Abstract
Abstract. This paper introduces a new scheme available in the library of algorithms for representing cloud microphysics in numerical models named libcloudph++. The scheme extends the particle-based microphysics scheme with a Monte Carlo coalescence available in libcloudph++ to the aqueous-phase chemical processes occurring within cloud droplets. The representation of chemical processes focuses on the aqueous-phase oxidation of the dissolved SO2 by O3 and H2O2. The particle-based microphysics and chemistry scheme allows for tracking of the changes in the cloud condensation nuclei (CCN) distribution caused by both collisions between cloud droplets and aqueous-phase oxidation. The scheme is implemented in C++ and equipped with bindings to Python. The scheme can be used on either a CPU or a GPU, and is distributed under the GPLv3 license. Here, the particle-based microphysics and chemistry scheme is tested in a simple 0-dimensional adiabatic parcel model and then used in a 2-dimensional prescribed flow framework. The results are discussed with a focus on changes to the CCN sizes and comparison with other model simulations discussed in the literature.
Highlights
1 Introduction libcloudph++ is an open-source library of schemes for representing cloud microphysics in numerical models
It allows for representing cloud microphysical processes from first principles and is especially well suited to track changes in the cloud condensation nuclei (CCN) size distribution that are caused by clouds
Aerosol particles that served as CCN are altered by cloud microphysical and chemical processes and return to the atmosphere after water drops evaporate
Summary
The particle-based scheme used in this work is described in detail in Arabas et al (2015) and this section only briefly summarizes its major concepts. In the particle-based approach to modelling cloud microphysics, the computational domain is filled with “numerical point particles” representing a specified number (called multiplicity) of real particles (aerosol particles, cloud droplets, or rain drops) of the same properties. The process of condensational growth from deliquescent aerosol particles to cloud droplets is resolved and no additional parameterization of cloud droplet activation is required as it is again often done in bulk microphysics schemes, see for example Morrison and Grabowski (2007). As discussed in Unterstrasser et al (2017), it does introduce statistical errors, i.e. fluctuations between different realizations of the same collision/coalescence scenario These errors are easier to minimize than diffusion numerical errors, for example by increasing the number of SDs in the computational domain or by averaging over an ensemble of simulation runs. None of the above included a description of the aqueous-phase chemical reactions happening within cloud droplets
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
