Nitric acid distributions for Northern hemisphere winter 2002/3 observed by the MIPAS on ENVISAT and preliminary comparisons to a denitrification model.


22 févr. 2014 (il y a 3 années et 3 mois)

67 vue(s)

Nitric acid distributions for Northern hemisphere winter 2002/3 observed
by the MIPAS on ENVISAT and preliminary comparisons to a
denitrification model.

J.J. Remedios, and A.M. Waterfall

Earth Observation Science, Space Research Centre, University of Leic
ester, University Road, Leicester,
LE1 7RH, England.

R. Spang

ICG I, Forschungszentrum Jülich, Germany

G. Mann, S. Davies, and K. Carslaw

Institute for Atmospheric Science, School of the Environment, University of Leeds, Leeds, UK


The Michelso
n Interferometer for Passive
Atmospheric Sounding (MIPAS) was launched on
ENVISAT on 1st March 2002. Its operational product
suite includes retrievals of nitric acid profiles, with
vertical resolutions approaching 3 km, as well as
temperature, ozone and tr
acers (methane and nitrous
oxide). In this extended abstract, preliminary results
are shown which illustrate the comparison of the
nitric acid data to simulations performed with a
denitrification model. The clear signature of
denitrification is demonstrate
d for early January
2004, both from the data and from the model. Initial
results of this comparison suggest that the
denitrification model is able to simulate the minima
of nitric acid observed by MIPAS in this time period.
For this case study shown here,

the model also
suggests that local uptake into polar stratospheric
cloud particles is small compared with the
denitrification signal.

I. Introduction

Denitrification or irreversible loss of total nitrogen
from a given level in the atmosphere is a key proc
in the evolution of polar winters in the stratosphere.
Both local denoxification and irreversible
denitrification are caused by polar stratospheric
clouds (PSCs) through uptake of nitric acid into PSC
particles and potential subsequent sedimentation.
hese loss processes are significant factors in
subsequent development of chlorine activation and
related ozone loss.

In the Arctic stratosphere, despite temperatures
seldom falling below T

on synoptic scales,
denitrification has been measured in severa
l winters
to occur intensively and probably over a wide area
et al.
, 1999, Kondo
et al.
, 2000, Santee
et al.
, 2000]
. Nonetheless, it has been difficult to
obtain good spatial sampling of the Arctic polar
vortex on a daily basis with sufficien
t precision to
compare with models of denitrification. This has
become a serious limiting factor in differentiating
between models and in particular in testing the role of
large nitric acid trihydrate (NAT) particles (greater
than 10

m diameter) in this process. Such particles
were observed for the first time during winter
1999/2000 [Fahey
et al.
, 2001].

The Michelson Interferometer for Passive
Atmospheric Sounding (MIPAS) on ENVISAT
provides new data that can potentially provide a
gnificant step forward in these comparisons. In this
extended abstract, a preliminary comparison is
undertaken between maps of gas phase nitric acid as
observed by the MIPAS instrument and results from
the denitrification model DLAPSE, which
incorporates t
he growth and sedimentation of NAT
particles as well as other PSC types.

II. The MIPAS instrument

The Michelson Interferometer for Passive
Atmospheric Sounding (MIPAS) on ENVISAT is an
infrared Fourier transform spectrometer which
observes radiation emerg
ing from the Earth’s limb.

Vertical profiles of the radiance are obtained by
scanning the limb between 6 and 68 km at a spacing
of 3 km in the lower atmosphere. The vertical
resolution is also of the order of 3 km. Coverage
extends from pole to pole in fo
urteen orbits per day
although not all the data are available as yet for

Nitric acid retrievals are performed operationally at
ESA processing centres. The retrieval scheme uses a

global fit approach for each profile with no
a priori


The err
ors for these data at the altitudes of
concern are approximately 10 % for both systematic
and random components, and are described in Carli

(in press). In particular for HNO
, it is estimated
that systematic errors are of the order of 10% with
m errors nominally less than 10%. These data
are in the process of extensive validation but the
estimates of systematic and random errors are
supported by tests against UARS reference
atmosphere data with very good agreement.

A particular issue is the effe
ct of PSC particles on the
accuracy of MIPAS retrievals in addition to the
physical effects in the atmosphere through uptake of
nitric acid from the gas phase. It is possible to
examine this effect by detecting the presence of PSCs
using a cloud detection
method adapted for MIPAS
as described in Spang
et al.

(in press). This defines a
cloud index, which is lowered by the presence of
increasing cloud. Preliminary results indicate that
PSCs with a cloud index of lower than 2.2 could
cause retrieval errors. In

the comparisons shown in
this extended abstract, this is not an issue.

III. The DLAPSE denitrification model.

In this study, data from the MIPAS instrument have
been compared to the DLAPSE denitrification model,
which is described in detail in Carslaw et
al. (2002).
Briefly, the model simulates the time
growth and evaporation of several thousand
individual nitric acid hydrate particles, with their 3
motion calculated using isentropic trajectories
combined with vertical motion due to gravitation
sedimentation, which depends on particle size.
average NAT nucleation rate of 8.1 x 10

particles cm


is set in the model. Equilibrium
formation of super cooled ternary solution (STS)
particles is also included.
Particle growth by diffus
transfer of HNO

removes HNO

from the gas phase
in each model time step. Gas phase HNO

is then
calculated on a 3
D grid and advected each time step
using the off
line global Eulerian chemical transport
model SLIMCAT (Chipperfield et al., 1996). For
runs in the present study we have used a SLIMCAT
grid with 36 vertical levels between 350K and
2720K, with a resolution of 2.8

x 2.8


x 10K in the
lower stratosphere.
The model is forced by 6
wind and temperature fields from European Centre
or Medium
Range Weather Forecasts (ECMWF)
operational analyses, while vertical tracer advection
is calculated in isentropic coordinates from heating
rates using the MIDRAD radiation scheme (Shine,

For the comparisons to the MIPAS data shown here,
e model was initialised with MIPAS nitric acid data
for November 7
, which is prior to major PSC
activity in the Arctic polar vortex. Other species were
initialised with

fields of chemical species from a
annual model run.

IV. Results

inary results are shown here for one day,
January 3
, in winter 2002/3 to illustrate the
identification of denitrification and the comparison
between data and model. This day is excellent for
these purposes since no PSCs were detected with
MIPAS (cloud in
dex threshold of 3.0). Hence one
would expect the concentration of PSC particles to be
rather low and the uptake of nitric acid to be
relatively small.

Figures 1 and 2 show the MIPAS data and the
DLAPSE model data (smoothed with the MIPAS
averaging kernels
) on the 505 K surface.
Although no
PSCs are detected on this day by the MIPAS
instrument there is a general area of low HNO

a collar region. In addition, there is a further
depletion of nitric acid near Northern Scandinavia
and there is a more int
ense low of HNO

northeast of
Nuvya Zemlya (about 80N, 80E). We can be
confident that these depletions equate to real
processes in the atmosphere rather than retrieval error
since the lack of detection of PSCs implies no cloud
offsets in the spectrum that
can affect the retrieval.
Hence the intense low of HNO

could be due to either
removal of gas phase HNO

by PSCs which are too
optically thin in the infrared to be detected, or to
denitrification. The most significant evidence for
denitrification arises f
rom the black dotted line in
Figure 2 which indicates the area of potential NAT
formation according to temperature. The intense low
in nitric acid lies outside this region and hence is
most likely to result from irreversible loss from the
gas phase through

the preceeding part of the winter.


data on the 505K level for the 3

January, 2003.


concentrations at 505K on 3/1/2003 calculated
with the DLAPSE model and smoot
hed with the MIPAS
averaging kernels. The solid black line shows the vortex edge,
whilst the black dotted line indicates the area of potential NAT
formation where the temperature is less than 195K.

Current runs of the DLAPSE model predict less than
0.5 pp
bv of nitric acid uptake to PSCs anywhere at
this level. This is despite the presence of a large area
with T less than T

(the temperature below which
NAT exists). This is because the cold pool is in
through flow and NAT particles do not have long

lifetimes to take up PSCs as they are advected
out of the cold pool after only about 1
2 days.

Correlation plots of MIPAS HNO

versus MIPAS
O confirm the extent of denitrification.
Correlations for latitudes greater than 45ºN describe
well the relation
ship between the two tracer species
from November to January for regions outside the
vortex. However, within the vortex there is a clear
decay of nitric acid relative to that which would be
expected at 505K in the absence of previous PSC
activity. The stre
ngth of the lowering of nitric acid
values is consistent with strong denitrification having
occurred at previous times.


Correlation between MIPAS HNO

and N
O at 505K
on 3

January, 2003. The solid line shows the obs
correlation for the 23

November 2002 prior to significant PSC
activity for this winter.

V. Conclusions

New data from the MIPAS instrument are providing a
good source of comparisons to a denitrification
model. Initial results indicate the data and m
show excellent comparisons where PSC activity is
weak enough to neither influence the retrievals of
nitric acid nor to cause significant local uptake of
nitric acid. Hence it is clear that significant
denitrification took place in Arctic winter 2002/3
Furthermore, the good agreement suggests that the
denitrification model has performed well in
simulating denitrification in this winter through the
incorportation of growth and sedimentation schemes
for NAT particles.


The work in this s
tudy was supported by the
EU Framework 5 Project MAPSCORE EVK2
The preliminary MIPAS data were supplied by the European
Space Agency.


Carli, B. et al., First results of MIPAS/ENVISAT with
operational Level 2 code, submitted to
vances in Space

Carslaw, K. S.; Kettleborough, J. A.; Northway, M. J.; Davies, S.;
Gao, R.
S.; Fahey, D. W.; Baumgardner, D. G.; Chipperfield,
M. P.; Kleinböhl, A. A vortex
scale simulation of the growth
and sedimentation of large nitric ac
id hydrate particles,
Geophys. Res.
, 8300 (D20), doi 10.1029/2001JD000467,

Chipperfield, M. P.; Santee, M. L.; Froidevaux, L.; Manney, G.
L.; Read, W. G.; Waters, J. W.; Roche, A. E.; Russell, J. M.;
Analysis of UARS data in the southern pola
r vortex in
September 1992 using a chemical transport model,
J. Geophys.

(D13), 18861
18881, 1996.

Fahey D. W. et al., The detection of large HNO
particles in the winter arctic stratosphere and their role in

, 1026
1031, 2001.

Kondo, Y.; Irie, H.; Koike M.; Bodeker, G. E.; Denitrification
and nitrification in the Arctic stratosphere during winter of
Geophys. Res. Lett.
, 337
340, 2000.

Santee, M. L.; Manney, G. L.; Livesey, N. J.; Waters, J
. W.;
UARS Microwave Limb Sounder observations of
denitrification and ozone loss in the 2000 Arctic late winter,
Geophys. Res. Lett.
, 3213
2316, 2000.

Shine, K. P.: The middle atmosphere in the absence of dynamical
heat fluxes,
Q. J. R. Meteorol. Soc
, 603
633, 1987.

Spang R. and Remedios, J.J.; Colour indices for the detection and
differentiation of cloud types in infra
red limb emission
spectra, accepted for
Adv. Space Res.,

in print, 2004.

Waibel, A. E.; Peter, T.; Carslaw, K. S.; Oelhaf, H.
; Wetzel, G.;
Crutzen, P. J.; Poschl, U.; Tsias, A.; Reimer E.; Fischer, H.;
Arctic ozone loss due to denitrification,
, 2064
2069, 1999.