Developing a downscaling method from global to regional ozone modeling: application for linking RAQMS and CMAQ

Abstract

The goal of this study is to assess the impact of the downscale linkage of global model output on the simulated regional scale O3 vertical/horizontal structure. In this research, we incorporated the analysis and forecasting of a global model with the regional model utilizing a specially designed satellite ozone data assimilation process. Two different methods can be applied to incorporate the satellite measurements in the regional ozone modeling. One is a method that utilizes the best-optimized analysis field directly as the initial conditions for the simulation, and the other is the downscale linkage of the output of a global scale model into which the satellite observations have been already assimilated. In this work, we focus on the second indirect method with the global RAQMS (NASA Langley Research Center and University of Wisconsin Regional Air Quality Modeling System), which adopts a simple optimal interpolation approach (SDF; Statistical Digital Filter), to conduct the TOMS column and solar occultation assimilation procedure. To do this, a linking tool to ingest the global RAQMS simulations into CMAQ (Community Multiscale Air Quality) model has been developed. We performed two sets of CMAQ simulations; one with the predefined lateral boundary condition (BC) profile and the other with the lateral BC generated from RAQMS outputs for the 1999 Southern Oxidant Study (SOS-99) period (Jun. 15-Jul. 14, 1999). The two simulation results were compared and evaluated with several ozone soundings and national wide surface observations (AIRS/EPA) in U.S. and Canada. In the upper troposphere and lower stratosphere, CMAQ simulations with the predefined BC profile failed to simulate the so-called chemical tropopause above which O3 concentrations rapidly increase with height. On the other hand, the simulation with BC provided with the global RAQMS outputs showed good agreement with the ozone soundings. However, there were no significant differences between the two in the spatial distributions of surface O3. These results indicate that the usage of BC generated from global simulations with a satellite data assimilation procedure certainly improves the performance of regional models above the middle troposphere in the northern part of the U.S. continent and Canada. In future work, we will apply both SDF and 3DVAR methods for CMAQ modeling with GOES ozone column and make a comparison with current results.