|
Model Development and Numerical Techniques
Prof. Turco's group has for many years contributed practical and efficient modeling tools to study atmospheric chemical and microphysical processes. With Xuepeng Zhao, a new technique was recently designed for solving comprehensive sets of photochemical rate equations that derives from an earlier "family" scheme. The new "super family matrix inversion" (SFMI) approach is more stable and reliable, is well suited for long term global chemical simulations, and is compatible with parallel computational architectures. Turco and Zhao also derived a simplified representation of stratospheric photodissociation rates based on the "curve of growth" behavior of broadband radiative absorption in the atmosphere. Prof. Turco collaborated in developing a new ocean biogeochemical model patterned after atmospheric chemical-microphysical models designed at UCLA. The biogeochemistry code has been integrated with the POP ocean model at the Los Alamos National Laboratory.
Photochemical Systems and Algorithms: To carry out global chemical and microphysical simulations effectively, particularly in the context of climate research, efficient, stable and accurate numerical algorithms are essential. In the past, Prof. Turco and his group have derived a number of practical chemical and microphysical algorithms, and applied these to a wide variety of situations in Earth and planetary sciences. Solutions of the photochemical rate equations of atmospheric chemistry are particularly difficult, and require specialized computational treatment if long-duration simulations are to be achieved for climate analyses, and other purposes. In one approach, introduced by Prof. Turco and Mark Jacobson, the highly reliable Gear integration algorithm was modified to increase efficiency a hundred-fold in three-dimensional applications. The resulting sparse-matrix vectorized Gear code (SMVGEAR) is now widely employed in atmospheric chemistry simulations (including the SMOG regional model described earlier). However, for very long time integrations, an even faster solver has been designed. This new "family" based photochemical scheme introduces stability into the calculation of families through a matrix formulation. The super-family matrix inversion (SFMI) scheme has the efficiency of other similar implicit approaches, but is more accurate and stable. Further, this new scheme is well suited for applications on massively parallel computers, which are favored for climate work. As with the other codes developed by his group, the SFMI code is modular and easily accepts different photoreaction sets.
Biogeochemical System Modeling: A major objective of the atmospheric chemistry and climate research community is to integrate biological processes into models in a comprehensive and self-consistent manner. Toward this end, Prof. Turco and co-workers have developed a framework in which biogeochemical processes are treated in the same manner as atmospheric chemical and microphysical processes. This approach adopts for ocean applications the well-tested algorithms developed for atmospheric tracers. In the ocean version of the tracer code, photochemical, thermochemical, biological and microphysical (detrital) mechanisms are incorporated in a manner analogous to their atmospheric counterparts. The first application of this new biogeochemical modeling approach has been carried out at the Los Alamos National Laboratory in the context of the POP ocean general circulation model. The first simulations with the coupled POP/biogeochemistry code considered the impacts of anthropogenic nitrates transported from mainland China on northern Pacific gyre phytoplankton. The computational results are encouraging, and the flexibility of this new approach is now apparent.
|
|
