Precipitation Efficiency’s Abstraction Art: Evaluating How Abstractions of Precipitation Efficiency in Microphysics Parameterizations Shape GCM Precipitation Outputs

Abstract:

Global circulation model (GCM) precipitation outputs reproduce observed average precipitation intensity values through compensating errors: specifically, lower precipitation intensity and higher precipitation frequency relative to observations. Here, we explore how parameterized representations of precipitation efficiency—the percentage of condensed water that clouds remove from the atmosphere to the surface—influence this general failure of GCMs to reproduce observed precipitation intensity distributions and frequencies. Current strategies for parametrizing precipitation efficiency prioritize temporally and spatially averaged microphysical relationships. We demonstrate from basic physical principles that, on short timescales, precipitation efficiency can be understood via in-cloud hydrometer mass distribution among three size classes. While hydrometer size distributions for fixed amounts of condensed water are highly stochastic, empirically their averages tend to converge when temporally and spatially averaged. We test whether prioritizing averaged microphysical relationships when parameterizing precipitation efficiency can explain the mixed success of GCM precipitation outputs in reproducing observations. We evaluate our hypothesis both from a process-based perspective using microphysics models and from a precipitation-output-based perspective using an aquaplanet GCM. We discuss how this work can generate new strategies for (1) parametrizing precipitation efficiency to improve agreement of GCM results with observations and (2) evaluating the physical robustness of parameterized representations of precipitation efficiency outside of the observable present day.