Our ability to predict future concentrations of significant
greenhouse gases such as methane and O
depends on our ability
to correctly predict tropospheric odd hydrogen chemistry. Prediction
of CH
CCl
concentrations have provided one test of our
understanding and theoretical ability to predict global OH
concentrations, but the test depends on accurate calibration of
the measured CH
CCl
and on accurate estimates of emissions.
Measurements of negative ocean saturation anomalies for CH
CCl
(Butler et al., 1992) have pointed to a previously unaccounted for ocean
sink that would require small (
10%) downward corrections to
model-predicted OH concentrations in order to maintain good agreement
with ambient CH
CCl
data. On the other hand, new rate data
for the reaction between CH
CCl
and OH are lower than
previously recommended rates by between 5 and 15% depending on
temperature (Finlayson-Pitt et al., 1992). This finding implies
that previously predicted OH concentrations would still be consistent
with the measured abundance of CH
CCl
even accounting for
the ocean sink. However, in contrast to model calibration studies with
CH
CCl
(Taylor et al., 1991; Spivakovsky et al., 1990),
studies which use 14CO for calibration indicate that model concentrations
of global OH are too low by about 45% (Mak et al., 1992).
An alternate method for testing our theoretical understanding of
odd hydrogen chemistry and our ability to predict OH is to compare
local photochemical model predictions of OH and peroxy radicals
with measurements in well-characterized field experiments. In contrast
to studies with CH
CCl
and 14CO, however, these studies indicate
that predicted OH may be too high. For example, Liu et al. (1992)
used observed and computed ratios of HNO
to NO
and observed and
computed CH
O and CH
OOH concentrations at Mauna Loa to argue that
model predicted OH is significantly overestimated, although other
factors such as measurement inaccuracy may also explain the
discrepancies between theory and observation. Studies in more polluted
areas by Cantrell et al. (1993) have shown that in some cases
model predicted peroxy radical concentrations agree with data, while
in others they do not. Significant progress has been made in
developing instrumentation to measure the concentrations of OH and HO
along with their organic analogues (Eisele and Tanner, 1991; Mount,
1992; Hard et al., 1992; Cantrell et al., 1993). In the future,
comparison of models with measurements along with more
accurate understanding of kinetic mechanisms should enable the
tropospheric chemistry community to derive a more accurate understanding
of odd hydrogen chemistry.