Remote Sensing of Atmospheric Trace Gases

Udo Frieß Institute of Environmental Physics University of Heidelberg, Germany

CREATE Summer School 2013 Lecture B, Wednesday, July 17

Remote Sensing of Atmospheric Trace Gases:

Outline

• Ground-Based Differential Optical Absorption

Spectroscopy (DOAS):

– The Beer-Lambert Law

– Multi-Axis DOAS

– Inverse modelling: Retrieval of trace gas and aerosol vertical

profiles from MAX-DOAS

– Longpath-DOAS

– Cavity-Enhanced DOAS

Why measuring Atmospheric

Trace Gases from the Ground?

University of

Heidelberg

Mean tropospheric NO2 column density (Sep 2007-Aug 2008) derived from GOME-2 spectra.

Steffen Beirle and Thomas Wagner, MPI Mainz.

Satellite Observations

• provide global distribution of atmospheric

trace gases

but

• only very limited information on the vertical

distribution of trace gases

• only one (or very few) measurement(s) per

day, always at (nearly) the same time

6 8 10 12 14 16

0.0

0.5

1.0

1.5

NO2 Profiles - Cabauw/Netherlands, 24.06.2009

Time [UTC]

Alti

tud

e [

km]

0.0

1.0

2.0

3.0

4.0

5.0

6.0N

O2 m

ixin

g r

atio

[p

pb

]

Ground-Based MAX-DOAS Observations

• provide vertical distribution of trace gases

in the planetary boundary layer

• high temporal resolution (~15 min.)

but

• measurements only at a single location

Absorption in the Earth‘s atmosphere

D

OA

S

Absorption spectra of trace gases

in the UV/Vis

300 400 500 600 700 800

Br2

ClO

OBrO

H2O

HONO

(CHO)2

HCHO

O2

O4

OClO

OIOI2

IOBrO

NO3

NO2SO

2

O3 Vis

Wavelength [nm]

O3 UV

(log)

Absorption of light by gases

Light source

(Sun, Lamp)

Diffraction grating

Spectrum

Detector

Electronic signal

400 450 500 550 600

Inte

nsity

Wavelength400 450 500 550 600

Inte

nsity

Wavelength

Absorbing gas

Electronic signal

The absorption strength (optical

density) depends on

• the concentration of the gas

• the length of the light path

• the ability of each molecule

to absorb light of a specific

wavelength (absorption

cross section)

The amount of gases along

a light path can be

determined using their

individual absorption

structure

Differential Optical Absorption Spectroscopy

- DOAS -

Assume a light beam traversing an infinitesimally

small air parcel containing a mixture of N gases

with concentrations i (i=1..N)

The extinction of light is described by the Beer-Lambert law:

L

i

airMieRaylii dssspTILI0

0 )()()()(),,(exp)(),(

I0,I: Incident and transmitted light intensity

k: Wavelength, pressure and temperature dependent absorption cross section of the kth absorber

Rayl: Rayleigh cross section

Mie: Mie cross section (scattering on particles with r>>)

Molecular absorption Scattering

Remote sensing of atmospheric trace gases:

Differential Optical Absorption Spectroscopy

(DOAS) When sampling the light intensity on a discrete wavelength grid k (and neglecting the pressure and temperature dependence of the absorption cross section), the Beer-Lambert law can be solved numerically by minimising

k i n

n

nikikk kcSII )()(ln)(ln 0

2

to determine the integrated concentrations along the light path (Differential Slant Column Density, dSCD):

ref

L

ii SdssdS 0

)(

The polynomial cnkn removes the broad-banded

spectral structure caused by Rayleigh- and Mie- scattering. Thus only compounds with high frequent absorption features can be detected. The high frequent parts of and are referred to as the differential absorption cross section and optical density ’ and ’. 350 355 360 365 370 375 380 385

-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0

'( ) = ( ) - b ( )

B ' [

10

-19

cm

2 ]

[nm]

4.5

5.0

5.5

6.0

b ( )

( ) A

[1

0 -1

9 c

m 2

] The Optical Density is defined as

)(

)(ln)()(

0

I

IS

DOAS Analysis

• Simultaneous detection of

several trace gases

• Retrieval algorithm based

non-linear leas squares

algorithm

• Optical densities of less

than 5x10-4 can be detected

• Detection limit depends on

– residual noise

– light path length

– absorption cross section

– possible interference with

other species

340 345 350 355

-0.50.00.51.0

[nm]

Fit Residual

Spectrum

320

340

360

380

Polynomial

-10

-5

0

Ozone

-40

-30

Optical density [x10

-3]

NO2

-6

-4

-2

0

Formaldehyde

-4

-2

0

Ring

Multi-AXis Differential Optical Absorption

Spectroscopy (MAX-DOAS)

• Variation of SCD with solar

zenith angle (SZA) allows to

determine the stratospheric VCD

• Sequential measurements at

different elevation angle allow for

the retrieval of the trace gas

vertical distribution

• Measurements of an absorber

with known vertical distribution

(oxygen collision complex O4)

allow for the retrieval of aerosol

extinction profiles Trace gas or

A typical MAX-DOAS Instrument

BrO DSCDs OASIS Campaign, Barrow, AK, 2009

06.04.200907.04.2009

08.04.200909.04.2009

10.04.200911.04.2009

12.04.200913.04.2009

14.04.200915.04.2009

16.04.2009

0

2

4

6

-2

0

2

4

6

dS

CD

BrO

[10

14 m

ole

c/c

m2]

1.500

2.500

6.000

11.00

89.00

01°

02°

05°

10°

20°

90°

dS

CD

O4

[10

43 m

ole

c2/c

m5]

Elevation

angle

U-shaped profile

Stratospheric BrO Separation of elevation angles

Tropospheric BrO

Clear sky

Low visibility

Monte-Carlo Modelling of Radiative Transfer

• Aim: to model the transport of radiation

through the atmosphere as a function of:

– Trace gas vertical profiles

– Viewing geometry

– Atmospheric state (T, p, humidity,

aerosols, …)

• Modelled quantities are: – Radiances

– (Box- ) airmass factors

– Derivatives of measured quantities with

respect to atmospheric state parameters

Deutschmann et al., JQSRT, 2013

Forward modelling Backward modelling

Random sampling of photon paths

Simulation of the radiative transfer through broken clouds

Rayleigh

Cloud and aerosols

Absorption

Ground reflection

Retrieval of Trace Gas and Aerosol Vertical Profiles

Rel.

Intensity

Intensity

errors

O4

dSCDs

O4

errors

Aerosol Retrieval

Radiative Transfer Model

a priori

aerosol profile

a priori

covariance

Aerosol

profile

Aerosol opt.

properties

Retrieval

covariance

Trace gas

dSCDs

Trace gas

errors

Trace Gas Retrieval

Radiative Transfer Model

a priori

trace gas profile

a priori

covariance

Trace Gas

profile

Retrieval

covariance

DOAS

Analysis

Spectra

Example for Trace Gas and Aerosol Retrieval CINDI Campaign, Cabauw/Netherlands, 2009

NO2 dSCDs

NO2 Profiles

O4 dSCDs

Aerosol Profiles

06:00 09:00 12:00 15:00 18:00

0.000

0.005

0.010

0.015

0.020

0.025

0.030

8°

4°

2°

30°

Measurement

Simulation

O4 o

ptic

al d

ep

th

Time [UTC]

3.000

7.000

14.00

29.00

15°

Trace Gas and Aerosol Retrieval – Averaging Kernels CINDI Campaign, Cabauw/Netherlands, 2009

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8

Aerosol Averaging Kernel

Altitu

de

[km

]

3900 m

3700 m

3500 m

3300 m

3100 m

2900 m

2700 m

2500 m

2300 m

2100 m

1900 m

1700 m

1500 m

1300 m

1100 m

900 m

700 m

500 m

300 m

100 m

ds = 1.9

NO2 Averaging Kernel

ds = 3.1

02.07.2009, 12:00

Hohenpeißenberg, Germany In collaboration with German Weather Service (DWD)

-10° 200 m

90°

10°

20°

2° 5°

1°

-2°

100 m

Hohenpeißenberg, Germany

NO2 Profiles, 8.7.2010

NO2 Surface Mixing Ratio

Comparison with in situ

Hohenpeißenberg, Germany

University of

Heidelberg

NO2 Surface Mixing Ratio – Comparison with in situ measurements

Ba

ck

sc

att

er

[A.U

.]

10-

5-

0-

Ex

tin

cti

on

[km

-1] 0.6-

0.4-

0.2-

0-

Alt

itu

de

[km

]

3-

2-

1-

0-

3-

2-

1-

0-

3-

2-

1-

0-

3-

2-

1-

0-

3-

2-

1-

0-

3-

2-

1-

0-

6 9 12 15 18

Time [hours UTC]

Ceilometer (20 min. average)

Ceilometer (Av. kernel applied)

BIRA

Heidelberg

JAMSTEC

MPI-Mainz

July 3, 2009

Ba

ck

sc

att

er

[A.U

.]

10-

5-

0-

Ex

tin

cti

on

[km

-1] 0.6-

0.4-

0.2-

0-

Alt

itu

de

[km

]

3-

2-

1-

0-

3-

2-

1-

0-

3-

2-

1-

0-

3-

2-

1-

0-

3-

2-

1-

0-

3-

2-

1-

0-

6 9 12 15 18

Time [hours UTC]

Ceilometer (20 min. average)

Ceilometer (Av. kernel applied)

BIRA

Heidelberg

JAMSTEC

MPI-Mainz

July 4, 2009

Comparison of aerosol extinction profiles

CINDI Campaign, Cabauw, Netherlands,

-120

-105

-90

180 -165 -150 -135

60

70

135

150

165

80

normalized BrO-VCD - 12/04/2009

BrO from GOME-2

(Holger Sihler)

Frieß et al., The Vertical Distribution of BrO and Aerosols in the Arctic: Measurements by Active and Passive DOAS,

JGR, 2011

MAX-DOAS Measurements of BrO and Aerosols in Barrow, Alaska

during the OASIS field Campaign, February-April 2009 Udo Frieß and Holger Sihler

200 km visibility

500 m visibility

100 m visibility

50 cm visibility

Extinction profiles - Barrow, Alaska 11:00 15:15

MAX-DOAS Measurements of BrO and

Aerosols in Barrow, Alaska

Example for the Diurnal Variation of Aerosol Extinction

University of

Heidelberg

MAX-DOAS Measurements of BrO and

Aerosols in Barrow, Alaska

Comparison with Ceilometer

Ceilometer

Ceilometer, vertical resolution degraded to MAX-DOAS

MAX-DOAS

3.4 4.4 5.4 6.4 7.4 8.4 9.4 10.4 11.4 12.4 13.4 14.4 15.4

0.1

1

10

AO

D

Date 2009

MAX-DOAS

Sun photometer

Sun photometer, quality filtered

MAX-DOAS Measurements of BrO and

Aerosols in Barrow, Alaska

Comparison with Sun Photometer

University of

Heidelberg

MAX-DOAS Measurements of BrO and Aerosols in

Barrow, Alaska

Longpath-DOAS

Direct measurement of the average trace gas concentration along the light path

Long-Path DOAS in Barrow The Sending/Receiving Telescope

Long-Path DOAS in Barrow Retro Reflector Setup

Long light path Short light path

Comparison of BrO

measured by LP-DOAS and CIMS (Chemical Ionisation Mass Spectrometry)

Liao, J.; Huey, L. G.; Neuman, J. A.; Tanner, D. J.; Sihler, H.; Frieß, U.; Platt, U.; Flocke, F. M.; Orlando, J. J.; Shepson, P. B.;

Beine, H. J.; Weinheimer, A.; Sjostedt, S. J.; Nowak, J. B. & Knapp, D.:

A comparison of Arctic BrO measurements by a chemical ionization mass spectrometer (CIMS) and a long path differential

optical absorption spectrometer (LP-DOAS), submitted to J. Geophys. Res., 2010.

Tomographic LP-DOAS

• Numerous light paths with

two LP-DOAS instruments

using mirrors and retro-

reflectors.

• The measured averaged

concentrations along the

light paths allow to

reconstruct the

concentration field using

tomographic methods.

Denis Pöhler, Ph.D. thesis, 2010

How to fit a long light path into a compact setup P

assiv

e

Op

tica

l

Resonatp

r

The NO2 CE-DOAS Instrument

Cavity:

• R≈99.975 (@445nm)

• Opt. Path Length L0≈1.8km

• Det. Limit 1ppb @ 2s

<0.5ppb @ 10s

• Weight ~ 5kg

• Blue Led (Peak at 445nm)

• Fit Range 436nm - 453nm

The NO2 CE-DOAS Instrument

CE-DOAS Theory

• Optical density is calculated relative to intensity

transmitted by a „zero air“ filled resonator.

• Difference to LP-DOAS:

• Wavelength dependent path length

• Path length must be calibrated: -> Purge cavity with gases of known extinction We use an alternating purge with

Helium and zero air

CE-DOAS Theory

• Length of optical light path depends on extinction in the resonator. Strong

absorber shorter light path.

• Path length correction depends

on absolute optical density.

• Direct correction with meas. DCE

possible but requires:

• light source with very

stable intensity

• very stable optomech. setup

Difficult to achieve.

Temp. stab. of LED required

L0=1.8

km

with [Platt et. al. 2009]

Mobile Meas. in Hong Kong Dec. 13-20, 2010

Central area of Hong Kong

Open Path CE-DOAS Measurements

of IO in Antarctica • Open path CE-DOAS

• Spectral region: 425-455 nm

• Mirror distance: ~1.9 m

• Light source: blue LED

• Spectrometer: Avantes AvaSpec-ULS2048

• Ring-down system with PMT tube

• Light path ~6 km

• Detection limit for IO ~0.5 ppt

• Power consumption: ~30 W

• Dimensions: 220 x 30 x 40 cm3, ~ 30 kg

• Temperature range: -45°C to +30°C

Summary

• Atmospheric remote sensing allows for the contact-free detection of atmospheric

constituents (trace gases, aerosols, clouds...)

• DOAS is a versatile measurement technique:

– Simultaneous measurement of numerous atmospheric trace gases

– Very sensitive (< 1 ppt for some species)

– Inherently self-calibrating

– Can be operated from various platforms (ground, air-borne, balloon-borne, satellite

borne)

• MAX-DOAS:

– Allows for the determination of stratospheric and tropospheric trace gases and aerosols

– Simple and robust instrumentation suitable for field experiments and long-term

measurements

• LP-DOAS: – Very accurate measurement of the mean concentrations of trace gases along a light

path of several km

• CE-DOAS: – Compact setup with high sensitivity due to light paths of several km on a desktop