Discussions about scope and goals of the WG

Discussions about scope and goals of the WG

Cédric JametKick-off meeting

Wimereux, FranceMay, 14, 2014

Why a new IOCCG WG?

• Complement of the IOCCG report #10: « Atmospheric Correction for Remotely-Sensed Ocean-Colour Products » (Wang, 2010)Update (could also be an update of IOCCG report #3)

• This WG focused mainly on open ocean waters• And in coastal waters??• Need for guidances on using the already developed AC

Purpose of this new WG

For a sediment-dominated Case-2 water with typical maritime aerosols: all four operational algorithms produced slightly larger errors in the nLw in the blue

bands, mainly due to an incorrect estimation of the NIR ocean contributions nLw in the longer visible wavelengths (e.g. 555 nm), and radiance band ratio values

can be retrieved accurately. Since nLw spectra for the sediment-dominated waters highest in green-red λ:

satellite-derived nLw and radiance ratio at the longer visible bands used for retrieval of sediment-dominated water optical and biological properties.

For a yellow substance-dominated Case-2 water: all four operational algorithms performed poorly, particularly for nLw in the blue nLw in blue should beused to obtain CDOM absorption information (i.e., to

distinguish it from phytoplankton absorption), even though the blue radiance is difficult to derive accurately

Spectral shape of nLw ratios, could be helpful for deriving the water property for yellow substance dominated waters.

For absorbing aerosols: All four operational algorithms failed to produce accurate nLw spectra in visible λ Satellite derived nLw values significantly underestimated

For aerosol retrievals: AOT data over oceans derived accurately using the vector aerosol LUT (errors within

5%. Ocean nLw spectra can be derived accurately using the scalar aerosol LUT

Why a new IOCCG WG?

• Complement of the IOCCG report #10: « Atmospheric Correction for Remotely-Sensed Ocean-Colour Products » (Wang, 2010)Update (could also be an update of IOCCG report #3)

• This WG focused mainly on open ocean waters• And in coastal waters??• Need for guidances on using the already developed AC

Purpose of this new WG

Summary

• Goal: Inter-comparison and evaluation of existing AC algorithms over turbid/coastal waters

Understanding retrievals differences between algorithms• Challenge: to understand the advantages and limitations of each

algorithm and their performance under certain atmospheric and oceanic conditions

• Only focus on AC algorithms that deal with a non-zero NIR water-leaving radiances.

• High demand for AC guidelines• Outputs timely Guidances on the use of AC over turbid waters Recommendations for improving and selecting the optimal AC• Not a sensor-oriented exercise

Questions

• How well do the aerosols need to be retrieved?• Does the technique matter?• How good is the extrapolation from SWIR to NIR to VIS?• How to improve AC with the historic wavelengths?– Better bio-optical models– New contrains (such as relationships between Rrs, Goyens et

al. (2014))– Regional AC?

• Do we really need extra wavelengths in NIR/SWIR? What info can offer the future satellite sensors?

Evaluation of AC

• Can round robin lead to improvements?– What can we learn?

• Range of validity and advantages• Limitations

– Sensitivity studies• Fixed aerosols Variation/change of the bio-optical model• Fixed bio-optical model Variation/change of the aerosol models

– Uncertainties propagation and budget on the hypothesis• Ruddick et al. (2000)• Bayseian statistics for NN (Aires et al., 2004a, 2004b, 2004c)

– Uncertainties on the NN parameters (weights)– Uncertainties on the outputs

• Others ?

Evaluation of AC


• Drawbacks and advantages• Limitations


– Uncertainties propagation and budget on the hypothesis• Ruddick et al. (2000)• Neukermans et al., (2012)• Bayseian statistics for NN (Aires et al., 2004a, 2004b, 2004c)


Evaluation of AC


• Drawbacks and advantages• Limitations


– Uncertainties propagation and budget on the hypothesis• Ruddick et al. (2000)• Neukermans et al., (2012)• Bayseian statistics for NN (Aires et al., 2004a, 2004b, 2004c)


Which sensors• SeaWiFS• MODIS-AQUA• MERIS• VIIRS• GOCI• ????

Which sensors• SeaWiFS• MODIS-AQUA for starting• MERIS• VIIRS• GOCI• ????

Which sensors• SeaWiFS• MODIS-AQUA for starting• MERIS after• VIIRS• GOCI• ????

many approaches exist, here are a few examples:• assign aerosols () and/or water contributions (Rrs(NIR)) e.g., Hu et al. 2000, Ruddick et al. 2000• use shortwave infrared bands e.g., Wang & Shi 2007• correct/model the non-negligible Rrs(NIR) Lavender et al. 2005 MERIS Bailey et al. 2010 used in SeaWiFS Reprocessing 2010 Shanmugam, 2012 any sensor Wang et al. 2012 GOCI• use a coupled ocean-atmosphere optimization e.g., Moore et al. (1999); Chomko & Gordon 2001, Stamnes et al. 2003, Jamet et al., 2005, Brajard et al., 2006a, b, 2008, 2012; Ahn and Shanmugam, 2007; Kuchinke et al. 2009; Steinmetz et al., 2010; • Other e.g., Chen et al., 2014; Doerrfer et al., (2007); He et al., 2013; Mao et al., 2013, 2014; Schroeder et al. (2007); Shanmugam and Tholkapiyan, 2014; Singh and Shanmugam, 2014

Which algorithms?

many approaches exist, here are a few examples:• assign aerosols () and/or water contributions (Rrs(NIR)) e.g., Hu et al. 2000, Ruddick et al. 2000• use shortwave infrared bands e.g., Wang & Shi 2007• correct/model the non-negligible Rrs(NIR) Lavender et al. 2005 MERIS Bailey et al. 2010 used in SeaWiFS Reprocessing 2010 Shanmugam, 2012 any sensor Wang et al. 2012 GOCI• use a coupled ocean-atmosphere optimization e.g., Moore et al. (1999); Chomko & Gordon 2001, Stamnes et al. 2003, Jamet et al., 2005, Brajard et al., 2006a, b, 2008, 2012; Ahn and Shanmugam, 2007; Kuchinke et al. 2009; Steinmetz et al., 2010; • Other e.g., Chen et al., 2014; Doerrfer et al., (2007); He et al., 2013; Mao et al., 2013, 2014; Schroeder et al. (2007); Shanmugam and Tholkapiyan, 2014; Singh and Shanmugam, 2014

Which algorithms?

All published/available atmospheric correction algorithms?

Selected only few depending on their hypothesis: (Hu et al. (2000) ~= Ruddick et al. (2000), is it necessary to take both?

Send datasets to developers and send their results back?

Every algorithm (when possible) implemented in SeaDAS so the entire processing chain is similar (differences in results only from AC) Is it possible/reasonable?

Which algorithms?

All published/available atmospheric correction algorithms?YES, WHEN POSSIBLE

Selected only few depending on their hypothesis: (Hu et al. (2000) ~= Ruddick et al. (2000), is it necessary to take both?

Email to developers to ask their participation to share their algorithms: natural selection?

Send datasets to developers and send their results back?YES

Every algorithm (when possible) implemented in SeaDAS so the entire processing chain is similar (differences in results only from AC) Is it possible/reasonable?

ACTION TO SEAN BAILEY for SeaDAS and C. Brockmann for BEAM (C. Jamet)

Which algorithms?

• Remote-Sensing reflectance (output of SeaDAS): Rrs• Normalized water-leaving reflectance: nρw• Normalized water-leaving radiance: nLw• Water-leaving radiance: Lw• Water Radiance-Luminance: RLw

• Do all algorithms take the same definition?

Which parameters?

• Remote-Sensing reflectance (output of SeaDAS): Rrs• Normalized water-leaving reflectance: nρw• Normalized water-leaving radiance: nLw• Water-leaving radiance: Lw• Water Radiance-Luminance: RLw

• Do all algorithms take the same definition?

ACTION: Table with all the possible AC algorithms with inputs and outputs (C. Jamet)

Which parameters?

Datasets (1/2)

• Classic match-ups analysis:– Which in-situ datasets:• AERONET-OC: moderately turbid waters• LOG/MUMM TriOS: moderately to very turbid waters• ASD measurements for VITO?• NOMAD, SeaBASS?• Data in estuaries?• Other?• Need for high-quality NIR nLw measurements

Datasets (1/2)• Classic match-ups analysis:– Which in-situ datasets:

• AERONET-OC: moderately turbid waters• LOG/MUMM TriOS: moderately to very turbid waters• ASD measurements for VITO• NOMAD, SeaBASS from NASA• DATA from CSIRO• DATA from X. He• DATA from P. Shanmugam

• ACTION: Ask PI of MERMAID in-situ data to share with WG

• ACTION TO K. Ruddick for his datasets (C. Jamet)

• Need for high-quality NIR nLw measurementsCOMPARISON OF NIR(nLw)

Datasets (1/2)

• What is a match-up?• Bailey and Werdell (2006)

–Δt= +/- 2h?–Spatial homogeneity criteria: std/mean on nLw or

on tau?–Spatial homogeneity criteria: valid if <0.15 or 0.20?–3x3 pixels box?–All valid pixels or a certain number (>5 or 6)?

Datasets (2/2)

• Image analysis over selected areas:– French Guiana, Eastern English Channel/North Sea,

Vietnam, Yangtze coasts, Amazon river????– Comparison of aerosol properties and nLw patterns– Transects from coasts to case-1 waters

Datasets (2/2)• Image analysis over selected areas:– French Guiana, Eastern English Channel/North Sea, Vietnam,

Yangtze coasts, Amazon river????

Each participant provides info about their regions of expertise to C. Jamet

Careful attention to sensor saturation: NOT AN ISSUE FOR THIS WG

– Comparison of aerosol properties and nLw patternsYES

– Transects from coasts to case-1 waters YES

Inter-comparison (1/3)• Optical water type:– Classification of nLw as a function of IOPs, contents– Sensitivities studies– Moore et al. (2009), Vantrepotte et al. (2012)

• Necessity to validate NIR nLw– Only validation of visible nLw– BUT goal of AC is to remove atmosphere from NIR

bands (most of the time)• Other way to validate– Determination aerosol optical properties– Use of AOP to estimate path reflectance via RTE– Reconstruction of Lrc using Lw + reconstructed

Lpath– Comparison of reconstructed Lrc to measured Lrc

Classification of Lw spectra per water type- 4 water type classes defined by Vantrepotte et al. (2012)

- Novelty detection technique:Assigns each spectra to one of the 4 water type classes using the Mahalanobis distance

METHODS

Focus on turbid waters only !

Distinguish classes based on normalized reflectance spectra

D'Alimonte et al. (2003)

Inter-comparison (2/3)• Optical water type:– Classification of nLw as a function of IOPs, contents– Sensitivities studies– Moore et al. (2011), Vantrepotte et al. (2012)

• Necessity to validate NIR nLw– Only validation of visible nLw– BUT goal of AC is to remove atmosphere from NIR bands (most

of the time)• Other way to validate– Determination aerosol optical properties– Use of AOP to estimate path reflectance via RTE– Reconstruction of Lrc using Lw + reconstructed Lpath– Comparison of reconstructed Lrc to measured Lrc

Inter-comparison (3/4)• Impact of observation angles:– Are AC sensitive to some observation angles?– CCI round-robin: YES– How do we do that? Images and/or sensitivity studies?

• Sensitivities studies:– Need for simulated datasets:

• Atmospheric RTE: Stamnes et al.; He et al.; other?– Which aerosol models? Shettle and Fenn, AERONET-like, ot

• Bio-optical model: Hydrolight How to fix the IOPs/oceanic constituents COASTCOLOUR + IOCCG round robin

• Fixed aerosol/varying IOPs• Varying aerosols/fixed IOPs• What do we simulate: L(TOA) or just L(rc)?

– If L(TOA), how do we simulate Lr, Lwc, Lg??





• Fixed aerosol/varying IOPs• Varying aerosols/fixed IOPs• What do we simulate: L(TOA) or just L(rc)?

– If L(TOA), how do we simulate Lr, Lwc, Lg??

ACTION: Definition of a set of few cases as benchmark to validate FUB RTM, C-VDISORT and PCOART(K. Stamnes, T. Schroeder and X. He)

Choice of the RTMTIMELINE: before next Ocean OpticsGeneration of atmosphere and ocean LUTs: before end of the year





• Fixed aerosol/varying IOPs• Varying aerosols/fixed IOPs• What do we simulate: L(TOA) or just L(rc)?– If L(TOA), how do we simulate Lr, Lwc, Lg??–Simulation of L(TOA) AND L(rc) with La, Lra, Lw and associated

transmittances (L and ρ)

Inter-comparison (4/4)

• Time series over selected regions– Correlation bt aerosols, IOPs, Rrs, ….

• Definition of objective criteria for comparison: – Need for metrics

• Validation of aerosol optical properties– Do we want accurate AOP of just accurate Lpath?– Is it important?– What info does it give us on the accuracy of AC?

• Vicarious calibration

Inter-comparison (4/4)• Time series over selected regions

– Correlation bt aerosols, IOPs, Rrs, ….• Definition of objective criteria for comparison:

– Need for metric– Which statistical parameters:

» RMSE» RELATIVE ERROR» Slope» Coefficient correlation» Bias

ACTION TO DAGMAR MULLER FOR EXPERIENCE FROM CCI (C. Jamet)



Inter-comparison (4/4)• Time series over selected regions– Correlation bt aerosols, IOPs, Rrs, ….

• Definition of objective criteria for comparison: – Need for metrics? (CCI)– Is it necessary?



Keep in mindACTION to S. Bailey for more inputs and possibility of vicarious

calibrated every AC algorithm

What next

• October 2014:– Ocean Optics conference: possibility to book a meeting

space 6 month progress– Who can attend?

• 2015:– Development/gathering synthetical datasets– Match-ups and transect analysis

• 2016:– Match-ups and transect analysis– Sensitivity analysis using synthetical datasets

What next• October 2014:

– Ocean Optics conference: possibility to book a meeting space 6 month progress

• Results of RTM comparison• Collection of in-situ datasets• Choice of AC algorithms• Tables on AC algorithms (inputs, outputs)• Choice of regions of interest

• 2015:– Development/gathering synthetical datasets– Match-ups and transect analysis

• 2016:– Match-ups and transect analysis– Sensitivity analysis using synthetical datasets

Documents

Discussions about scope and goals of the WG