Statistics and parameterization of shortwave horizontal photon transport effects

Poster PDF

Description

The ultimate goal of our project is to help improve the accuracy of radiative heating calculations in regional and global cloud resolving simulations, and to better understand the role of horizontal photon transport in cloud development and in the Earth’s radiative budget. For this, we are developing a neural network-based algorithm to provide corrections to the currently used 1D radiative calculations of cloud- resolving models, which cannot include horizontal photon transport effects. The first step of developing a correction algorithm was to create a data set of observed cloud structures and corresponding simulated radiation fields that are suitable for training a neural net. The poster includes new results from our statistical analysis of this data set, discussing the typical magnitude of 2D radiative effects over three ARM sites: NSA, SGP, and TWP. In addition, the poster also reports on our progress toward parameterizing horizontal photon transport effects. This includes examining both the effectiveness of the “Tilted Independent Column Approximation” to incorporate some horizontal processes into 1D radiative calculations and the effectiveness of a neural net to estimate horizontal photon transport effects using cloud texture parameters.