Environmental Engineering Reference
In-Depth Information
meteorologically complex subdomains had larger numbers of clusters. Meteoro-
logical complexity here refers to the degree of interaction between the large-scale
pressure gradient and more localized terrain effects to produce the observed flow
field. Differences in complexity among the subdomains justify their independent
clusterings. SFBA exhibits four clusters and is the least complex. It is directly
affected by marine flows of variable strength which nearly always arrive from the
Pacific Ocean to the west. Relative to SFBA, SJV has one additional cluster (five
total) and is more complex. Bulk flows along the SJV major axis are nearly
always oriented from the north, and a number of localized, terrain-induced flow
features significantly impact regional air quality. SV exhibits seven clusters and
is the most complex. Bulk flows along the SV major axis can be oriented from
either the north or south, and localized, terrain-induced flow features significantly
impact regional air quality.
The clusters were named according to their dominant upper-level features.
Clusters R and H exhibit relatively calm conditions and have the highest ozone
pollution potentials. R and H reflect upper-level (500-hPa) ridges and onshore
high pressure centers, respectively. This pair of clusters generally has reduced
marine air flows through the Carquinez Strait and into the Central Valley, as well
as limited vertical dispersion. A variety of ventilated patterns (named using V)
exhibit relatively high winds through the Carquinez Strait and have lower ozone
pollution potentials. Cluster H/V is a hybrid pattern capturing high winds through
Carquinez Strait, limited vertical dispersion, and moderate to high ozone pollution
potential, depending on the subdomain. The clusters are seasonally distinct. H and
H/V account for more exceedances during the middle of the ozone season,
whereas R accounts for more exceedances near the beginning and end of the
ozone season. Specific to each subdomain, the clusters are associated with loca-
lized flow features which may trigger NAAQS exceedances. These features include:
sea breeze activity along the I-680 Corridor and Santa Clara Valley for SFBA;
lack of ventilating flows through minor Coast Range gaps such as Altamont Pass
for N-SJV; the Fresno Eddy for C-SJV; reversal of flows exiting SJV toward
Mojave Desert for S-SJV; and slope flows over the Sierra Range extending over
the Sacramento source area for SV.
Two subsequent analyses were applied to the wind field cluster labels to better
characterize meteorological impacts on ozone levels. The clusters alone categorize
the conditions on a given day in a given subdomain among a small number of
idealized patterns. They do not directly account for temporal patterns longer than a
single day (i.e., multi-day evolution of episodes) or spatial patterns larger than a
single subdomain (i.e., meteorological coupling of connected basins). First, we
tracked the day-to-day evolution of weather systems impacting ozone levels by
considering multi-day sequences of the daily cluster labels. This sequence analysis
distinguished between various episodic scenarios in which stagnating air masses
develop. Different scenarios reflect weather systems of onshore (over the Mojave
Desert or Four Corners) and offshore (over the Pacific Ocean) origins. Second, a
domain-wide synthesis of the clustering and sequencing results was constructed by
combining the independent results for all subdomains. Qualitatively, each of the
