Optical Absorption of Aerosols

Here it is described the process of validation of the absorption diagnostics by specie and wavelength. Note that for wavelength of 550nm we have the optical depth by species but it is not directly an output for other wavelength (by specie).

Theory

The total extinction coefficient (or simply extinction) β_ext is the sum of the scattering β_sca and absorption coefficients β_abs . These coefficients are related with the refractive index and size/shape properties of the particles.

The single scattering albedo ω is defined by the ration β_sca ⁄ β_ext so when absorption is dominant over scattering ω will be small.

Given ω it is possible to calculate the absorption fraction of the total optical depth by

τ_abs = (1 − ω)τ_ext

but also if we apply the definition of optical depth:

τ_ext(z) = σ_extn(z)Δz = (σ_sca + σ_abs)n(z)Δz

this second approach is the one implemented at LMDZ. Specifically for an aerosol specie i

τ_extⁱ(x, z) = f(β_extⁱ)n⁽ⁱ⁾(x, z)Δz

the function f would be just the extinction coefficient if we have a insoluble aerosol, or a more complex function that includes the humidity for a soluble aerosol. Therefore:

τ_absⁱ(x, z) = f(β_absⁱ)n⁽ⁱ⁾(x, z)Δz

is the absorption fraction of the total optical depth.

Implementation

We use the last expression to estimate the τ_absⁱ that in practical terms is store the partial sum of total τ_abs (for all aerosols) in arrays for each specific specie.

The direct test is:

(τ_abs)/(τ_ext) = (1 − ω) = (β_abs)/(β_ext)

The following table shows the values of the extinction and absorption coefficients for each aerosol absorbing specie:

Aerosol	βext(440nm)	βext(550nm)	βabs(440nm)	βabs(550nm)
dust	0.788	0.818	0.081	0.048
bc-insoluble	5.342	5.159	2.861	2.806
oa-insoluble	5.300	4.569	0.170	0.145

so the ratios are:

Aerosol	βabs ⁄ βext 440nm	βabs ⁄ βext 550nm
dust	0.1028	0.0586
bc-insoluble	0.5355	0.5439
oa-insoluble	0.0321	0.0317

If we consider that:

f(β_ext) = β_ext +  corrections

and

f(β_abs) = β_abs +  corrections

Then the ratios of the table above would be similar also for soluble plus insoluble in the sense of order of magnitude

Aerosol	τabs ⁄ τext 550nm	βabs ⁄ βext 550nm
dust	0.0586 (cte)	0.0586
bc-insoluble	0.52 (+- 0.02)	0.5439
oa-insoluble	0.025 (+- 0.008)	0.0317

Therefore we see a reasonable consistency (exact for insoluble dust) approximate, as expected, for the other aerosol species.

The second test, is to validate that only BC, OA and DUST are contribution to absorption, that is shown in the next figure where we see the τ_abs^all , τ_abs^oa , τ_abs^bc , τ_abs^dust , and the difference between all aerosols absorption and the sum of these three species. We can see that the difference is just numerical noise.

Mean values

Comparison of global mean values of several optical properties for one single month

Aerosol	τext	τabs	ω	(1 − ω)	(1 − ω)τext
dust	0.02399	0.00141	0.8414	0.0586	0.00139
bc-all	0.00148	0.00075	(0.4561)	(0.5439)	(approx) 0.00080
oa-all	0.01083	0.00025	(0.9683)	(0.0317)	(approx) 0.00034
bc-insoluble	5.342	5.159	0.4561	0.5439
oa-insoluble	5.300	4.569	0.9683	0.0317

Comparison with CAM-OSLO of AeroCOM Phase III

Aerosol (ABS550)	IPSL (Feb 2006)	CAM53-OSLO (full year)
all	0.0024	0.00442
dust	0.0014	0.00314
bc	0.00075	0.000534
oa	0.00024	0.000750

Comparison of soluble and insoluble parts

Here we are compare yearly values of optical absorption (AAOD here named also τ_abs ).

For a simulation with flag_bc_internal_mixture=.FALSE.

Aerosol AAOD	λ = 440nm	λ = 550nm	λ = 870nm
total BC	0.000886551	0.000852139	0.000713484
bc insoluble	9.06376e-05	8.88952e-05	7.74269e-05
bc sol. int. mix	0	0	0
bc other soluble	0.000795914	0.000763244	0.000636057

Here we are compare yearly values of optical absorption (AAOD here named also τ_abs ).

For a simulation with flag_bc_internal_mixture=.TRUE.

Aerosol AAOD	λ = 440nm	λ = 550nm	λ = 870nm
total BC	0.00243913	0.00193918	0.00113904
bc insoluble	9.0574e-05	8.8832e-05	7.7372e-05
bc sol. int. mix	0.00234855	0.00185035	0.00106167
bc other soluble	0.000795914	0.000763244	0.000636057

Note that for bc insoluble the differences are very small.

Comparison with references

Here we compare with several references [R.Wang-2016] being the main one. Initially the BC is treated as: BC in the soluble accumulation mode was treated as an internal mixture considering insoluble BC inclusions in a soluble hygroscopic aerosol matrix, while the Mie-Scattering is used to ascertain the optical properties. Now the dry aerosols have a MEC (mass extinction cross-section) and a single scattering albedo of a internal mixture where the absorption is attributed to BC so the mass absorption cross section is estimated from the MEC and the single scattering albedo of BC. When there is an absorption enhancement, that is related with the soluble mode.

Regarding emissions according to R. Wang paper and based on global emission inventories the median typical global value is about 8.9 Tg/yr.

Optical properties of BC

Usually it is divided the BC in soluble and insoluble parts (modes). For insoluble a common assumption is a external mixed scheme with a given refraction index. The soluble can be treated as external or internal mixture, or insoluble parts on a soluble aerosol mixed state.

m_e² = m₀²(m_A² + 2m₀² + 2v_A(m_A² − m₀²))/(m_A² + 2m₀² − v_A(m_A² − m₀²))

Update 2020 on BC internal mixing

In principle the equation

τ_ext = τ_abs + τ_sca

However the results of the evaluation of the version with BC internal mixing have other results. Here we have a global mean (and time average) of monthly fields and we obtain.

Diagnostic	Name	Internal Mixing	External Mixing
τextBC	od550bc	0.0013462	0.0016742
τabsBC	abs550bc	0.0021083	0.0008521
τabsBCI	abs550bcI	0.0000960	0.0000889
τabsBCS	abs550bcs	0.0000000	0.0007632
τabsBCS	abs550bcsin	0.0020123	0.0000000

The files for BC-internal has been taken from AeroCOM-CTRL-2019-PD-BC and year 2010 The files for BC-external has been taken from AeroCOM-CTRL-2019-PD and year 2005

The problem is that the

τ_extⁱ(x, z) = f(β_extⁱ)n⁽ⁱ⁾(x, z)Δz

τ_absⁱ(x, z) = f(β_absⁱ)n⁽ⁱ⁾(x, z)Δz

the lookup tables of R.Wang has values: β_ext^BC <  , in the case of the internal mixing.

Note

This test is located at _research/test_abs_vs_od_BC

[R.Wang-2016]

R. Wang et al, Estimation of global black carbon direct radiative forcing and its uncertainty constrained by observations, JGR-Atmos. 2016