Atmospheric particulate matter (aerosols) decrease visibility and are believed to affect climate by scattering and absorbing solar radiation aloft. Scattering by aerosols leads to an increase in planetary albedo (reflectivity) while light absorption can lead to warming of the atmosphere. In this work, an instrument to simultaneously measure aerosol scattering, extinction, and albedo is summarized. It is believed this method may find use in both field and laboratory studies of aerosol optics.

Background and Rationale

The attenuation of a beam of monochromatic light through an aerosol cloud can be modeled through the Beer-Lambert law relationship:

Formula (1)

where b_ext is the extinction coefficient (here we use Mm^-1 units) and z is the path length. In turn, the extinction coefficient is the sum of scattering and absorption coefficients, and single scatter albedo (ω) the ratio between the effects of scattering and extinction:

Formula (2)

Formula (3)

The reduction in local visibility and net climate effect of the aerosol depends on a number of factors including aerosol scattering coefficient (b_scat), extinction coefficient (b_ext), and the single scatter albedo (ω). Therefore, precisely measuring these variables on both lab generated aerosol mimics and genuine ambient aerosols is of considerable interest.

Traditionally, scattering and extinction have been measured separately through nephelometry and long-path length optical loss/extinction measurements (transmissometry). A major step forward occurred in 2001 when Smith and Atkinson applied cavity ring-down spectroscopy (CRDS) to aerosol extinction measurements [1]. Soon thereafter additional groups reported aerosol extinction measurements based on CRDS [2,3]. CRDS provides the requisite sensitivity for the measurement within a portable, compact instrument package. Similarly, interest in integrating sphere (reciprocal) nephelometry by several groups [4, 5] has led to improvements in device performance. All nephelometers cannot collect light over all angles equally. This leads to an angular truncation.

For commercial devices, the truncation angle is often 5-15°. The main technical advantage of integrating sphere nephelometry is this angle can be reduced to < 5°. Additionally, scattered light is collected over a solid angle of nearly 4π steradians - a condition that can help improve limits of detection. The albedometer directly builds upon these approaches by combining the technical advantages of CRDS with integrating sphere nephelometry.

Device Function

Figure 1 illustrates the experimental setup similar to that originally reported in our recent technical works [4,5]. Ambient air is drawn through either the aerosol inlet or an air filter. The filter can remove particles from the sample which provides a spectroscopic blank. The sample is then drawn through either the internal volume of a sphere nephelometer itself or a transparent tube placed within the sphere. The albedometer employs CRDS with a frequency doubled Nd:YAG laser to make extinction measurements at 532 nm. In CRDS, the rate of light attenuation is measured as a short pulse of light circulates in an optical resonator formed between two highly reflective mirrors (R > 0.999). After the light is introduced into the resonator, the beam is switched off and the light intensity then exponentially decays in time (first order) with a time constant τ. The time constant τ is the time required for the intensity to fall to 1 / e of its original value. Since mirror reflectivity is fixed, only light absorption and scattering by the sample (placed between the mirrors) leads to a change in rate of optical loss and a corresponding change in τ. The cavity time constant (τ) can then be linked to sample extinction coefficient through the equation shown in figure 1 if mirror reflectivity (R), distance between mirrors (L), and time required for light to make one "round-trip" through the cell (tr) is known. The effects of both particles and absorbing gases can lead to changes in τ. Experimentally, the effect of gases is subtracted by using filtered air as the spectroscopic blank. Ring down times on the order of 25-35 µs are often encountered for our system, offering detection limits for b_ext < 1 Mm^-1. In the setup, ambient pressure and temperature are also monitored at the measurement cell outlet. This allows for corrections in Rayleigh scattering due to changes in air density to be accounted for. Sample relative humidity (RH) is also recorded since aerosol optical properties are known to change with RH.

Simultaneously, on a second channel the device measures light scattered from the reflecting beam through use of an integrating sphere nephelometer and second photomultiplier tube. The interior of the 30.5 cm sphere we use is coated with a Lambertian diffuse reflectance material (also illustrated in fig. 1) with R > 0.95 for the visible region. The scatter channel signal also exhibits an exponential decay in time. For quantitative analysis, the scatter channel detector signal (ISCAT) is ratioed to the CRDS channel detector signal (ICRDS) at all points along the curve. Strawa et al. [6] have shown this ratio is linearly proportional to scattering coefficient (b_scat) through a constant K' that takes into account efficiency of light collection, detector electronic gains etc. The ISCAT / ICRDS ratio is then averaged for many points and this measurement related to scattering coefficient through calibration with gases of known Rayleigh scatter coefficient (often CO₂ and R-134a is used). In the first generation design, aerosol filled the sphere. The second generation instrument contains the aerosol within a transparent glass tube to reduce sample volume and instrument response time. Differential reflection off the glass tube can bias light collection efficiency as a function of scatter angle.