Constraining Microphysical Processes of Warm Rain Formation Using Advanced Spectral Separations, an Ensemble Retrieval Framework and Machine Learning Techniques

Abstract

Drizzle, common in maritime warm clouds, plays a crucial role in determining cloud lifecycle and thus has a significant impact on Earth’s energy budget.  Yet, many models struggle to produce drizzle in the right amount with the right frequency, making it difficult to determine the response of clouds to a warmer climate.  This calls for a need for improved understanding and model representation of drizzle formation.

The need will be addressed using the wealth of advanced and comprehensive cloud observations from the Atmospheric Radiation Measurement (ARM) user facility.  The high-resolution ARM data provide an excellent opportunity for us to analyze autoconversion and accretion processes in which drizzle forms via collision and coalescence of droplets.  In this project we will exploit ARM measurements from the Azores and create concurrent cloud and drizzle properties and vertical profiles, critical observables for determining autoconversion and accretion rates.

Our product will be unique, because it:

• Maximizes the synergy between active and shortwave radiation observations, directly constraining cloud optical properties that are crucial for radiative effect quantification;
• Combines a clean, mathematical spectral separation between cloud and drizzle particles, with a physically-based retrieval framework that provides robust estimates and uncertainty; and
• Exploits the advances in machine-learning techniques to determine autoconversion and accretion rates.

This dataset will be the first direct vertical profile estimates of the process rates.  Given our present understanding, this will be a huge step forward, creating an unprecedented opportunity for the ARM users and the wider community to address key science questions in drizzle initiation and formation, and their interactions with aerosols, cloud, radiation and dynamics.