Cloud screening and aerosol retrievals from MAX-DOAS measurements at Ukkel Clio Gielen Michel Van Roozendael, Francois Hendrick, Caroline Fayt, Christian.

Cloud screening and aerosol retrievals from MAX-DOAS measurements at Ukkel

Clio GielenMichel Van Roozendael, Francois Hendrick, Caroline Fayt,

Christian Hermans, Gaia Pinardi, Didier Gilllotay, and Tim Vlemmix

+ Hugo De Backer, Veerle De Bock, and Quentin Laffineur from RMI

BIRA contribution to WP3:

Goals

Qualify sky and cloud conditionsfrom MAX-DOAS data

Apply to Ukkel station to improve aerosol retrievals

Tasks 3.1 (algorithm improvements) and 3.2 (operating a MAX-DOAS instrument at Ukkel)

Sky conditions: the colour index

Trace gas and aerosol retrievals affected by clouds

→ introduce: additional photon absorption multiple scattering strong temporal variation Degenerate problem: thin/thick clouds, broken clouds, fog, haze, white clouds, grey clouds, high/low aerosol load, ....

Use observed colour of the sky for characterisation

→ Colour Index = ratio of intensities

Colour Index characteristics

Investigation of the colour index shows that:

- the CI = high for clear skies

- the CI < in the presence of clouds, aerosols,

fog,...

- the CI = a constant value over time in the

presence of

optically thick clouds

- the CI = highly variable in the presence of

broken/scattered clouds

Cloud-Screening flags

Using the colour index we define two flags:

- a flag to describe the general sky condition (clear/polluted/full-cloud cover)

→ Colour-Index flag

- a flag to note the presence of scattered clouds

→ Broken-Cloud flag

Broken-Cloud flag

Temporal variation to determine presence of scattered clouds

→ model observations with double-sine function and detect outliers→ no information on presence of full-cloud cover!

Colour-Index measurements: Ukkel

Colour-Index flag: sky condition

Good

Mediocre

Bad

By combining measurements and simulations we define 3 sky-condition regions

Good= clear sky

Med= thin clouds/ low

aerosols

Bad= thick clouds/ high

aerosols

Impact on aerosol retrievals

+ no cloud screening

+ CI-flag=good/med

+ CI-flag=good/med and no broken clouds

◊ AOD measurements (Aeronet + Brewer)

+ AOD retrievals (MAX-DOAS + bePRO profiling

code)

Impact on aerosol retrievals

+ no cloud screening

+ CI-flag=good/med

+ CI-flag=good/med and no broken clouds

Cloud-cover from thermic infrared data

Interpretation of aerosol data at Ukkel

Exploiting the ensemble of data at BIRA-RMI (task 3.3):

- BIRA: Aerosol AOD and extinction profiles (MAX-

DOAS) Aerosol AOD (CIMEL) Cloud screening (MAX-DOAS) Cloud-cover fraction (thermic infrared)

- RMI: Aerosol AOD (Brewer) Backscatter profiles/mixing-layer height

(lidar) Cloud screening (lidar/Brewer)+ Aethalometer in the future

Vertical Profiles: BIRA-RMI

Aerosol AOD (BIRA/RMI)

Backscatter/ cloud/MLH

(RMI)

Aerosol vert.prof+

cloudscreening(BIRA)

Aerosol vert.prof(BIRA)

Status

- Operational cloud-screening method for MAX-DOAS measurements

- Aerosol retrievals for full Ukkel dataset (2011-2013): AOD and extinction profiles

- Applying cloud screening strongly improves aerosol retrievals

- Good agreement between BIRA cloud screening and RMI data

- Good agreement between MAX-DOAS cloud screening and thermal infrared cloud-cover data

Future tasks for the last part of the project

- Further improvement of cloud-screening method: Include O4 DSCDs → distinguish

clouds/aerosols/fog... + cloud bottom height

- Detailed study of aerosol retrievals at Ukkel (BIRA + RMI):

Trends (seasonal/daily/weekly) Correlation with meteorology

- Estimation of the radiative forcing effect at Ukkel (task 3.3)

Colour-Index simulationsA

ero

sol lo

ad

Clo

ud

op

tica

l depth

Simulations performed by Tim Vlemmix, with DAK plan-parallel model