Flexible Radiation Codes for Numerical Weather Prediction Across Space and Time Scales
Abstract
Our long-term goal is to develop radiation parameterizations for dynamical models that are computationally efficient and work seamlessly across models at all time and space scales, especially from regional models to global models. The objective of this project is to build radiation codes for use in the Navy's limited area model (COAMPS) and global model (NOGAPS/NAVGEM). The codes will be scale-aware, computationally efficient across a range of computer architectures, and operate continuously rather than at discrete time steps. We are building on two algorithms developed by the PI: Monte Carlo Spectral integration (Pincus and Stevens, 2009) for time continuity and Monte Carlo Independent Pixel Approximation (Pincus et al., 2003) for sampling sub-grid-scale variability in cloud optical properties. We have nearly completed initial development of these codes, which are highly modular. Gas optics come from the state-of-the-art RRTMG radiation parameterization. Cloud optics use lookup tables developed at the Max Planck Institute for Meteorology (Hamburg), though other parameterizations are possible. We have written vectorized longwave (emission/absorption) and shortwave (multiple scattering) solvers and higher-level drivers. We have developed a general method for enapsulating the spectral sampling strategy used by MCSI, including the possibility of using one or more teams of spectral points to limit the maximum error realized. We are exploring methods for more accurate and efficient diagnosis of cloud radiative effects.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 30, 2011
- Accession Number
- ADA557119
Entities
People
- Robert Pincus
Organizations
- University of Colorado Boulder