IPCC Fourth Assessment Report: Climate Change 2007
Climate Change 2007: Working Group I: The Physical Science Basis Changes in the Hydroxyl Radical Over Time Impact of emissions

Because of its dependence on CH4 and other pollutants, tropospheric OH is also expected to have changed since the pre-industrial era and to change in the future. Pre-industrial OH is likely to have been different than today, but because of the counteracting effects of higher CO and CH4 (decreasing OH) and increased NOx and ozone (increasing OH) there is still little consensus on the magnitude of this change. Several model studies suggest a decline in weighted global mean OH from pre-industrial time to the present of less than 10% (Shindell et al., 2001; Lelieveld et al., 2002a; Lamarque et al., 2005a). Other studies have reported larger decreases in global OH of 16% (Mickley et al., 1999), 25% (Wong et al., 2004) and 33% (Hauglustaine and Brasseur, 2001). The model study by Lelieveld et al. (2002b) suggests that during the past century, OH concentration decreased substantially in the marine troposphere through reaction with CH4 and CO. However, on a global scale it has been compensated by an increase over the continents associated with strong emissions of NOx.

Karlsdottir and Isaksen (2000) used a three-dimensional CTM accounting for varying NOx, CO and NMVOC emissions and found a positive trend in OH of 0.43% yr–1 over the period 1980 to 1996. Dentener et al. (2003a,b), with a three-dimensional CTM accounting for varying emissions of ozone precursors and CH4, meteorology and column ozone, derive a positive trend of 0.26% yr–1 over the 1979 to 1993 period. J. Wang et al. (2004) also use a three-dimensional CTM accounting for interannual variations in CH4 and CO emissions, transport and column ozone to analyse the trend in CH4 from 1988 to 1997. They do not account for interannual variability of a number of other variables that affect OH such as concentrations of NOx, tropospheric ozone and NMVOCs. They also derive a positive trend in OH over the period considered of 0.63% yr–1. Their calculated trend in OH is associated primarily with the negative trend in the overhead column ozone over the period considered and the trend is reduced to 0.16% yr–1 when the total ozone column is held constant.

Future changes in OH depend on relative changes in hydrocarbons compared with NOx abundances. In the TAR, Prather et al. (2001), using scenarios reported in the IPCC SRES (IPCC, 2000) and on the basis of a comparison of results from 14 models, predicted that global OH could decrease by 10 to 18% by 2100 for five emission scenarios and increase by 5% for one scenario (which assumes large decreases in CH4 and other ozone precursor emissions). Based on a different emission scenario for future emissions, Wang and Prinn (1999) also predicted an OH decrease of 16 ± 3% in 2100. Effects of climate change

In addition to the emission changes, future increases in greenhouse gases could also induce changes in OH, arising through direct participation in OH-controlling chemistry and indirectly through stratospheric ozone changes that could increase solar UV radiation in the troposphere. OH will also be affected by changes in temperature, humidity and clouds or climate change effects on biogenic emissions of CH4 and other ozone precursors. Changes in tropospheric water could have important chemical repercussions. The reaction between water vapour and electronically excited oxygen atoms constitutes the major source of tropospheric OH. So, in a warmer climate characterised by increased specific humidity, the abundance of OH is expected to increase. This effect was proposed by Pinto and Khalil (1991) to explain the variation of OH during the cold dry Last Glacial Maximum (LGM). It was quantified by Martinerie et al. (1995) who calculated that the global mean OH concentration during the LGM was 7% lower than at present because the atmospheric water vapour concentration was lower during that period. Valdes et al. (2005) estimate that the cold and dry LGM climate was responsible for a 7% decrease in global OH. Brasseur et al. (1998) and Johnson et al. (1999) estimated that in a warmer (doubled atmospheric CO2) climate, the global and annual mean OH concentration would increase by 7% and 12.5%, respectively. More recently, Hauglustaine et al. (2005) use a climate-chemistry three-dimensional model to estimate a 16% reduction in global OH from the present day to 2100 accounting solely for changes in surface emissions. The effect of climate change and mainly of increased water vapour in this model is to increase global OH by 13%. In this study, the competing effects of emissions and climate change maintain the future global average OH concentration close to its present-day value. The importance of the water vapour distribution to global OH is illustrated by Lamarque et al. (2005a), who show that under reduced aerosol emissions, a warmer and moister climate significantly increases global OH concentration.

Changes in lightning NOx emissions in a warmer climate may also affect OH. Labrador et al. (2004) show that global OH is sensitive to the magnitude of lightning NOx emissions, and increases by 10% and 23% when global lightning is increased by a factor of 2 and 4, respectively, from a 5 TgN yr–1 best estimate. Similar sensitivity of global OH to the lightning source was estimated by Wang et al. (1998), who calculated a 10.6% increase in OH for a doubling of the source (from 3 to 6 TgN yr–1). Regarding the large uncertainty about lightning emissions and the sensitivity of OH to the total amount of N emitted, an improved understanding of this source appears important for the ability to simulate OH accurately over time.