We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

physics.ao-ph

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Physics > Atmospheric and Oceanic Physics

Title: PySDM v1: particle-based cloud modelling package for warm-rain microphysics and aqueous chemistry

Abstract: PySDM is an open-source Python package for simulating the dynamics of particles undergoing condensational and collisional growth, interacting with a fluid flow and subject to chemical composition changes. It is intended to serve as a building block for process-level as well as computational-fluid-dynamics simulation systems involving representation of a continuous phase (air) and a dispersed phase (aerosol), with PySDM being responsible for representation of the dispersed phase. The PySDM package core is a Pythonic high-performance implementation of the Super-Droplet Method (SDM) Monte-Carlo algorithm for representing collisional growth, hence the name. PySDM has two alternative parallel number-crunching backends available: multi-threaded CPU backend based on Numba and GPU-resident backend built on top of ThrustRTC. The usage examples are built on top of four simple atmospheric cloud modelling frameworks: box, adiabatic parcel, single-column and 2D prescribed flow kinematic models. In addition, the package ships with tutorial code depicting how PySDM can be used from Julia and Matlab.
Subjects: Atmospheric and Oceanic Physics (physics.ao-ph); Computational Engineering, Finance, and Science (cs.CE)
Cite as: arXiv:2103.17238 [physics.ao-ph]
  (or arXiv:2103.17238v2 [physics.ao-ph] for this version)

Submission history

From: Sylwester Arabas [view email]
[v1] Wed, 31 Mar 2021 17:37:09 GMT (165kb,D)
[v2] Fri, 22 Oct 2021 17:24:28 GMT (166kb,D)

Link back to: arXiv, form interface, contact.