Feedbacks between a warming atmosphere, emission of aerosols, and clouds and
precipitation are some of the most difficult aspects for climate models to accurately
capture. While climate models operate at resolutions of tens or hundreds of kilometers,
many of the physics that determine how and where clouds form or precipitate
function at the micron droplet scale. Due to this disparity in physical scales, most
of these cloud physics must be modeled with only a few approximate quantities and
physical equations. These simplifications lead to large uncertainties about climate
forcings such as the sensitivity of global warming to human-emitted aerosols.
In this talk, I will introduce promising new techniques for representing hydrometeors in the climate system, with a particular focus on collisional processes between droplets. First, I illustrate how resolving a particle size distribution using collocated basis functions can recover the accuracy of a more expensive method, while remaining feasible for global scale simulations. Next, I generalize this approach to a flexible method of moments which attains an even higher ratio of accuracy to complexity. In this method, each basis function corresponds to a subdistribution of the overall droplet size spectrum and transfer of moments between basis functions is computed from explicit integration over the marginals of interacting and adjacent subdistributions. Both approaches eliminate the need for artificial and uncertain conversion rates while avoiding mathematical challenges of non-uniqueness and numerical instability. More importantly, these methods offer a flexible way to represent cloud microphysics over a wide range of complexity, providing a practical means to improve the representation of clouds in our climate system.
