Preliminary testing of new approaches to retrieve aerosol properties from joint photometer-lidar inversion

Transcription

1 ESA/IDEAS Project- WP Preliminary testing of new approaches to retrieve aerosol properties from joint photometer-lidar inversion Q. Hu, P. Goloub, O. Dubovik, A. Lopatin, T. Povdin, T. Lopyonok, B. Torres, Y. Derimian, I. Popovici 1

2 SHADOW campaign Objectives (March-April, 2015 & December, 2015-January, 2016) Saharan dust properties exploration New retrieval approaches testing Produce new and high quality aerosols parameters (needed for satellite validation) Site selection Dust investigation near the source Transported biomass burning and other aerosol types Keep continuous and relevant aerosol study Collaboration/Cooperation Cappa Labex ACTRIS community Fig 1. M Bour IRD (Institut de Recherche pour le Développement) 2

3 Aerosol parameters investigated H Vertical concentration Size distribution R Optical properties Extinction Absorption Scattering Microphysical properties Vertical concentration Size distribution Complex Refractive Index (CRI) Particle shape Chemical properties Fig 2. Aerosol parameters investigated Chemical speciation 3

4 Instrumentation Multi-wavelength Raman Lidar 2α : Extinction coefficient at 355nm, 532nm 3β: Backscattering coefficient at 355nm, 532nm, 1064nm 2δ: Particle Depolarization Ratio (PDR) at 355nm, 532nm LR: Lidar Ratio BAE,EAE: Backscattering Angstrom Exponent and Extinction Angstrom Exponent. WVMR: Water Vapor Mixing Ratio Sun/Lunar photometer Direct sun/moon extinctions Angular sky radiance & polarization measurements 4

5 PDR (δ) is defined as the ratio of the cross-polarized Lidar returned signal to the parallel polarized signal, with respect to the polarization plane of the transmitted laser beam. δ can be used as an indicator of the particle shapes: Spheres such as fogs, droplets : ~0% Molecular atmosphere: ~0.37% % Dust particles: ~30%-35% Cirrus clouds: ~40%-45% LR is the ratio of aerosol extinction and backscattering coefficient LR is independent of aerosol concentration, only depends on the aerosol property. 5

6 Instrumentation CIMEL Lidar single wavelength Lidar Backscattered signal at 532nm Doppler Lidar Vertical resolved wind speed and directions Airborne photometer PLASMA Spectral AOTs 6

7 Particle sizing instruments - GRIMM, OPC Particle size distributions Nephelometer Scattering coefficient (450, 525 and 625nm) Aethalometer Absorption coefficient (370, 470, 520, 590, 660, 880 and 950nm) Other instruments Particle mass concentration, chemical speciation 7

8 Instrumentation Fig 3. Instruments in SHADOW campaign 8

9 Case studies Case 1: 25 January, 2016 Fig 4. Diurnal AOD and Angstrom exponent variation. (25 January, 2016) 9

10 532nm Layer 2 Layer 1 Fig 5. Lidar quicklook. (25 January, 2016) 10

11 Raman Retrieval (a) Fig 6. (a)extinction and backscattering. (b) Particle depolarization and Angstrom Exponent (355/532) (25 January, 2016). Data is averaged between 14:00-16:00 UTC. (b) 11

12 Case 2 :20 January, 2016 (UTC) Fig 7. Diurnal AOD and Angstrom exponent variation. (20 January, 2016) 12

13 532nm Fig 8. Lidar quicklook. (20 January, 2016) 13

14 Raman Retireval (a) Fig 9. (a)extinction and backscattering coefficient. (b) Particle depolarization and Angstrom Exponent (355/532) (20 January, 2016). Data is averaged between 08:20-09:40 UTC. (b) 14

15 Aerosol classification Fig 11. Aerosol classification of PDR and LR at 355. (Picture taken from presentation of Ulla Wandinger) Fig 10. Retrieved Lidar ratio. (20 January, 2016) 15

16 GARRLiC/GRASP retrieval AOTs τ(λ i ) scattered radiances I(λ i,θ j ) Backscattered Lidar signal LS(λ m, h n ) GARRLiC/GRASP Aerosol vertical concentration C f (h n ) C c (h n ) Aerosol size distribution dv f (r) dlnr dvc (r) dlnr Complex Refractive Index n f (λ k ),n c (λ k ) κ f (λ k ),κ c (λ k ) Spherical particles fraction C sph Fig 12. Inputs and outputs of GARRLiC/GRASP 16

17 Extinction profile Raman 355-Raman 532-GARRLiC 355-GARRLiC 532EXT-in situ EXT_GARRLiC 532EXT_GARRLiC 532EXT-in situ 1 532EXT-Lidar 355EXT- Lidar Height(m) Height (m) ? Extinction coefficient(1/km) (a) Extinction coefficient (1/km) Fig 13. Extinction retrieved from GARRLiC and comparison with Raman retrieval. (a)25 January (b) 20 January (b) 17

18 Aerosol vertical concentration Fine mode Coarse mode 5000 Fine mode Coarse mode Height(m) Aerosol concentration(ppb) (a) Aerosol concentration(ppb) Fig 14. Retrieved aerosol vertical concentration. (a) 25 January. (b) 20 January. (b) 18

19 Size distribution and CRI Contamination from incomplete sun photometer measurements Fig 15. Size distribution. Fig 16. Complex refractive index(cri). Real part (top) and imaginary part (bottom) 19

20 SSA profile nm-20 Jan 532nm-20 Jan 1064nm-20 Jan 355nm-25 Jan 532nm-25 Jan 1064nm-25 Jan 2500 SSA at 2700m Height(m) SSA at 1250m SSA Fig 17. Profile of single scattering albedo (SSA) 20

21 Absorption profile ABS 532ABS 1064ABS 532ABS-in situ 1 355ABS-in situ ABS 532ABS 1064ABS 532ABS-in situ 1 532ABS-in situ Height (m) 3000 Height (m) Absorption coefficient (1/km) (a) Absorption coefficient(1/km) Fig 17. Retrieved aerosol vertical absorption profile. (a) 25 January. (b) 20 January. (b) 21

22 Conclusions Conclusions and perspectives WP Deliverable : Technical Report on "Preliminary testing of new approaches to retrieve aerosol properties from joint photometer-lidar inversion GARRLiC/GRASP has extended the retrieved parameter dataset by providing bimodal and vertical distributed aerosol properties. Promising results,coincident with present understanding of aerosol properties, are obtained for the measurements from SHADOW campaign. Raman retrievals generally show good agreements with the reported aerosol characteristics in the previous campaigns. However, there are still some discrepancies to be further explored. 22

23 Perspectives Conclusions and perspectives Retrieval products validation, more in-situ/remote sensing measurements will be analyzed and samples studied by chemical and physical laboratorial methods. Algorithm implementation, ground-based measurements and depolarization information maybe employed into the retrieval. Uncertainty study for both measurements and retrieval. GARRLIC/GRASP routine processing of European sites in preparation 23

24 Thanks for your attention! 24

25 Extra Slides 25

26 Water vapor mixing ratio (kg/kg) Fig 18. Angstrom Exponent and water vapor mixing ratio 02:00-04:00, 20 January 26

27 Retrieved Angstrom Exponent Height (m) Angstrom Exponent 27

28 5H_6H Fig 19. Angstrom Exponent and water vapor mixing ratio 05:00-06:00, 20 January 28

29 Extinction comparison (a) Fig 20. Extinction comparison between PLASMA, Lidar and GRIMM. Lidar extinction computed from Klett method. (a) 355 channel. (b) 532 channel (b) 29