A Python-based tool for fitting multi-mode aerosol size distributions

Atmospheric aerosols and precursors to cloud condensation nuclei span a range of sizes comprising several orders of magnitudes. While in experiments or field campaigns it is often convenient to express these aerosol size distributions in a histogram or sectional form, when applied to numerical simulations studying aerosol-cloud processes, often times other representations of the size distributions are necessary. For instance, several CMIP5 climate models featuring enhanced aerosol physics track the development of aerosol size distributions using a modal representation - a discrete set of pre-fitted, typically lognormal size distributions with just two free parameters. Necessarily, a modal aerosol size distribution is an approximation to any observed aerosol size distribution, often based on the assumption that the aerosol population is comprised of three major "modes" of aerosol sizes - nucleation, aitken, and accumulation - characterized by their mean size. The work presented in this paper highlights the development of a new analysis package, written in Python, for developing best-fitting, multi-mode lognormal distributions to aerosol size distribution observations in order to better implement these observations as initial conditions in numerical simulations.