 |
|
|
Key issues arising from the 2006 WMO/UNEP Ozone Assessment by T.G. Shepherd and W.J. Randel In the 2002 Assessment it was indicated that the total atmospheric burden of ozone depleting substances was responding as expected to the controls on production imposed by the Montreal Protocol, and that the ozone-layer depletion from the Protocol’s controlled substances is expected to begin to ameliorate within the next decade or so. The Executive Summary of the 2006 Assessment noted that an important next step is to ask whether stratospheric ozone and surface UV radiation are responding as expected to the controls imposed by the Protocol. In addressing this question it is necessary to consider factors other than ozone-depleting substances that also influence ozone and UV radiation. These factors include natural dynamical variability, volcanic eruptions, solar variations, aerosols, and climate change. The status of our understanding and the key questions for several of these issues are discussed below. Ozone-depleting substances The observed tropospheric abundances of HCFCs are increasing more slowly than anticipated, and those of bromine-containing gases are declining more quickly than anticipated. While both facts are good for the ozone layer, it is important to reconcile them with estimates of the relevant emissions. The importance of stratospheric bromine from very short-lived species (VSLS) appears to be significantly greater than previously estimated (WMO, Figures 2–3), and needs to be better quantified. We are still waiting for the observed decline of tropospheric bromine to be reflected in the stratosphere. Tropical ozone trends Observations
of column ozone from both ground-based and
Figure 1. (a) Meridional cross section of ozone trends during 1979-2005 derived from SAGE satellite and polar ozone-sonde data. Trends are derived from regression onto EESC, and expressed in terms of net percentage change during 1979-2005. Contours are -4, -8, -12, -16, -20, -30, -40%. (b) Latitudinal structure of annual mean column ozone trends during 1979-2005,derived from vertically integrated SAGE/sonde data and merged TOMS/SBUV data. Trends are expressed in terms of net ozone change over 1979-2005. The heavy dashed line denotes trends derived from SAGE data, integrated only over 25-50 km. From Randel and Wu, 2007. Short-term ozone recovery Ozone depletion has levelled off in every region of the atmosphere, consistent with the levelling off of stratospheric EESC (equivalent effective stratospheric chlorine). In some regions, ozone abundance has increased notably in the last 5 years or so (for example, over NH midlatitudes below 20 km; WMO Figures 3-11). This cannot yet be considered ozone recovery (since EESC has not notably declined), and the reasons for these increases need to be better understood since such variations will confound the detection of the onset of ozone recovery. Polar ozone and PSC microphysics There is now unambiguous evidence from Arctic measurements that NAT (nitric acid trihydrate) polar stratospheric cloud (PSC) particles can nucleate above the ice frost point, and that their occurrence can be widespread. Incorporating this process in chemical transport models (CTMs) improves the simulation of denitrification in the Arctic, but discrepancies remain in properly representing the effects of interannual variability, pointing to an incomplete understanding. Moreover, many of the specifics of PSC formation, such as freezing rates, remain empirical. Without a reliable representation of PSC processes, CCM predictions of past and future polar (especially Arctic) ozone are significantly compromised. Volcanoes The impact of the Mount Pinatubo volcanic eruption on stratospheric ozone remains something of a puzzle. While ozone amounts declined sharply in the Northern Hemisphere (NH) following the eruption, no such decline was evident in the Southern Hemisphere. Moreover, a number of modelling studies have suggested that the NH decline was mainly associated with changes in transport. As there are likely to be one or more volcanic eruptions during the ozone recovery period, there is a need to better understand the likely impact of such an eruption on stratospheric ozone. Stratospheric temperature trends There are substantial improvements in understanding
the uncertainties in historical stratospheric temperature data sets.
The long-standing differences between lower stratospheric trends derived
from Microwave Sounding Unit (MSU) satellite data and radiosonde-based
results can be reconciled by recognizing cooling biases in many individual
radiosonde stations (associated with instrumentation improvements over
time). Omitting the stations with largest biases allows more accurate
estimates of past variability and change (Figure 2). There is also improved
understanding of satellite data in the middle and upper stratosphere
(from the Stratospheric Sounding Unit, SSU), including quantifying the
effects of increasing CO2 on the measurements (which can significantly
influence trend results). These improved observational data sets will
provide critical tests for simulations of past stratospheric changes.
Figure 2. Comparison of near-global deseasonalized temperature anomalies calculated from MSU4 satellite data (top), vertically-integrated radiosonde data (middle), and their difference (bottom). MSU4 represents a weighted mean of temperatures in the layer ~13-22 km. The radiosonde results are averages over 35 individual stations over 60°N-S, using a subset of the Lanzante-Klein-Seidel data set (Lanzante et al. 2003), and are vertically weighted using the MSU4 weighting function. The MSU4 data here have been sampled at these same 35 station locations. Dynamical variability Long-term variability in wave forcing and other dynamical
quantities appears to have had a significant effect on observed ozone
abundance, especially in the NH, and has the potential to affect ozone
recovery on both short and long time scales. It is therefore important
to understand the extent to which long-term variability in dynamics
may be associated with climate change, and to better understand causes
of natural variability (including the apparent “trends”
associated with decadal-scale variability). Figure 3 shows the observational
record of winter-average planetary wave forcing of the NH stratosphere
for 1979-2006, together with winter average polar stratospheric temperatures.
These data show significant interannual variability across a range of
scales (yearly to decadal); the fundamental causes of such variability,
and potential shifts in a changing climate, are poorly understood.
Figure 3. Lower time series show winter-averaged
eddy heat flux (a proxy for planetary wave forcing) at 100 hPa for the
NH for 1979-2006 (averaged over December-March for each year). Upper
curves show the corresponding January-March averaged polar 100 hPa temperatures
(averaged over 60°– 90°N).Both sets of curves show results
derived from NCEP and ERA40 reanalyses, plus NCEP Climate Prediction
Center (CPC) data. Tropical tropopause temperature and water vapour CCMs generally predict a warming of the tropical tropopause region from climate change, and a modest increase in stratospheric water vapour, but these predictions do not appear to be consistent with past observations. CCM simulations of both fields often show large biases, with significant differences among models (Eyring et al., 2006, Figure 7). It is possible that long-term changes predicted in the models are more robust, but this would need to be demonstrated. There are also remaining uncertainties regarding decadal-scale changes in the observational record. Brewer Dobson circulation and age of air CCMs suggest an increase in tropical upwelling and thus decrease in age of air throughout the stratosphere, due to climate change (WMO, Figures 5-19). The extent of the increase varies substantially among models. The mechanism for the increased upwelling has yet to be determined, and its robustness assessed. Changes in age of air call into question the ODS scenarios used by CCMs, which impose tropospheric concentrations and thus cannot represent the effects of a faster removal of ODSs. Solar signal in ozone The ozone solar signal provides a key physical link between solar variability and climate, and is also important for interpreting low frequency ozone variability. However, there are substantial uncertainties in quantifying effects of the 11-year solar cycle on stratospheric ozone and temperature, both in comparisons of models and observations, and even among different observational data sets. The main differences regard the magnitude of the solar signal in column ozone (Figure 4), and the vertical profile of the solar signal in the tropics; much of the uncertainties result from the relatively short observational data records, and possible confusion of volcanic and QBO effects.
Figure 4. Latitudinal profile of the solar cycle variations in column ozone, derived from vertically integrated SAGE I+II data (over 20-50 km), and three column ozone data sets (groundbased, SBUV, and merged TOMS/SBUV data). Error bars on the TOMS/SBUV curve denote 2*sigma uncertainty in the fit. Ozone simulation by CTMs When driven by observed meteorology, CTMs should, in principle, be able to reproduce the observed behaviour of ozone. This makes CTMs potentially useful tools for separating (to the extent this is possible) the effects of chemical and dynamical processes on observed ozone changes. Although CTMs have been used very successfully to identify chemical processes in the context of particular winters, decadal-timescale simulations by CTMs are still plagued by errors in transport (e.g. age of air) from assimilated winds. This limits our ability to attribute past ozone changes. Ozone simulation by CCMs CCM predictions of future ozone are limited by a persistent young bias in age of air, although the situation has improved markedly in recent years. CCM simulations of midlatitude ozone can reproduce the overall features the past record, but there are substantial uncertainties in detail and differences among models (WMO Figures 3-26). While the observational record contains significant effects of dynamical variability, especially in the NH, such variability should also be evident in the CCMs. Reconciling the past observations with CCM simulations remains an essential task. Polar ozone and long-term recovery Model predictions of future Arctic ozone are highly uncertain because of large uncertainties in the future dynamical state of the Arctic polar vortex (WMO Figures 6-12, 6-13). It will be important to understand the sensitivity of modelled dynamical behaviour to various model parameters, such as horizontal/vertical resolution, dynamical wave forcing, radiative balances, etc. Model simulations need to become more robust and provide better estimates of the uncertainty associated with natural variability, as well as the effects of climate change. Radiative forcing from ozone changes The radiative forcing from stratospheric ozone changes (e.g. as used by IPCC) assumes that all the ozone changes are due to ODSs and thus that the ozone radiative forcing is an indirect forcing which can be set against the direct radiative forcing from the ODSs themselves. However, it seems clear that a significant fraction of the observed ozone changes are associated with changes in transport rather than with ODSs. Moreover, these transport-induced changes appear to be located preferentially in the lowest part of the stratosphere, where they have a maximum impact on radiative forcing. It is thus necessary to quantify the vertical profile of ozone changes attributable to ODSs, and its associated radiative forcing. References Eyring, V. et al., Assessment of temperature, trace species and ozone in chemistry-climate model simulations of the recent past, J. Geophys. Res., 111, D22308, doi:10.1029/2006JD007327, 2006. Lanzante, J.R., S.A. Klein, D.J. Seidel, 2003: Temporal homogenization of monthly radiosonde temperature data. Part I: Methodology. J. Climate, 16, 224-240. Randel, W. J., and F. Wu (2007), A stratospheric ozone profile data set for 1979– 2005: Variability, trends, and comparisons with column ozone data, J. Geophys. Res., 112, D06313, doi:10.1029/2006JD007339. Scientific Assessment of Ozone Depletion: 2006, WMO, Report No. 50.
|
|
|
|
|
