Interactive Photochemistry in Earth System Models to Assess Uncertainty in Ozone and Greenhouse Gases

Funding Program: 

Objectives: This effort will deliver a photochemistry module working within CESM that is more complete than our existing super-fast chemistry, and that is efficient enough computationally to become a standard component of century-long ensembles used to assess future climate. Optimization will include both scientific choices on approximating the chemical mechanisms and computer science algorithms addressing some of the unique aspects of photochemistry (e.g., uneven loads). Such optimization is necessary because current chemistry-transport models with a full description of stratospheric and tropospheric chemistry are too costly to include regularly in climate ensembles. The new capability will enable more accurate climate simulations than those with off-line or climatology chemistries because the radiative forcing from ozone and aerosols will now have the sharp features representative of the local meteorology and climate change (instead of a smoothed climatology). Also, it will allow for analysis of climate statistics on the chemical processes controlling non-CO2 greenhouse gases and aerosols, such as abrupt climate change initiated by rapid release from natural CH4 reservoirs. A working chemistry-climate module embedded within CESM will provide chemistry-climate-consistent links across anthropogenic emissions of greenhouse gases and short-lived reactive species, changing natural emissions with climate, aerosols, and radiative forcing.

Description and Methods: Our new photochemistry module will build upon four major capabilities at UC Irvine and Lawrence Livermore National Lab:

  1. The fast-J photolysis methodology developed at UCI includes full scattering for low computational cost by optimizing wavelength integration, cloud and aerosol scattering functions, and radiative transfer. It is now used in many of the world's top CTMs.
  2. The super-fast tropospheric chemistry module was developed at LLNL to approximate the response of O3, OH, and SO2 oxidation to changing emissions and climate with just 16 chemical species. Preliminary tests of the super-fast chemistry in CAM3 show that it responds similarly to the full chemistry module, and is efficient enough to have been used in some of the CMIP5 climate simulations.
  3. UCI's linearized stratospheric O3 chemistry module has been used in century-long runs with the UCI CTM for the characterization of N2O-CH4 coupled modes. It now includes the chemistry of O3, N2O, NOY, and CH4; and it can be adjusted to match decadal changes in CFCs and stratospheric H2O.
  4. Computer science research at UCI has expertise in parallel compilers and algorithms that will be applied to the specific problems of chemistry in the CESM, e.g., atmospheric chemistry calculations suffer from highly uneven computational loads because of day-night and stratosphere-troposphere differences. There is additional opportunity for optimization in that many trace species need not be transported in large regions of the atmosphere, and this will also create load imbalance.

Potential Impact: Climate simulations will be more accurate with interactive O3 and aerosols and their associated local heating/cooling. Climate statistics from all the ensembles will now include these climate/meteorology-driven variations in short-lived forcings. Changes in air quality will also become a standard diagnostic. Other (independent) aerosol modules can use the gas-phase photochemical oxidants from this module for in-line, interactive aerosol chemistry. Biogeochemistry modules can use the fast-J calculation of photosynthetically active radiation that incorporates O3, aerosols and clouds. Terrestrial ecosystem modules can use the ozone and nitrogen deposition from this module. These latter connections allow for meteorological as well as climate-driven couplings across the system. External solar-cycle changes in ultraviolet flux and stratospheric ozone can also be simulated with Linoz.

Project Type: 
University Funded Research

Research Highlights:

None Available