Environmental Engineering Reference
In-Depth Information
1. Introduction
Air Quality forecast models are employed to provide forecasters with numerical
guidance to issue timely ozone and particulate matter concentration forecast
pertinent to human health exposures to the communities they serve. The National
Centers for Environmental Prediction (NCEP) North American Mesoscale Model
(NAM) has been coupled with the EPA Community Multi-scale Air Quality
(CMAQ) model to form the National Air Quality Forecast Capability (NAQFC).
Coupling of the turbulent mixing processes of the meteorological driver and
CMAQ has been a challenge. Both Eta and NAM use the Mellor-Yamada-Janjic
(MYJ) turbulence mixing scheme in the boundary layer (Janjic, 2001). The option
of directly using MYJ provided vertical eddy diffusivity to calculate turbulent
mixing of the atmospheric constituents in CMAQ has been considered. This study
investigated the redistribution of ozone by various temporal averaging treatments
of vertical eddy diffusivity in the boundary layer. Model predicted concentration
of ozone resulted from these various temporal treatments of vertical eddy diffusivity
has been studied and verified using AIRNow, a surface ozone concentration observ-
ation network compiled by U.S. EPA, and ozone-sondes data. Observation sites
representative of various geographic and chemical settings to reflect their effect on
the diurnal cycle of the pollutant plumes are chosen, and the fidelity of vertical
diffusivity thus successfully captured or not by the schemes was studied.
2. Verical Mixing Schemes Sensitivity
In 2006, NAQFC (Otte et al., 2005) used a version of CMAQ very similar to
CMAQ version 4.5 (CMAQ-4.5) but customized computationally for NCEP
production requirements. Beside this real time forecast, two sensitivity cases were
run. They are stipulated hereby as Cases-
a
: Turbulence Kinetic Energy-Planetary
Boundary Layer height (TKE-PBL) derived Vertical Eddie Diffusivity,
K
z
; or
dubbed the (TKE-PBL Kz) Case - the baseline real time case; Case-b: (NAM-K
z
);
and Case-c: (NAM hrly K
z
). The latter two cases parameterize turbulent mixing of
tracer gases within CMAQ by using NAM predicted
K
z
.
2.1. RADM scheme with NAM TKE-based PBL height
The RADM (Regional Acid Deposition Model) scheme (Chang et al., 1987)
addresses boundary layer mixing based on a parameterization of turbulent mixing
in the surface and convective boundary layers using an application of the similarity
theory. The scheme computes vertical mixing using the eddy diffusion formulation,
the so-called K-theory. One benefit of the K-theory is the assumption of similar
diffusivity characteristics between tracer species and potential temperature: namely
