The Phase III questionnaire is available in the following Google spreadsheet:

https://docs.google.com/spreadsheets/d/1pFq6f7gP8nQLz6IYLqMtCPQN4t0eVe2Xd3R3_ra4pg4/

If you have problems accessing this spreadsheet, please contact jonasg@met.no or michaels@met.no. The spreadsheet is also available as .ods file in the supplementary material of Gliß et al, 2020 (https://www.atmos-chem-phys-discuss.net/acp-2019-1214/).

Questions on optical property calculations in AeroCom phase II models with the purpose of a correct interpretation of diagnosed AOD properties in the AeroCom database:

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

2b) If you differentiate, have you investigated the difference?

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

6) Which humidity growth factor and variation do you use ? How does it depend on RH? Do you include sub-grid variability in relative humidity?

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

9) Do you average nighttime and daytime values of AOD into daily averages?

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

11) What is the absorption coefficient of BC? How is it calculated?

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species.

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

50% BC from fossil fuel is internally mixed with POM. All BCs from the other sources are internally mixed with POM. The other aerosol species are completely in external mixture. Size of the internally-mixed BC is according to that of POM. It is assumed to mix homogeneously.

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

The relative humidity is available only for the all-sky.

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

No. AOD is calculated at the specific wavelengths. The broadband information is essential to calculate the RF. The aerosol mass mixing ratio is transferred to the radiation code, then the RF is calculated in it.

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

Yes.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

The clear-sky AOD can be calculated when the 2D cloud fraction is under 0.2 in each time step. If the cloud fraction is over 0.2, the clear-sky AOD is a missing value. The time-averaged AOD for the clear-sky is estimated by excluding the missing value.

6) Which humidity growth factor and variation do you use? How does it depend on RH? Do you include sub-grid variability in relative humidity?

The growth factors are according to Tang and Munkelwitz (1994) for sulfate, Hobbs et al. (1997) for POM, and Shettle and Fenn (1979) for sea salt.

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

The aerosol water mass is diagnosed.

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

Yes. The all-sky RH is used.

9) Do you average nighttime and daytime values of AOD into daily averages?

Yes.

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

At 550nm: 0.0165 for BC; 0.958 in 0%RH, 0.967 in 50%RH, 0.969 in 70%RH, 0.987 in 80%RH, 0.993 in 90%RH, 0.995 in 95%RH, 0.998 in 98%RH, 0.999 in 99%RH for POM; 0.987 in 0.13µm, 0.962 in 0.33µm, 0.987 in 0.82µm, 0.954 in 1.27µm, 0.983 in 3.20µm, 0.975 in 8.02µm for DUST.

11) What is the absorption coefficient of BC? How is it calculated?

2.774 m2/g at 550nm with the Mie theory (radius: 0.0118µm, standard deviation: 2.0, density: 2.3 g/cm3 with log-normal size distribution).

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species.

At 550nm: (1.430,1e-8) for sulfate; (1.381,4.26e-9) for sea salt; (1.530,6e-3) for POM; (1.750,0.440) for BC; (1.530,2e-3) for dust; The homogeneous mixture with water is assumed in the hygroscopicity with the refractive index of (1.333,1.96e9) for water.

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

We assume that the hydrophilic BC is internal mixed. We assume the core shell type of mixing and an increase in absorption by 50% based on Bond et al. (2006)

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

All-sky RH is used and reported

2b) If you differentiate, have you investigated the difference?

Relatively small difference is found in sensitivity calculations between all-sky and clear-sky AOD of less than 10% over land. The clear-sky AOD is calculated from the grid-box mean relative humidity assuming saturation in the cloudy-fraction.

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

Yes

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

Yes, see 2b

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

See 2b). No threshold and daily data are generated.

6) Which humidity growth factor and variation do you use ? How does it depend on RH? Do you include sub-grid variability in relative humidity?

The humidity growth is parameterized based on Fitzgerald (1975) and for sulphate tested that this is very similar to simulations based on the Kohler equation (see Myhre et al., 2004). No sub-grid relative humidity is implemented in the AeroCom simulations for OsloCTM2

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

No

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

Yes at ambient RH and all-sky RH

9) Do you average nighttime and daytime values of AOD into daily averages?

Yes, AOD is an average of night-time and daytime values.

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

BC has a single-scattering albedo at 550 nm of 0.21 which is not dependent on size or RH POM has a single-scattering albedo at 550 nm of 1.0 which is not dependent on size or RH DUST has a single-scattering albedo at 550 nm of around 0.97, but it is strongly dependent on size (from 0.72 – 0.99 for the 8 size bins). No dependence on RH.

11) What is the absorption coefficient of BC? How is it calculated?

The absorption coefficient of BC is 7.3 m2/g for hydrophobic BC and 50% for hydrophilic. Mie theory is used in the calculation of the optical properties (see Myhre et al., 2007)

1) Do you assume and how internal mixing when computing optical properties of total aerosol?

4 out of 7 modes are internally mixed

Is Size varying?

Yes, variation of median radius for each mode accounted in calculation of optical properties

Which mixing assumption?

Submitted is volume weighted mixing of refractive indices. Differences owing to alternative mixing rules explored in Stier et al., ACP (2007) were negligible as compared to the uncertainty in BC refractive indices.

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

Clear-sky relative humidity, calculated from the grid-box mean assuming saturation in the cloudy-fraction, is used in the AOD calculation.

2b) If you differentiate, have you investigated the difference?

Yes, long ago…

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

Yes, online in the model.

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

Well, in as much as satellite retrievals are able to separate clear-skies. Ultimately, we will need to work with satellite simulators to answer this question.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use?

See above.

Cloud fraction weighting to derive a monthly average Clear-sky AOD?

“Clear-sky” AOD is reported even for fully cloudy cells. In this case the grid-box mean RH is at saturation which is then used in the water uptake calculation. It would be good to sample better.

6) Which humidity growth factor and variation do you use ?

Kappa-Koehler theory.

How does it depend on RH?

Obvious.

Do you include sub-grid variability in relative humidity?

No.

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

Mass based on Kappa-Koehler, aerosol water AOD based on volume weighted attribution of AOD to the components (this works well for AOD, but component specific AAOD cannot be accurately computed).

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

All based on RH as above.

9) Do you average nighttime and daytime values of AOD into daily averages?

Yes. We also have satellite overpass sampled AOD but this reduces sampling strongly for single year runs.

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

Calculated from Mie theory based on volume weighted mixing of refractive indices and mode size. Details in Stier et al., ACP, (2007).

11) What is the absorption coefficient of BC? How is it calculated?

Not directly calculated. Can be diagnosed as for Phase I.

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

External mixing only, size distributions prescribed for each aerosol component.

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

The model differentiates between clear-sky and all-sky relative humidity, and AOD calculations use the clear-sky RH.

2b) If you differentiate, have you investigated the difference?

Quickly (unpublished).

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

No. Radiative flux calculations use waveband-averaged specific scattering and absorbing coefficients, and aerosol mass mixing ratios. AODs are diagnostic only.

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

Yes, diagnosed AODs represent clear-sky conditions.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

Clear-sky AOD uses clear-sky RH, with no threshold on RH (i.e. calculations go up to 100% RH). Cloud fraction is only used when computing clear-sky RH (assuming cloudy-sky specific humidity is at qsat).

6) Which humidity growth factor and variation do you use ? How does it depend on RH? Do you include sub-grid variability in relative humidity?

Hygroscopic growth is parametrized as a function of RH following Fitzgerald [1975]. RH is gridbox mean only, although q = qsat is assumed in cloudy sky when computing clear-sky RH.

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

Aerosol water AOD is diagnosed as the difference between standard AOD and the AOD computed without accounting for hygroscopic growth (i.e. aerosol size and refractive index remain those of the dry aerosol).

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

Use clear-sky RH computed from ambient gridbox-mean RH and cloud fraction.

9) Do you average nighttime and daytime values of AOD into daily averages?

Yes.

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

Please refer to Table 4 of Bellouin et al. (2011)

11) What is the absorption coefficient of BC? How is it calculated?

For FFBC, absorption coefficient is 3.2 m2g-1 at 0.55 microns. Computed using Mie theory from prescribed size distribution and refractive index.

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species.

Please refer to Table 4 of Bellouin et al. (2011), and the Appendix of the manuscript for details on aerosol optical properties in HadGEM.

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

External mixing.

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

No. We use ambient relative humidity.

2b) If you differentiate, have you investigated the difference?

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

The AOD reported to AeroCom is 550 nm. For RF calculations, we use the SW spectral dependence of AOD and other properties (SSA, Asymmetry factor) to calculate RF.

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

Not necessary - but we are doing it anyway. We don't have quantitative numbers to show on monthly averaged basis how much the differences are.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

N/A

6) Which humidity growth factor and variation do you use ? How does it depend on RH? Do you include sub-grid variability in relative humidity?

Based on Tang and Munkelwitz (1996) water activity formula determined at the HeHe laser wavelength of 0.633 microns for ammonium nitrate and ammonium sulfate. All others based on GADS (aka OPAC). No we don't include sub-grid variability of RH.

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

Not for AeroCom II aerosol direct radiative forcing activity (not required). Yes for AeroCom HTAP activity. Aerosol water = ambient AOD - dry AOD (i.e., no hygroscopic growth).

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

Yes they are computed at ambient RH under all sky conditions.

9) Do you average nighttime and daytime values of AOD into daily averages?

Yes.

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

At 550 nm: SSA for BC: 0.2080, 0.2086, 0.2139, 0.2481, 0.3227, 0.3830, 0.4594, 0.5359 at 0, 50, 70, 80, 90, 95, 98, 99% RH.

SSA for POM: 0.9618, 0.9773, 0.9819, 0.9851, 0.9897, 0.9932, 0.9960, 0.9972 at 0, 50, 70, 80, 90, 95, 98, 99% RH.

SSA for dust: 0.9916, 0.9909, 0.9864, 0.9767, 0.9604, 0.9353, 0.8933, 0.8399 for dust effective radius of 0.14, 0.24, 0.45, 0.80, 1.4, 2.4, 4.5, 8.0 um. No hygroscopic growth is considered.

11) What is the absorption coefficient of BC? How is it calculated?

I assume you mean “mass absorption efficiency” with a unit of m2/g. This is calculated as MAE=MEE*(1-SSA).

At 550 nm: MAE for BC: 7.3529, 7.4091, 0.4491, 8.5266, 8.0192, 9.7856, 10.190, 10.521 m2/g at 0, 50, 70, 80, 90, 95, 98, 99% RH.

They are calculated from the Mie code with refractive indices (BC/water volume weighted), size distributions, and particle density.

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species.

BC: 1.75 - 0.44i; POM: 1.53-0.006i; (NH4)2SO4: 1.53-10^(-7)i; NH4NO3: 1.556-1.1^(-9)i; Dust: 1.53-0.0055i; Seasalt: 1.50-10^(-8)i

The refractive indices of ammonium sulfate and ammonium nitrate are calculated using lacis-refrac.rhwmri.f provided by Andrew Lacis [ftp://ftp.giss.nasa.gov/pub/crmim/lacis/lacis_refrac.rhwmri.f]. All others see Table 2 in Chin et al., Ann. Geophys., (2009).

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

Aerosol are externally mixed, size distributions are prescribed

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

We do not differentiate between clear sky and all sky humidity, but we provide both, clear and all sky AOD for AeroCOm.

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

Yes

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

We provided both clear and all sky.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

Clear sky is defined when the Monte Carlo sampling of cloud fraction sees zero cloud fraction in the column.

6) Which humidity growth factor and variation do you use? How does it depend on RH?

Sea salt, nitrate and sulfate get humidified following Lacis and Oinas 1991 and depends on ambient RH

Do you include sub-grid variability in relative humidity?

No

We use different parameterizations for sulfate and sea salt.

Yes

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

No

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

Yes we use ambient all sky RH.

9) Do you average nighttime and daytime values of AOD into daily averages?

Yes

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

11) What is the absorption coefficient of BC? How is it calculated?

BC absorption is calculated based on BC mass, an BC enhancement factor of 1.5 is applied to take into account BC enhancement effect due to internal mixing.

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species.

The following refractive indices at 550nm are used in this study: Sulfate (1.528-1.e-7i), nitrate (1.528-1.e-7i), OC (1.527-0.014i), BC (1.85-0.71i), sea salt (1.45-0.i), dust (1.564-0.002i) and water (1.334-3.91e-8i).

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

Mixing state is taken into consideration. Particles including BC have core shell structure; other particles use volume mixing approach. The size is prognostic and the mixing state assumption follows the population definitions in Bauer et al 2008 ACP

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

We do not differentiate between clear sky and all sky humidity, but we provide both, clear and all sky AOD for AeroCOm.

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

Yes

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

We provided both clear and all sky.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

Clear sky is defined when the Monte Carlo sampling of cloud fraction sees zero cloud fraction in the column.

6) Which humidity growth factor and variation do you use ? How does it depend on RH? Do you include sub-grid variability in relative humidity?

We calculate the uptake of water following the thermodynamical model EQSAM and for Sea salt using the Lewis parameterization (Lewis and Schwartz 2004)

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

No, aerosol water is one component that determines the mixing state of aerosols. Aerosol water is tracked for each aerosol population.

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

Yes we use ambient all sky RH.

9) Do you average nighttime and daytime values of AOD into daily averages?

Yes

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

11) What is the absorption coefficient of BC? How is it calculated?

BC absorption is calculated depending on aerosol mixing state – see question 1)

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species.

The following refractive indices at 550nm are used: Sulfate (1.528-1.e-7i), nitrate (1.528-1.e-7i), OC (1.527-0.014i), BC (1.85-0.71i), sea salt (1.45-0.i), dust (1.564-0.002i) and water (1.334-3.91e-8i).

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption? Reply: No, all species of aerosols are external mixture. The size spectrum of each aerosol is divided into 12 bins with radii.

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom? Reply: No, we don’t differentiate between clear-sky and all-sky relative humidity. We use all-sky relative humidity for the AOD calculation.

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations? Reply: Yes, all-sky AOD is used in the RF calculations.

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies? Reply: Yes.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD? Reply: Cloud fraction weighting is used to derive a clear-sky AOD ((1-CLDTOT)*OD550AER), but with no explicit RH screening.

6) Which humidity growth factor and variation do you use? How does it depend on RH? Do you include sub-grid variability in relative humidity? Reply: Kohler theory is used to calculate the humidity growth. We don't include sub-grid variability in relative humidity.

7) Do you diagnose the aerosol water mass or aerosol water AOD? How? Reply: No.

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH? Reply: Yes. The all-sky RH is used.

9) Do you average nighttime and daytime values of AOD into daily averages? Reply: Yes.

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range? Reply: The single-scattering albedo is calculated from Mie theory. At 550 nm, the single-scattering albedos are 0.00118 - 0.575 for BC, 0.0233 - 0.999 for POM, and 0.0355 - 0.975 for dust. They vary with particle size and RH.

11) What is the absorption coefficient of BC? How is it calculated? Reply: The absorption coefficient of BC is 5.11, 5.30, 6.18, 9.34, 10.7, 6.11, 2.72, 1.26, 0.600, 0.289, 0.141, and 0.0695 m2/g in 12 bins at 550 nm. It's calculated using Mie-theory.

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species. Reply: Refractive index at 0.55 micron is (1.75, 0.44), (1.53, 0.0059), (1.53, 0.008), (1.43, 1.e-8), (1.5, 9.7e-9) for BC, POM, dust, sulfate, and sea salt, respectively. Kohler theory is used to calculate the aerosol hygroscopicity.

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

Internal vs. external mixing is determined on aerosol process level: coagulation, condensation, and cloud processing gives internal mixing with pre-existing particles, which grow in size due to these processes. For mixing of absorbing and transparent constituent the Maxwell-Garnett mixing rule is applied, otherwise volume mixing.

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

With respect to aerosol treatment, no. Both OD550_AER (all-sky AOD) and OD550CS_AER (clear-sky AOD) are calculated at ambient relative humidities (RH), but OD550CS_AER is a clear-sky fraction (=1-CLDTOT) weighted AOD.

2b) If you differentiate, have you investigated the difference?

For CAM4-Oslo the two are found to differ quite significantly, especially at mid to high latitudes. Negligible differences in areas with low RH. Globally averaged, the relative difference is 16% (clear-sky 0.128 vs. all-sky 0.153).

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

All-sky AOD is used in the RF-calculations*. The only RH limitation is an upper cut-off value of RH=99.5% in the optics look-up tables. *More precicely, spectrally resolved 3-dimensional extinction coefficient, single scattering albedo and asymmetry parameter are used in the radiative transfer code.

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

Since AERONET uses cloud-screening, I would consider some kind of clear-sky AOD estimate to be better than the all-sky AOD.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

OD550CS_AER is a clear-sky fraction (=1-CLDTOT) weighted OD550_AER, but with no explicit RH screening.

6) Which humidity growth factor and variation do you use ? How does it depend on RH? Do you include sub-grid variability in relative humidity?

Growth factors are as in described Seland et al. (2008) and references therein. Linear growth factors are applied in the hysteresis regime for sea-salt and ammonium sulfate. For internally mixed aerosols the growth factor is calculated from the Kohler theory, taking the hygroscopicity of each mixed constituent into account. No sub-grid variability included.

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

Yes, both. AOD: od550aerh2o is calculated as OD550_AER (ambient RH) - OD550_AER (RH=0%). Water mass: mmraerh2o is the mixing ratio for aerosol-water (from hygroscopic growth).

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

Yes, using all-sky assumptions (for RH and cloud cover).

9) Do you average nighttime and daytime values of AOD into daily averages?

Yes, 24h AOD values are used.

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

For a given wavelength, SSA varies both with size (distribution) and state of mixing with other aerosol constituents and condensed water (due to RH). SSA at 550nm for dry and pure aerosol components: BC nucleation-mode (compact monomers) 0.124; BC accumulation-mode (fluffy agglomorate) 0.354; POM Aitken mode 0.960; DUST accumulation-mode 0.981; DUST coarse-mode 0.900.

11) What is the absorption coefficient of BC? How is it calculated?

It's calculated using Mie-theory and Maxwell-garnett mixing for refractive index when internally mixted. It varies with size and state of mixing (also with RH when internally mixed). For pure (not internally mixed) particles: BC nucleation mode (compact monomers): 6.01 m2/g, BC accumulation mode (fluffy agglomorate): 5.27 m2/g.

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species.

Refractive index at 0.55 micron is (2.0,1.0), (1.53,0.006), (1.56,0.0025), (1.5,4.1.e-8), (1.43,1.e-8) for BC (with monomer density), POM, dust, sea-salt, sulfate, respectively. A more general description and further references concerning refractive indices and hygroscopicities are found in section 2.8 of Seland et al. (2008).

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption? internal mixing within each of 3 modes. size varies with mass/number. volume mixing of refractive indices of components within mode.

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom? No.

2b) If you differentiate, have you investigated the difference?

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations? Yes.

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies? Sort of. The grid cell mean RH is used, which is an approximation. Caveat: we only treat optics of interstitial aerosol.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

6) Which humidity growth factor and variation do you use ? How does it depend on RH? Do you include sub-grid variability in relative humidity? Kappa Kohler theory using volume mean kappa. Dry if RH<RH_crystalization. Wet if RH>RH_deliquescence. Linear in RH between. No subgrid variability in RH.

7) Do you diagnose the aerosol water mass or aerosol water AOD? How? Aerosol water from volume mean wet radius.

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? Not sure what is meant by ambient RH. And eventually is that using clear-sky or all-sky RH? All-sky RH is used.

9) Do you average nighttime and daytime values of AOD into daily averages? No. Day only.

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range? Properties diagnosed from volume mean refractive index and wet surface mode radius.

11) What is the absorption coefficient of BC? How is it calculated? Diagnosed from Mie theory assuming refractive index of (1.95,0.79). Depends on size. Emitted with 0.08 micron number mode radius and 1.8 geometric standard deviation.

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species. Refractive index at 0.55 micron is (1.95,0.79), (1.53,0.005665), (1.53,0.00627), (1.53,0.005665), (1.43,1.e-8), (1.49,1.e-8) for BC, POM, dust, SOA, sulfate, respectively. Hygroscopocity is 0, 0.10, 0.068, 0.14, 0.507, 1.16 for BC, POM, dust, SOA, sulfate, sea salt, respectively.

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

External mixing.

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

No. We use ambient relative humidity.

2b) If you differentiate, have you investigated the difference?

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

The AOD reported to AeroCom is 550 nm. For RF calculations, we use the SW spectral dependence of AOD and other properties (SSA, Asymmetry factor) to calculate RF.

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

Not necessary - but we are doing it anyway. We don't have quantitative numbers to show on monthly averaged basis how much the differences are.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

N/A

6) Which humidity growth factor and variation do you use ? How does it depend on RH? Do you include sub-grid variability in relative humidity?

Based on GADS (aka OPAC). No we don't include sub-grid variability of RH.

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

Yes. Aerosol water = ambient AOD - dry AOD (i.e., no hygroscopic growth).

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

Yes they are computed at ambient RH under all sky conditions.

9) Do you average nighttime and daytime values of AOD into daily averages?

Yes.

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

At 550 nm: SSA for BC: 0.2080, 0.2086, 0.2139, 0.2481, 0.3227, 0.3830, 0.4594, 0.5359 at 0, 50, 70, 80, 90, 95, 98, 99% RH.

SSA for POM: 0.9618, 0.9773, 0.9819, 0.9851, 0.9897, 0.9932, 0.9960, 0.9972 at 0, 50, 70, 80, 90, 95, 98, 99% RH.

SSA for dust: 0.9675, 0.9652, 0.9507, 0.9196, 0.8717, 0.8107, 0.7314, 0.6601 for dust effective radius of 0.14, 0.24, 0.45, 0.80, 1.4, 2.4, 4.5, 8.0 um. No hygroscopic growth is considered.

11) What is the absorption coefficient of BC? How is it calculated?

I assume you mean “mass absorption efficiency” with a unit of m2/g. This is calculated as MAE=MEE*(1-SSA).

At 550 nm: MAE for BC: 7.3529, 7.4091, 0.4491, 8.5266, 8.0192, 9.7856, 10.190, 10.521 m2/g at 0, 50, 70, 80, 90, 95, 98, 99% RH.

They are calculated from the Mie code with refractive indices (BC/water volume weighted), size distributions, and particle density.

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species.

BC: 1.75 - 0.44i POM: 1.53-0.006i SO4: 1.43-10^(-8)i Dust: 1.53-0.0055i Seasalt: 1.50-10^(-8)i (see Table 2 in Chin et al., Ann. Geophys., 2009)

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

Five out of seven modes are internally mixed. Volume weighted mixing of refractive indices within each mode, using non-linear mixing rules. Yes, the actual median radius for each mode is taken to infer optical properties. Bruggeman and Maxwell-Garnett (BC and dust are treated as inclusions, if applicable).

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

The model differentiates between clear-sky and all-sky RH. See 5). All-sky RH is reported to AeroCom, clear-sky RH is used for AOD calculation.

2b) If you differentiate, have you investigated the difference?

No.

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

No, reported-to-AeroCom AOD is daily means, while (offline) RF calculations use local noon-time AOD.

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

Yes.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

• All-sky RH (RHa) is bounded to [0.0-0.999].
• For grid cells with RHa > 0.9, the clear-sky RH (RHc) is derived by scaling RHa with cloud cover fraction for which RH is at saturation value. RHc is bounded to [0.75, 0.999].
• For grid cells with RHa ⇐ 0.9, RHc equals RHa.
• RHc is then used for water uptake and subsequent AOD calculations.

6) Which humidity growth factor and variation do you use ? How does it depend on RH? Do you include sub-grid variability in relative humidity?

The soluble particles are assumed to be in equilibrium with water vapor. Only sulphate and sea salt are influencing the water uptake, as described in Vignati et al. (JGR, 2004). No sub-grid variability in RH.

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

Aerosol water mass is diagnosed from hygroscopic growth. Aerosol water AOD is the difference between total AOD and the AOD inferred using dry radii and dry internal mixtures.

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

Clear-sky scaling of grid-box mean RH, as described above.

9) Do you average nighttime and daytime values of AOD into daily averages?

Yes.

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

For internally mixed modes: volume weighted mixtures of individual refractive indices based on Mie theory. Pure, dry aerosols at 550nm, d=particle diameter:

• BC: [0.172,0.482,0.563] for d=[0.05,0.5,5.0] microns (Aitken/Accum/Coarse), respectively
• POM: [0.973,0.991,0.924] for d=[0.05,0.5,5.0] microns (Aitken/Accum/Coarse), respectively
• DUST: [0.991,0.924] for d=[0.5,5.0] microns (Accum/Coarse), respectively

11) What is the absorption coefficient of BC? How is it calculated?

The absorption coefficient of internal mixtures is diagnosed from extinction coefficient and SSA. For pure, dry BC at 550nm, d=particle diameter: [9.198,3.205,0.364] m2/g for d=[0.05,0.5,5.0] microns (Aitken/Accum/Coarse), respectively.

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species.

• SO4: (1.43,1.E-08i)
• BC: (1.75,0.44i)
• SS: (1.50,1.E-08i)
• POM: (1.53,0.0055i)
• DUST: (1.52,0.0011i)

Please see 6).

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

We compute optical properties over six “optical” species externally mixed. The species are: inorganic ions (sulfate + nitrate + ammonium), organic carbon (primary and secondary OC), blank carbon, sea salts, and soil dust (4 size bins). A size distribution (log-normal for all but dust, for which we assume gamma-distributions in the 4 size bins). The size distribution varies by hygroscopic growth.

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

No.

2b) If you differentiate, have you investigated the difference?

—

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

We didn’t carried out RF calculations so far in this frame.

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

Tricky question. The AOD is diagnosed for all-sky conditions, but the RH used for hygroscopic growth is capped at 90% to avoid strong non-linearities within the clouds. However, no cloud fraction filter is applied for AOD diagnostic. How is the impact of such a filter is an open question.

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

—

6) Which humidity growth factor and variation do you use ? How does it depend on RH? Do you include sub-grid variability in relative humidity?

We use growth factors given in GADS/OPAC database (M. Hess, P. Koepke, and I. Schult (1998): Optical Properties of Aerosols and clouds: The software package OPAC, Bull. Am. Met. Soc., 79, 831-844). We do not account for sub-grid variability of RH.

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

No.

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

Yes, ambient RH is used to scale AOD of all species. We do not diagnose clear-sky AOD.

9) Do you average nighttime and daytime values of AOD into daily averages?

Yes.

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

BC and POM single-scattering albedo (SSA) are calculated with Mie theory for spherical particles using the refractive indices given in GADS/OPAC database (Hess et al., 1998). They vary as a function of RH and wavelength. At 0.55 microns, we obtain the following SSA at 0, 50, 70, 80, and 90% RH:

BC –> 0.12, 0.12, 0.12, 0.16, 0.20

POM –> 0.962, 0.978, 0.982, 0.986, 0.990

For DUST the refractive indices are taken from Syniuk et al., 2003 (Sinyuk, A., O. Torres, and O. Dubovik, Combined use of satellite and surface observations to infer the imaginary part of refractive index of Saharan dust, Geophys. Res. Lett., 30(2), 1081, doi:10.1029/2002GL016189, 2003.) and do not vary with RH, but they vary with wavelength and size. They range 0.993 to 0.897 at 0.55 microns from smaller to larger size bins.

11) What is the absorption coefficient of BC? How is it calculated?

The absorption coefficient of BC is calculated from its SSA.

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species.

As mentioned above, the refractive indices and hygroscopic growth factors are taken from GADS/OPAC database (Hess et al., 1998) for all species but dust, whose optical properties are taken from Sinyuk et al. (2003).

For full details and data on how AOD is calculated within GEOS-Chem, the user may refer to the FlexAOD post-processing tool: http://wiki.seas.harvard.edu/geos-chem/index.php/FlexAOD

1) Do you assume and how internal mixing when computing optical properties of total aerosol? Is Size varying? Which mixing assumption?

INCA uses 5 modes, two insoluble (accumulation, coarse) , three soluble (accumulation, coarse, super coarse). Size varies, but sigma of lognormal distributions is fixed. Each mode has one number and several mass tracers which are advected separately. Size varies as a function of number and mass affected by mixing, different source and removal processes. Internal mixing is assumed with respect to some properties of each mode, namely removal by sedimentation, wet scavenging. Optical properties are calculated assuming external mixing, though the size of all species in a given mode are assumed to be equal. Size variation is not taken into account in the AOD calculation. Optical properties assume mean size for each mode.

2) Do you differentiate between clear-sky and all-sky relative humidity and which one do you use for the AOD calculation, reported to AeroCom?

In the LMDzT host model instantaneous clear-sky relative humidity is computed. This is used to compute hygroscopic growth factors, used in tabulated form, to get to AOD.

NOTE: The meaning of the clear-sky humidity in the LMDzT model has recently changed. Not clear how this affected the results.

2b) If you differentiate, have you investigated the difference?

All-sky AOD was never calculated.

3) Is the reported-to-AeroCom AOD as asked under 2) used directly for the RF calculations?

No - The AeroCom AOD diagnostics are computed for a given wavelength. RF uses the broadband AOD for the SW and NIR bands used in the RF calculation.

4) Would you consider the diagnosed AOD provided to the Aerocom database as representative for clear-skies? Eg for comparison to Aeronet and satellite retrievals in clear skies?

yes

5) If you diagnose a clear-sky AOD ⇒ How do you compute it? What are the assumptions on RH? Any threshold of RH you use? Cloud fraction weighting to derive a monthly average Clear-sky AOD?

Growth factors for each mode are interpolated.

6) Which humidity growth factor and variation do you use ? How does it depend on RH? Do you include sub-grid variability in relative humidity?

11 tabulated growth factors between 0 and 90% RH and one at 95% RH. No inclusion of sub-grid variability.

7) Do you diagnose the aerosol water mass or aerosol water AOD? How?

Yes, but it does not correspond strictly to the growth factors used for AOD computation.

8) Is the diagnosed sea salt and sulfate AOD computed at ambient RH? And eventually is that using clear-sky or all-sky RH?

Yes, clear-sky RH.

9) Do you average nighttime and daytime values of AOD into daily averages?

daily averages

10) What is the single-scattering albedo of your BC, POM, DUST? Does it vary as a function of size, RH? How roughly? Which range?

Depending on wavelength and RH BC soluble SSA varies between 0.41 and 0.53 BC insoluble SSA varies between 0.30 and 0.42 OA SSA varies between 0.94 and 0.99 DUST SSA varies between 0.944 and 0.987

11) What is the absorption coefficient of BC? How is it calculated?

from SSA

12) Please provide the refractive index at 0.55 microns for each aerosol species? And eventually further information on the hygroscopicity for each species.