A Novel, Efficient Approach to Investigating Sensitivities of Ice Crystal Concentration: Development and Evaluation of the Adjoint of a Physically-Based Cirrus Cloud Formation Parameterization

Cirrus clouds can significantly influence climate through their impact on Earth's radiative balance, stratospheric circulation, and rain production. Despite their climatic importance, the ice-cloud-climate connection is still poorly understood. Attempts to resolve this relationship often involve making small perturbations in global climate models (GCM) inputs and observing the results of the perturbation. While being conceptually straightforward, this finite difference approximation of sensitivities is time consuming and forces the state of the system to be altered when the evaluations are made and lacks the ability to simultaneously evaluate output sensitivity to multiple inputs parameters. The use of an adjoint has the potential to alleviate these problems.

In this study we develop and evaluate the adjoint of the Barahona and Nenes (2009) cirrus cloud parameterization (which is physically-based and considers competition between homogeneous and heterogeneous freezing for water vapor). The adjoint sensitivities were evaluated against finite differences, and then applied globally to study the relative importance of input parameters to cirrus ice crystal number concentration. The necessary inputs to the adjoint, including aerosol size distribution, ice nuclei characteristics, and updraft velocity, were obtained from a year long simulation of the NASA Global Modeling Initiative Chemical Transport Model driven by climatological meteorology obtained with the GISS GCM Model II'.