size
is in the transition range between the Rayleigh
limit (IR) and full Mie scattering (UV).
9
The tale of the five blind people trying to
describe an elephant ....

• The main cause of ambiguity of interpretation of
  scattering measurements in temrs of an analytical
  size distribution function
• Different-size sub-groups of the size
  distribution dominate different scattering
  features such as
• Indicatrix - Polarisation state - Extinction
• at different scattering angles and wavelengths
• Polarisation effects as well as the colour index
  are most conspicuous in the UV limit and at
  scattering angles beyond 90

10
This brings up the question of how well the size
distribution can be retrieved from optical
measurements. In the size range of mesospheric
particles, the scattering is dominated by the
largest particles. To promote the small-particle
tail, measurements need to be made at a
wavelength where the scatterin is still a steep
function of the size while thegradient of the
large particles becomes modest ? Scattering at
Lyman-alpha (121.5 nm)
Attempted in the e-ARI experiment SLAM/ASLAF on
HotPay I in June with a catastrophic end ......
11

• One of the experimental problems related to this
  question is the size distribution of the
  particles
• The size range accessible to optical measurements
  is restricted to particles large enough to cause
  a deviation from atmospheric molecular
  scattering.
• Two ways to compensate for the lack of
  information
• Short wavelength observations Scattering of
  Lyman-alfa by NLC particles (Attempted by the
  SLAM experiment in 2007
• The combination of photometry with physical
  impactor techniques (Robertson and Horányi) and
  conductivity probes (Mitchell)

The question of particle formation Is it an
impulsive event with an explosive generation of
nuclei that grow to an optically sensible size,
limited nly by the available supply of water
molecules and the particles temperature ...
or is there a permanent population of small ice
clusters built up gradually during the period of
cooling that, given the thermal conditions at the
right altitudes produces NLC ?
12
HABIT AND STRUCTURE. Below 100 K, ice is
deposited as an amorphous substrate In the range
130 to 160 K cubic-symmetric ice is formed Above
180 K the ice has the normal, hexagonal
symmetry Then the optical constants can be
derived from standard optical measurements in
thin ice layers
BUT Does the assumption of structural
homogeneity hold for NLC particles ? And how much
of porosity can one expect to arise from the
collisional interactions between NLC particles
? This as well as the particle shape depend
crucially on the possibility of coagulation of
the NLC particles during their life cycle

13
Collisions between the mesospheric particles are
rare events and although coagulation is possible,
the process should not expected to mimic the
process as known in th troposphere. (which,
eventually, leads to the log-normal size
distribution) To illustrate the process, consider
the coagulation of hard spheres
14
The effect of clustering on the optical
properties can be treated by means of T-matrix
codes for interacting spheres (Mackowski) BUT the
particles are not hard spheres, and the collision
at thermal speed is not a simple billiard-ball
encounter...(P.Andersson al) The appropriate
modelling of the kinetics of collisional
coagulation in free-flow conditions remains to be
done
ABOUT THE COLLISIONAL COAGULATION PROBLEM
Starting from a rigid bi-sphere
15
Composition
The information is in the spectrally structured
extinction of sunlight. For H2O ice the spectral
region the important feature is the OH stretching
vibration at 3240 cm-1 (ACE-FTS) and the thermal
emission peaking near 12 microns (CRISTA).
Evidence from earlier experiments (HALOE) but
the proof (in the legal sense) that the main
component in NLC particles is H2O(s) comes from
the ACE Fourier extinction spectra.
16
Extinction in the IR region the OH stretching
feature
Temperature-dependent peak
Shape-dependent shoulder
17
Now to some highlights of the recent
development....
the rockets, to begin with
Experience The insight that the optical
measurements need to be complemented by other
techniques MISU C-59 and the development of
impactor techniques such as the recently flown
particle detectors (Rapp, Horányi) and
conductivity measurements (Mitchell)
In the range of scattering angles accessible to
photometric detection, the polarisation is not
sufficient to characterize the particle size. The
colour index and the symmetry of the phase
function are more sensitive indicators. (Gumbel
Witt)
18
1965
Current MISU photometer configuration
and reliable retrieval of the attitude...
19
LIDAR-Particle microphysics and dynamcis The
lidar technique developed from the primitive ruby
lidar attempts in the late 60-ies into powerful
systems allowing daytime high-resolution
detection of NLC. The depolarisation of the
scattered laser light has clearly shown that NLC
particles are not spherical.

• However, the lidar technique is hampered by the
  singular scattering angle of 180 which limits
  the information regarding polarisation.
• Possible new development for infirmation on hape
  and orientation
• IMAGING the illuminated spot
• CIRCULARLY polarized laser light
• BI-STATIC observation geometry

20
The global distribution and characterisation of
the light scattering layer at the mesopause
Rockets Novel techniques extending the range of
in-situ measurements to the small particle range
unaccessible by optical methods MASS (Horányi)
and the satellites SME UV symmetry of
the phase function Odin Scattered radiance
spectrum water abundace from thermal
rad. temperature ACE IR extinction spectrum,
H2O abundance and temperature AIM to be
told by others at this meeting
21

• Some highlights of Odin
• Sub-mm spectrometer
• Re-distribution of water molecules by ice-phase
  transport (S. Lossow)

22

1. Global Observations of Noctilucent CloudsB.
   Karlsson

northern hemisphere
southern hemisphere
- Particles tend to grow slightly larger towards
the pole
23
ACE highlights-the baseline of the IR spectrum is
the non-structured extinction by solid particles
and drops The important parameters of the
extinction spectra Peak optical thickness
(occasionally) 0.07 (!!!)
Frequency of the peak
32391 cm-Relative height of the shoulders
variable A general problem of IR occultation is
the scarce information on the low-temperature
refractive index of ice The only available data
set published by Clapp al is derived from
scattering measurement on artificial ice
particles by means of Mie theory
24
Maximum optical thickness observed 0.04-0.07
25
Large extinction signature of dominance of
small particles ?
Compare the integrated extinction with the number
of particles required to generate an NLC of equal
brightness as the molecular atmosphere above 60
km and 50 km.
26
New challenges...
A new experimental problem in the focus of
interest meteoric smoke as a possible nucleus
for water clusters and NLC .
27

• 2.Dreamed
• High-resolution H2O and temperature profiles
• The modification of the MISU optical hygrometer
  for daytime use
• mm-wave radiometer from rocket

Future experimental activities at MISU 1.
On-going
The PHOCUS science objectives.
smoke
heterogeneous chemistry ?
nucleation ? dirty ice ?
MISUIAP, FFI, TUG, NRL, Colorado,
ChalmersEsrangeEsrange lidar, ESRAD
ice
chemistry
bite-outs ? altered HOx chemistry ?
28
and finally, a hint for the future
exploit the ground-based CCD camera for
polarisation measurements from the ground. It is
time to re-examine the circular polarisation of
NLC reported by Gadsden al and disputed by
Bohren......... Thank you !
Concluding remarks

• Despite the increasingly intensive research
  related to NLC problems still exist which have
  not yet been resolved and motivate further work.
  
• The relation of the generation mechanism to
  organised and random atmospheric motions. This
  includes both large-scale dynamics and local
  structure variations, notably the temperature and
  local H2O abundance. /Optical hygrometer for
  daytime/
• The microphysics of particle growth including the
  mechanism determining the particle shape
• Improved polarimetric methods for better
  characterisation of the particle shape