The impact of stratospheric ozone depletion and CO

The impact of stratospheric ozone depletion and CO 2 on AAO trends and regional climate change at surface of Southern Hemisphere Syktus J. , I. Smith 2, G. Mc. Keon 1, M. Dix 2 & W. Cai 2 1 Department of Natural Resources, Mines and Energy, Brisbane, Australia 2 CSIRO Atmospheric Research, Melbourne, Australia C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Motivation – the source of observed 20 th. C trends in the SAM is a matter of scientific debate with stratospheric ozone losses, greenhouse gas increases and natural variability all possible players –my motivation is from practical point of view; does it influence regional climate trends and change – since it is difficult to separate the contribution of these various forcings in the observed record, a global climate model has been used C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Southern Hemisphere Annular Mode - Discussion • It is a dominant mode of low frequency variation of the SH circulation. Characterized by zonally symmetric fluctuations in the extra-tropical atmospheric circulation that extends from surface to the stratosphere • The observation show that AOO index has trend towards a positive values since the late 1970 s, coincident with occurrence of ozone depletion (Marshall 2003) • The issue here is whether AGCM forced with observed SST and radiative forcing factors such as CO 2 & O 3 can simulate the trends in surface MSLP and other surface characteristics at regional scales • Climate models forced by changing CO 2 or by decreasing stratospheric ozone usually show a trend in the SH annular mode. • Still an open question how good a representation of stratosphere is needed to accurately model tropospheric climate and climate change, however link to stratosphere is not essential for MSLP to have SH annular mode like response C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Martin Visbeck Todd Mitchell C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Problem and approach • During the past decade large parts of eastern Australia experienced unprecedented and persisting drying trends • Drying trends in eastern Australia in recent decades Water supply restrictions in major cities of Eastern Australia • Identification of possible forcing factors; GHG, O 3, land cover change … • Limitations of the approach • Review of the current state of knowledge • Simulation results and discussion C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Rainfall Relative to Historical Records Australia– July 1992 to June 2003 Percentiles Red = Last 10 years are in the bottom 10% of all previous ten July to June year periods from 1890 C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Trends in annual rainfall expressed in terms of the trend correlation coefficient (r, not b). 1952 to 2002 MJJASO 1952 to 2002 Annual 1952 to 2002 NDJFMA C 20 C Workshop Trieste, Italy 19 – 23 April 2004

All-Australian annual rainfall indices (a)area-averaged total, mean=461 1974 1902 r(SOI)=0. 50 2000 1905 1973 2002 1950 1961 1956 1994 (b)area-averaged mean decile value 1974 1905 1973 1928 2000 1961 1975 1994 r(SOI)=0. 44 C 20 C Workshop Trieste, Italy 19 – 23 April 2004 1917 2002

• Despite 2002 being one of the driest years on record, Australia appears to have become wetter • This positive trend in NW Aus appears because of relatively wet years in the second half of the record (e. g. 1974, 2000). • The chances of these trends arising due to random fluctuations of the data are about 5% We do not know: • where the trends have occurred • if there any compensating negative trends • which seasons are involved C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Second dominant EOT of Australian Observed Rainfall - S 2 Central west (15%) S 2 represents variability over that potion of the continent where rainfall stations are most sparse. It is possible that T 2 may be dominated by the relatively small number of stations whose effect has been interpolated over large distances. (Smith 2004) T 2 has also positive & statistically significant trend during the 1900 -2002 period, strongest over the past decade NDJFMA C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Summary • Increases in all-Australian annual rainfall over the past 100 years are the result of increases in summer rainfall over much of the western half of the continent over the past 50 years. • The magnitude and pattern of the increases appears unusual and unlikely to be the result of “natural variability”. • It is not possible to describe winter rainfall trends as unusual. • Increases in summer rainfall have been suggested as a response to increased greenhouse gases. • Temperatures have risen substantially over both Australia and the Indian Ocean over recent decades. • These may be linked to the rainfall changes. • These may, in turn, Trieste, be attributable to. April EGG 2004 - but earlier than C 20 C Workshop Italy 19 – 23 expected.

C 20 C Workshop Trieste, Italy 19 – 23 April 2004

March 2002 Rainfall Forecast for July to September IRI Net Assessment Forecasts NR&M contributes NCEP model output at monthly intervals to IRI consensus forecast since late 1998 • Forecast skill assessment (Goddard et al. , 2004, BAMS) http: //iri. columbia. edu/climate/forecast/net_asmt/ C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Murray-Darling Basin – April-September rainfall and maximum temperature 1952 - 2002 C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Water Storages Levels Gladstone and Gold Coast water restriction in place for number of years now. December: 21 NSW towns required water carting Gold Coast Canberra: First water restrictions since 1966 Melbourne: First water restrictions since 1982 C 20 C Workshop Trieste, Italy 19 – 23 April 2004 Follows a 6 -year dry spell in southern Australia

Natural Annual Flows into Hume Weir (MDBC) Lowest 4 -year period on Natural Annual Flows into Hume Weir (MDBC)record C 20 C Workshop Trieste, Italy 19 – 23 April 2004

CSIRO Mk 3 AGCM Model • Atmosphere Grid: T 63 (1. 88 o x 1. 88 o) 18 levels - hybrid , p with top at 4. 5 h. Pa (~ 36 km) Semi-Lagrangian moisture transport UKMO convection (Gregory & Rowntree) Liquid water clouds (Rotstayn) • Land surface Soil model - 6 levels Temperature, water, ice 9 soil types 13 land surface and/or vegetation types Snow-cover model - 3 layers C 20 C Workshop Trieste, Italy 19 – 23 April 2004

CSIRO T 63/L 18 C 20 C Experiments • 5 runs: 1949 -2003 SST only 1961 -2003 SST and O 3 • • 1871 -2003 SST only 1871 -2003 SST and solar 1871 -2003 SST, solar and CO 2 1961 -2003 SST, solar, CO 2 and O 3 5 runs: • 10 runs: 1949 -2003 SST, solar, CO 2 and O 3 in progress • C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Ozone scenario used in CSIRO AGCM simulations Observations C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Ozone recovery projections C 20 C Workshop Trieste, Italy 19 – 23 April 2004

C 20 C Workshop Trieste, Italy 19 – 23 April 2004 Trends in the Southern Hemisphere annular mode in CSIRO Mk 2 transient greenhouse runs reverse when the forcing is stabilised. However, this may take several centuries to take place, depending on the emission scenario used.

CSIRO Mk 2 Model, Warming Run C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Correlation between the AAO and JJA Rainfall CSIRO Mk 2 Model, greenhouse run, showing when the AAO goes up rainfall over SWWA decreases, and vice-versa! C 20 C Workshop Trieste, Italy 19 – 23 April 2004

MSLP change in Mk 3 DJF JJA MAM SON ANN C 20 C Workshop Trieste, Italy 19 – 23 April 2004

SAM SWWA Rainfall Winter C 20 C Workshop Trieste, Italy 19 –Spring 23 April 2004

Queensland summer rainfall in GHG scenario with Mk 3 coupled AGCM C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Model Simulated and NCEP R 1 JJA MSLP for JJA 1971 -2000 C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Annual Trend NCEP PSL 1971 -2001 C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Model Simulated MSLP difference for JJA 1971 -2000 C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Model Simulated MSLP difference for JJA 1971 -2000 C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Model Simulated MSLP difference for JJA 1971 -2000 C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Model Simulated MSLP difference for JJA 1971 -2000 C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Model Simulated MSLP difference for JJA 1971 -2000 C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Trends in the SAM: impact on Antarctic temperatures Time series of the SAM based on NCEP-NCAR Z 850 data and Z 500 Antarctic radiosonde data (Thompson & Solomon, 2002). Data are normalized to 1968 -98. Regression of Antarctic surface temperature anomalies vs unit change in the SAM, 1982 -98. (Kwok & Comiso, 2002) Source: Todd Mitchell (http: //tao. atmos. washington. edu/aao/) C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Response of surface temperature Warming over Antarctic Peninsula and cooling over Eastern Antarctica seen in both CO 2 and O 3 forcing, but stronger in O 3 C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Response of surface temperature DJF 1993 -2003 cooling over Eastern Australia due to CO 2 effect and an opposite effect when O 3 is added C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Response of precipitation NDJFM 1993 -2003 wetter over Eastern Australia due to CO 2 effect and an opposite effect when O 3 is added, follows from surface temperature changes in previous slide. SPCZ changes C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Response of JJA precipitation JJA 1993 -2003 weaker Asian monsoon due to CO 2 effect and an lesser effect when O 3 case is considered. The area of precipitation decrease/increase in Southern Ocean consistant with AAO trend is visible C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Summer Rainfall Difference between 1961 -1994 and 1995 -2002 Observed Difference SSTs + Solar + CO 2 SSTs only SSTs + Solar + CO 2 + Ozone C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Summer Rainfall Difference between 1961 -2003 and 1998/9 -2000/1 Observed Difference SSTs only SSTs + Solar + CO 2 + Ozone C 20 C Workshop Trieste, Italy 19 – 23 April 2004

Conclusions Both CO 2 and O 3 forcing have contributed to pressure changes in Southern Hemisphere These changes are similar to SH Annular Mode The model has rather weak response to a radiative forcing when 3040 years long ensemble mean climatology is compared Model show some systematic response in surface air temperature and precipitation, significance of which needs to be assessed At regional scale contribution of CO 2 and O 3 to simulated rainfall anomalies is positive C 20 C Workshop Trieste, Italy 19 – 23 April 2004