CVPS: An operator solving complex chemical and vertical processes simultaneously with sparse-matrix techniques

摘要

We present a locally, one-dimensional operator that couples complex Chemical and Vertical Physical processes with Sparse-matrix techniques (CVPS) for multi-dimensional regional photochemical transport models. The CVPS operator solves fundamental interaction between chemical reactions and vertical physical processes in the atmospheric boundary layer at each time step, and may be used to simulate chemicals sensitive to both vertical mixing and photochemistry at a time step. The CVPS operator is numerically stable and computationally efficient in atmospheric boundary layers over California. The computational advantage originates from sparse-matrix techniques and the low frequency for communicating feedbacks between CVPS and other local operators. Based on surface Ox (O_3 + NO_2) simulations in the Southern California Air Quality Study domain (Harley et al., 1993; Jacobson et al., 1996) on a dual quad-core Linux processor, the ratio of simulation/computer times may reach two for three-dimensional modeling using a classic horizontal advection solver and the CVPS operator.