School of Earth and Environment

Impact of 11-yr Solar Cycle on Stratospheric Ozone

Sandip Dhomse and Martyn Chipperfield

Background

There are regular changes in solar radiation (or sunspot numbers) with an average period of 11 years (known as the 11-year solar cycle). Over this period, total solar irradiance (TSI) varies by only ~0.1% (~1.3 Wm-2) but this has various pathways to impact the Earth’s climate. As the largest radiation changes (up to 100%) occur in the ultra-violet part of the spectrum, solar flux variations can influence the production and destruction of stratospheric ozone. Due to its radiative properties, this ozone change then modifies the background stratospheric temperatures that in turn alter the upward propagation and breaking of the planetary waves that drive the Brewer-Dobson Circulation. Changes in tropical lower stratospheric ozone can also alter the equator-pole temperature gradient and influence the tropospheric Walker and Hadley cell circulations.

However, even with the availability of long-term satellite data, the exact structure and magnitude of the solar signal in stratospheric ozone is still not well understood. Satellite data sets from the solar occultation instruments SAGE I & II (1979-2005) and the nadir-viewing SBUV (1979-2005) show a “double-peaked” solar signal in tropical stratospheric ozone, with positive signals in the upper and lower stratosphere and no signal in the middle stratosphere. In contrast, data from another solar occultation instrument, HALOE (1992-2005), shows a positive solar signal in the middle stratosphere and negligible signals in the upper and lower stratosphere. We have investigated these satellite differences using our off-line 3-D chemical transport model (CTM).

Results

In Dhomse et al. (2011) we used the SLIMCAT 3-D CTM to investigate the 11-yr solar signal in tropical stratospheric ozone. The model was forced with ECMWF reanalyses (ERA-40 and ERA-Interim) from 1979–2005. We compared the modelled ozone solar signal to satellite observations. The standard model simulation is partially able to simulate a “double peak” ozone solar signal with a minimum around 30 km, in better agreement with HALOE than SAGE or SBUV (Figure 1a, green line). The model suggests that photochemistry contributes to the ozone solar signal throughout the stratosphere. The largest model-observation differences occur in the upper stratosphere where SBUV and SAGE data show a significant ozone solar signal (up to 4%) whereas the standard model and HALOE do not. For the model, this is partly due to a positive solar signal in the ECMWF temperatures which reduces the modeled upper stratosphere ozone signal through the well-known ozone-temperature anti-correlation. The large positive upper stratospheric ozone solar signal in SBUV/SAGE and SAGE data can, however, be reproduced in a model run with fixed dynamical fields (i.e. no inter-annual meteorological changes, red line in Figure 1a). As this run effectively assumes no long-term temperature changes, it should provide an unrealistic upper limit to the ozone solar signal, rather than a good fit to the observations. This highlights that quantification and attribution of the solar signal in stratospheric ozone remains an open scientific question.

Currently, this is even more critical as the present solar cycle (known as Solar Cycle 24) is one of the weakest in the last 100 years. Also, recent solar flux measurements from various instruments on the Solar Radiation and Climate Experiment (SORCE) satellites show significantly different variations in the solar fluxes than previous solar cycles. Using SORCE measured solar fluxes, Haigh et al. (Nature, 2010) and Merkel et al., (GRL, 2011) argued that the upper stratospheric ozone solar signal during solar cycle-23 is out of phase with the TSI changes. We performed SLIMCAT simulations with different solar flux sets for 2001-2010 and compared them with two satellite ozone data sets: SABER (2002-2010) and MLS (2004-2010). Overall we find that that model does not show an out-of-phase ozone solar signal in the upper stratospheric with the new solar fluxes (Figure 1b). Furthermore, due to the large uncertainties in MLS and SABER ozone observations in the upper stratosphere, and the limited length of the data record, we cannot establish the exact nature of the upper stratospheric ozone solar signal during solar cycle 23. However, our model simulations, as well as MLS and SABER data, do show a larger positive ozone solar signal in the middle stratosphere compared to earlier solar cycles, confirming the uniqueness of the ozone solar signal during solar cycle 23 (see Figure 1b).

<b>Figure 1(a):</b> Tropical ozone solar signal (25<sup>o</sup> S–25<sup>o</sup>N) from SLIMCAT 3-D model simulations for 1979–2010 with ECMWF ERA-40 and fixed dynamics (green and red lines), HALOE (1992–2005, black line) and a 2-D model (blue line). The estimated ozone solar signal using SBUV/SAGE data (triangles) and SAGE-based data (stars) are also shown. <b>1(b):</b> Estimated ozone solar signal using multivariate regression model from 3 SLIMCAT runs (2001–2010, 120 months), SABER (2002–2010, 108 months) and MLS (2004–2010, 77 months). Estimated errors (1-&#963;) for solar coefficients are shown with coloured horizontal lines. The large error bars (±10 %) at all levels for MLS data and in the lower stratosphere for SABER and model data are not shown. From Dhomse et al. (2013).

Publications

Dhomse, S., M.P. Chipperfield, W. Feng and J.D. Haigh, Solar response in tropical stratospheric ozone: A 3-D chemical transport model study using ERA reanalyses, Atmos. Chem. Phys., 11, 12773-12786, doi:10.5194/acp-11-12773-2011, 2011.

Dhomse, S.S., M.P. Chipperfield, W. Feng, W.T. Ball, Y.C. Unruh, J.D. Haigh, N.A. Krivova, S.K. Solanki and A.K. Smith, Stratospheric O3 changes during 2001-2010: The small role of solar flux variations in a chemical transport model, Atmos. Chem. Phys., 13, 10113-10123, doi:10.5194/acp-13-10113-2013, 2013.

Search site