Aerosol Inherent Optical Properties

Transcription

1 Aerosol Inherent Optical Properties Aerosol Inherent Aerosol Inherent Optical Properties Optical Properties R. Santer, F. Zagolski and O. Aznay Université du Littoral 1

2 Outline Validation of the aerosol product Alternative aerosol IOPS Perpectives 2

3 The test sites equipped with CIMEL Venice Lampedusa El Arenosillo Lanai San Nicolas 3

4 Validation days site date time s Venise 10/08/ :34: Venise 13/08/ :40: Lampedusa 20/08/ :20: Gotland 07/09/ :56: Gotland 08/09/ :20: Helgoland 16/09/ :10: Lampedusa 21/09/ :13: Venise 16/10/ :30: Venise 25/10/ :49: Calcofi 28/10/ :18: Venise 04/11/ :35: Calcofi 10/11/ :10: Venise 10/11/ :47: Venise 13/11/ :53:

5 0,000-0,500 alpha MERIS -1,000-1,500-2,000-2,500-3, ,5-2 -1,5-1 -0,5 0 first processing second processing Linéaire (first processing) alpha CIMEL Linéaire (second processing) y = 0,3236x - 0,5081 R 2 = 0,1357 y = 0,9965x + 0,0266 R 2 = 0,6663 5

6 0,4 0,3 AOT 865 MERIS 0,2 0, ,1 0,2 0,3 0,4 AOT 870 CIMEL first processing second processing Linéaire (first processing) MAVT-2006 ESRIN - Linéaire March (second 2006 processing) y = 0,5486x + 0,0873 R 2 = 0,6708 y = 0,4564x + 0,0612 R 2 = 0,5727 6

7 Conclusion 1 A good retrieval of α MERIS overestimates AOT_865 7

8 The atmospheric correction Formulation of the signal L toa = ( L atm + T * L g + t * (L w ) )T g sunglint Level-2 What do we need? Compute the atmospheric functions: -with a radiative transfer code -and standard values for aerosol IOPs: Phase function, extinction and scattering coefficients 8

9 Derive P from sky radiance measurements 1- Correct from multiple scattering: f = L L (1) theo L L (1) mes SO Code 2- Primary scattering approximation : P( Θ) = 4 (1) L mes τ 1 exp( ) 1 exp( τ ( µ µ o 1 ) µ 1 µ o µ µ o 1 Devaux et al., 1989 ωaτ apa ( Θ ) + τrpr ( Θ) P( Θ ) = τa+ τr 9

10 Multiple scattering correction Sucessive Order of Scattering Inputs : - total optical thickness τ - single scattering albedo ω time phase matrix M hypothesis : - Cox and Munk for wave slope distribution - inhomogeneous atmosphere H a = 3 km H r = 8 km 10

11 More to know Derive aerosol phase functions from sky radiance measurements in the frame of MERIS aerosol remote sensing validation Zagolsky Françis and Richard Santer (Université du Littoral) A new aerosol climatology for MERIS atmospheric correction over water using ground based measurements of the solar extinction and of the sky radiances Richard Santer and Zagolsky Françis (Université du Littoral) 11

12 The Aeronet data base 12

13 Data filtering Expected range of: Alpha Single scattering albedo case (i) α 1 ω1 ω2 ω1+ω2 N

14 Spectral dependancy 14

15 Single scattering albedo (ω a ) versus α 15

16 Phase function versus α Θ =90 o P a (Θ,675) (a) α Θ = 130 o P a (Θ,675) (b) α 16

17 aerosol phase function for mean value of α, 25 classes in α with α =0.2 Between 100 to 200 phase functions per class Filtering at 1 sigma 17

18 aerosol phase function:α =

19 ptheta land47 CIMEL-2.0 delta theta 19

20 Readjust AOT_865 We know from α the 2 SAMs used in MERIS GS We then know the phase functions used We assume than the aerosl path radiance in primary scattering is correctly retrieved We then get τ aϖ a 0 P a We correct the initial AOT_865 by Pcimel/ Pmeris τ IOP a = ( τaϖ Pa ) ( ϖ Pa ) a 0 a 0 SAM IOP 20

21 Readjust AOT_865 AOT at 865 nm MERIS CIMEL Série1 Série2 21

22 Perpectives Generate a significant IOPs data base at 865 nm, 670 nm and 440 nm Classify by alpha Generate LUTs Process level 2 with this new LUTs Evaluate improvements 22

23 23