Information content and aerosol property retrieval potential for different types of in situ polar nephelometer data

International audiencePolar nephelometers are in situ instruments used to measure the angular distribution of light scattered by aerosol particles. These types of measurements contain substantial information about the properties of the aerosol being probed (e.g. concentrations, sizes, refractive indices, shape parameters), which can be retrieved through inversion algorithms. The aerosol property retrieval potential (i.e. information content) of a given set of measurements depends on the spectral, polarimetric, and angular characteristics of the polar nephelometer that was used to acquire the measurements. To explore this issue quantitatively, we applied Bayesian information content analysis and calculated the metric degrees of freedom for signal (DOFS) for a range of simulated polar nephelometer instrument configurations, aerosol models and test cases, and assumed levels of prior knowledge about the variances of specific aerosol properties. Assuming a low level of prior knowledge consistent with an unconstrained ambient/field measurement setting, we demonstrate that even very basic polar nephelometers (single wavelength, no polarization capability) will provide informative measurements with a very high retrieval potential for the size distribution and refractive index state parameters describing simple unimodal, spherical test aerosols. As expected, assuming a higher level of prior knowledge consistent with well-constrained laboratory applications leads to a reduction in potential for information gain via performing the polarimetric measurement. Nevertheless, we show that in this situation polar nephelometers can still provide informative measurements: e.g. it can be possible to retrieve the imaginary part of the refractive index with high accuracy if the laboratory setting makes it possible to keep the probed aerosol sample simple. The analysis based on a high level of prior knowledge also allows us to better assess the impact of different polar nephelometer instrument design features in a consistent manner for retrieved aerosol parameters. The results indicate that the addition of multi-wavelength and/or polarimetric measurement capabilities always leads to an increase in information content, although in some cases the increase is negligible, e.g. when adding a fourth, near-IR measurement wavelength for the retrieval of unimodal size distribution parameters or if the added polarization component has high measurement uncertainty. By considering a more complex bimodal, non-spherical-aerosol model, we demonstrate that performing more comprehensive spectral and/or polarimetric measurements leads to very large benefits in terms of the achieved information content. We also investigated the impact of angular truncation (i.e. the loss of measurement information at certain scattering angles) on information content. Truncation at extreme angles (i.e. in the near-forward or near-backward directions) results in substantial decreases in information content for coarse-aerosol test cases. However for fine-aerosol test cases, the sensitivity of DOFS to extreme-angle truncation is noticeably smaller and can be further reduced by performing more comprehensive measurements. Side angle truncation has very little effect on information content for both the fine and coarse test cases. Furthermore, we demonstrate that increasing the number of angular measurements generally increases the information content. However, above a certain number of angular measurements (∼20-40) the observed increases in DOFS plateau out. Finally, we demonstrate that the specific placement of angular measurements within a nephelometer can have a large impact on information content. As a proof of concept, we show that a reductive greedy algorithm based on the DOFS metric can be used to find optimal angular configurations for given target aerosols and applications