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
Altitude [km] NO 2 mixing ratio [ppb] Why measuring Atmospheric Trace Gases from the Ground? Satellite Observations provide global distribution of atmospheric trace gases but Mean tropospheric NO2 column density (Sep 2007-Aug 2008) derived from GOME-2 spectra. Steffen Beirle and Thomas Wagner, MPI Mainz. 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 1.5 NO 2 Profiles - Cabauw/Netherlands, 24.06.2009 Ground-Based MAX-DOAS Observations 1.0 0.5 6.0 5.0 4.0 3.0 2.0 1.0 0.0 provide vertical distribution of trace gases in the planetary boundary layer high temporal resolution (~15 min.) but 0.0 6 8 10 12 14 16 University of Heidelberg Time [UTC] measurements only at a single location
Absorption in the Earth s atmosphere DOAS
Absorption spectra of trace gases in the UV/Vis O 3 UV (log) O 3 Vis SO 2 NO 2 NO 3 BrO IO OIO I 2 ClO OClO OBrO HCHO Br 2 (CHO) 2 H 2 O HONO O 4 O 2 300 400 500 600 700 800 Wavelength [nm]
Intensity Absorption of light by gases Light source (Sun, Lamp) Spectrum Detector Electronic signal Absorbing gas 400 450 500 550 600 Wavelength Diffraction grating 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: I(, L) I ( ) exp L i(, T, p) i ( s) Rayl ( ) Mie( ) 0 i 0 air( s) ds Molecular absorption Scattering I 0,I: Incident and transmitted light intensity k : Wavelength, pressure and temperature dependent absorption cross section of the k th absorber Rayl : Rayleigh cross section Mie : Mie cross section (scattering on particles with r>> )
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 n ln I( k ) ln I0( k ) i( k Si cn k 2 ) k i n to determine the integrated concentrations along the light path (Differential Slant Column Density, dscd): The Optical Density is defined as ds i L i ( s) ds S 0 The polynomial c n n k removes the broad-banded 1.0 spectral structure caused by Rayleigh- and Mie- 0.8 '( ) = ( ) - 0.6 b ( ) scattering. Thus only compounds with high frequent 0.4 0.2 absorption features can be detected. The high 0.0-0.2 frequent parts of and are referred to as the -0.4-0.6 differential absorption cross section and optical B -0.8 density and. 350 355 360 365 370 375 380 385 ref I( ) ( ) ( ) S ln I0( ) [10-19 cm 2 ] ' [10-19 cm 2 ] 6.0 5.5 5.0 4.5 b ( ) A [nm] ( )
Optical density [x10-3 ] DOAS Analysis 380 360 340 320 0-5 -10-30 -40 0-2 -4-60 -2-4 1.0 0.5 0.0-0.5 Spectrum Polynomial Ozone NO 2 Formaldehyde Ring Fit Residual 340 345 350 355 [nm] 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
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 O 4 ) allow for the retrieval of aerosol extinction profiles Trace gas or
A typical MAX-DOAS Instrument
dscd BrO [10 14 molec/cm 2 ] BrO DSCDs OASIS Campaign, Barrow, AK, 2009 6 Clear sky 4 dscd O 4 [10 43 molec 2 /cm 5 ] Low visibility 2 0-2 6 U-shaped profile Stratospheric BrO 4 Separation of elevation angles Tropospheric BrO Elevation angle 01 02 05 10 20 90 1.500 2.500 6.000 11.00 89.00 2 0 06.04.2009 07.04.2009 08.04.2009 09.04.2009 10.04.2009 11.04.2009 12.04.2009 13.04.2009 14.04.2009 15.04.2009 16.04.2009
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 Backward modelling Forward modelling Random sampling of photon paths Simulation of the radiative transfer through broken clouds Rayleigh Cloud and aerosols Absorption Ground reflection Deutschmann et al., JQSRT, 2013
Spectra Retrieval of Trace Gas and Aerosol Vertical Profiles DOAS Analysis O 4 dscds O 4 errors Rel. Intensity Intensity errors Trace gas dscds Trace gas errors Aerosol Retrieval Radiative Transfer Model a priori aerosol profile a priori covariance a priori trace gas profile a priori covariance Trace Gas Retrieval Radiative Transfer Model Aerosol profile Aerosol opt. properties Retrieval covariance Trace Gas profile Retrieval covariance
O 4 optical depth 0.030 0.025 Example for Trace Gas and Aerosol Retrieval CINDI Campaign, Cabauw/Netherlands, 2009 O 4 dscds Measurement Simulation NO 2 dscds 0.020 0.015 0.010 30 29.00 15 14.00 8 7.000 4 3.000 2 0.005 0.000 06:00 09:00 12:00 15:00 18:00 Time [UTC] Aerosol Profiles NO 2 Profiles
Altitude [km] Trace Gas and Aerosol Retrieval Averaging Kernels CINDI Campaign, Cabauw/Netherlands, 2009 4.0 02.07.2009, 12:00 3.5 3.0 2.5 2.0 1.5 1.0 d s = 1.9 d s = 3.1 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 0.5 0.0 0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8 Aerosol Averaging Kernel NO 2 Averaging Kernel
Hohenpeißenberg, Germany In collaboration with German Weather Service (DWD) 90 10 20 100 m 200 m 5 2 1-2 -10
Hohenpeißenberg, Germany NO 2 Profiles, 8.7.2010 NO 2 Surface Mixing Ratio Comparison with in situ
Hohenpeißenberg, Germany NO 2 Surface Mixing Ratio Comparison with in situ measurements University of Heidelberg
Extinction [km -1 ] Extinction [km -1 ] Altitude [km] Altitude [km] Backscatter [A.U.] Backscatter [A.U.] Comparison of aerosol extinction profiles CINDI Campaign, Cabauw, Netherlands, July 3, 2009 July 4, 2009 3- Ceilometer (20 min. average) 3- Ceilometer (20 min. average) 2-2- 1-0- 3-2- Ceilometer (Av. kernel applied) 1-0- 3-2- 10- Ceilometer 5- (Av. kernel applied) 0-10- 5-0- 1-1- 0-3- BIRA 0-3- BIRA 2-2- 1-1- 0-3- Heidelberg 0-3- Heidelberg 2-1- 0-3- JAMSTEC 2-1- 0-3- JAMSTEC 0.6-0.4-0.2-0.6-0.4-0.2-2- 2-0- 0-1- 1-0- 3- MPI-Mainz 0-3- MPI-Mainz 2-2- 1-1- 0-6 9 12 15 18 Time [hours UTC] 0-6 9 12 15 18 Time [hours UTC]
MAX-DOAS Measurements of BrO and Aerosols in Barrow, Alaska during the OASIS field Campaign, February-April 2009 Udo Frieß and Holger Sihler normalized BrO-VCD - 12/04/2009 80 135-90 70 150-105 165-120 60 180-165 -150-135 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
200 km visibility
500 m visibility
100 m visibility
50 cm visibility
MAX-DOAS Measurements of BrO and Aerosols in Barrow, Alaska Example for the Diurnal Variation of Aerosol Extinction Extinction profiles - Barrow, Alaska 11:00 15:15 University of Heidelberg
MAX-DOAS Measurements of BrO and Aerosols in Barrow, Alaska Ceilometer Comparison with Ceilometer Ceilometer, vertical resolution degraded to MAX-DOAS MAX-DOAS
AOD MAX-DOAS Measurements of BrO and Aerosols in Barrow, Alaska Comparison with Sun Photometer 10 MAX-DOAS Sun photometer Sun photometer, quality filtered 1 0.1 University of Heidelberg 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 Date 2009
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 retroreflectors. 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 Passive Optical Resonatpr
The NO 2 CE-DOAS Instrument Cavity: R 99.975 (@445nm) Opt. Path Length L 0 1.8km Weight ~ 5kg Blue Led (Peak at 445nm) Fit Range 436nm - 453nm Det. Limit 1ppb @ 2s <0.5ppb @ 10s
The NO 2 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. with [Platt et. al. 2009] Path length correction depends on absolute optical density. Direct correction with meas. D CE possible but requires: light source with very stable intensity very stable optomech. setup Difficult to achieve. Temp. stab. of LED required L 0 =1.8 km
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 cm 3, ~ 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