Beyond the Big-Leaf Model for NOAA’s Unified Air Quality Forecasting Capabilities

The ongoing development at NOAA for next-generation, Unified Forecast System-Limited Area Models (UFS-LAM) has led to unification around a future high resolution Rapid-Refresh Forecast System (RRFS) for short-term weather regional UFS applications. Under this vision, NOAA has also developed a novel UFS-based regional air quality forecasting application for future (summer 2023) operational implementation, known currently as the “Online-CMAQ (Community Multiscale Air Quality)” model. Similar to a myriad of other community and operationally-based air quality models, however, Online-CMAQ continues to rely on the approximate but useful “big-leaf” model, that uses a bulk single leaf layer, with no representation of the underlying canopy interactions with chemistry and scalar transport. In this work, we take the first steps to advance beyond the big-leaf approach for NOAA’s UFS/Online-CMAQ application, which impacts model chemical performance via the effects of the forest canopy on (1) photolysis attenuation and (2) vertical turbulent transport and diffusivity. This work also addresses a long standing issue of systematic ozone overpredictions for NOAA’s air quality models that are linked to distinct vertical gradients of ozone measured within dense forest canopies of the U.S. This is made possible by incorporation of additional canopy/vegetation characteristics into the Online-CMAQ modeling system, including gridded forest canopy height, fraction, and clumping index, as well as cumulative leaf area index fractions for different heights within the canopy. The initial implementation of in-canopy photolysis is based on the probability of beam attenuation (i.e., fractional light attenuation), while the modulated turbulence/vertical diffusivity parameterization is based on Raupach’s (1989) near-field theory. Preliminary results from a proof-of-concept, experimental version of the current NOAA/NWS National Air Quality Forecasting Capability (NAQFC; also based on the CMAQ model) show that the in-canopy photolysis and turbulence attenuation parameterizations lead to improved predictions of near-surface ozone, while reducing the mean bias (compared against ground-based ozone observations) by 50% for an average contiguous U.S. domain. Preliminary results of in-canopy photolysis attenuation parameterizations in Online-CMAQ consistently show reduced overpredictions of near surface ozone in regions of contiguous canopies. Future incorporation of the turbulence parameterization in Online-CMAQ is expected to further improve ozone predictions.