1. Climate Correlation – DNSC Troposphere and Surface Nigel Marsh, Freddy Christiansen, Torsten Bondo, Henrik Svensmark ESTEC, March 22, 2006

2. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 2 Monthly Solar and Climate Data Solar Modulated Forcing Parameters Tropospheric Climate Data TSI (1978-present Temperature (1958-present) UV and EUV (1978-present) Vertical Winds (1948-present) SW (1963-present) Clouds (1983-2004) HPI (1978-2003) Ozone (1978-present) GCR (1958-present) Aerosols (1978-present) SSN Surface Response

3. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 3 Solar Modulated Forcing Parameters TSIUVEUVSWHPIGCRSSN TSI0.800.790.310.09-0.630.72 UV0.980.480.17-0.790.77 EUV0.500.19-0.850.96 SW0.59-0.540.46 HPI-0.440.15 GCR-0.79 TSIUVEUVSWHPIGCRSSN TSI0.700.800.790.310.09-0.630.72 UV0.300.980.480.17-0.790.77 EUV0.270.500.19-0.850.96 SW0.870.59-0.540.46 HPI0.99-0.440.15 GCR0.61-0.79 TSIUVEUVSWHPIGCRSSN TSI0.700.800.790.310.09-0.630.72 UV0.590.300.980.480.17-0.790.77 EUV0.500.770.270.500.19-0.850.96 SW-0.02-0.060.070.870.59-0.540.46 HPI-0.030.040.140.600.99-0.440.15 GCR-0.21-0.44-0.42-0.29-0.530.61-0.79

4. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 4 Internal Modes of Variability NAVolc. TSIUVEUVSWHPIGCRSSN Nino30.00.22-0.05-0.130.010.05-0.08-0.020.00 NAO 0.090.160.100.090.050.09-0.050.09 Volc. 0.060.02-0.080.150.22-0.02-0.09

5. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 5 Method of Analysis Climate parameters: subtract linear trend and internal modes of variability subtract solar activity (SSN) Solar modulated forcing parameters: subtract linear trend subtract solar activity (SSN) Significance: random phase test

6. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 6 Hadley Radiosonde Temperatures Monthly data over land (balloon) Revised data set

7. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 7 Global Tropospheric Temperature TSIEUVSWHPIGCRSSN Observ.0.13(0.16)0.33(0.35) 0.20(0.31)-0.08(-.07)-0.31(-.33)0.25(0.27) Niño, etc.0.31(0.44)0.42(0.50) 0.19(0.35) 0.07 (0.14)-0.47(-.57)0.40(0.48) Solarcycle0.05(0.00)0.07(0.07)-0.02(.02)-0.01(0.03)-0.27(-.33)

8. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 8 Zonal Averages

9. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 9

10. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 10 MSU temperatures

11. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 11 Temperatures Solar cycle responsible for the major part of the correlations observed Limiting the time period to 1978-present removes the correlation also of GCR When viewed over all 5 solar modulated parameters the correlations are strongest in the tropics UV shows correlations comparable to those with GCR, but they dissapear after removal of the solar cycle GCR the only parameter with a moderately strong correlation with global temperatures after solar cycle is removed

12. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 12 Low Cloud Coverage TSIUVSWHPIGCRSSN Observ.-0.40-0.58-0.24-0.080.40-0.36 Niño,etc.-0.41-0.50-0.24-0.310.52-0.42 Solar-0.21-0.50-0.07-0.180.35 -0.54 -0.61 -0.34 -0.30 0.62 -0.57 -0.49 -0.57 -0.27 -0.28 0.59 -0.54 -0.23 -0.41 -0.03 -0.10 0.32

13. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 13 Low Clouds – UV and GCR Before solar activity Significant After solar activity

14. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 14 UV and GCR display very similar behaviour and the effect of these two parameters is therefore hard to distinguish from each other. TSI is negatively correlated with low clouds. Some of this correlation survives the removal of the solar cycle. Low Clouds The solar cycle is negatively correlated with low clouds over most of the globe and is responsible for most of the correlation seen in SW, HPI, and TSI. UV and GCR show negative(positive) bands of correlations at mid to high latitudes. These bands are strengthened by removal of the solar cycle which also causes positive(negative) bands of correlation to appear in the tropics.

15. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 15 Vertical Winds

16. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 16 Vertical Winds All solar modulated forcing parameters show correlations in alternating positive and negative latitudinal bands stretching from ground level into the stratosphere. The correlations for the different parameters are very similar but not at the same latitudinal position. No clear conclusion can thus be reached for this climate parameter.

17. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 17 Ozone TSIUVSWHPIGCRSSN Niño, etc.0.450.560.290.14-0.520.48 Solar cycle0.150.250.030.08-0.20

18. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 18 Ozone Two strong symmetric bands of positive correlations are seen in the sunspot number around the equator. Only small amounts of correlations are observed for the solar modulated forcing parameters after removal of the solar cycle, but is stronger in the UV. Although a strong correlation is found in global averages for GCR, very little is seen in the spatial plots.

19. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 19 Aerosol Index TSIUVSWHPIGCRSSN Niño, etc.-0.28-0.130.050.350.03-0.03 Solar cycle-0.190.190.160.37-0.26

20. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 20 Aerosol Index Almost no correlation with the solar cycle. Strong correlations in the southern hemisphere for UV and GCR, and at the equator for HPI.

21. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 21 Surface Response Very little is seen in the correlation between the solar modulated forcing parameters and the various surface parameters, except possibly for surface pressure.

22. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 22 Summary TSI Zonally averaged temperature under solar min conditions showed some significant positive correlation in the equatorial troposphere. HPI Zonally averaged temperature under solar min conditions showed some significant negative correlation in the equatorial troposphere. No significant correlation was found in any of the other climate parameters. Aerosols displayed a significant positive correlation at the equator. HPI Zonally averaged temperature under solar min conditions showed some significant negative correlation in the equatorial troposphere.. SW Zonally averaged temperature under solar min conditions showed some significant positive correlation in the stratosphere. No significant correlation was found in any of the other climate parameters.

23. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 23 Summary UV Zonally averaged temperatures displayed significant positive correlation coefficients that were comparable to those found with GCR, but then disappeared after removal of the solar cycle. Aerosols displayed a significant positive correlation in the southern hemisphere. Ozone displayed limit areas that were correlated with UV but these were signficantly stronger than with the other solar modulated parameters. Low clouds displayed negative bands of correlations at mid to high latitudes. These bands were strengthened by removal of the solar cycle leading to additional positive bands of correlation to appear in the tropics.

24. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 24 Aerosols displayed a significant correlation in the southern hemisphere which could not be distinguished from the signal found with UV. Summary GCR Zonally averaged temperatures displayed a significant negative correlation. GCR was the only parameter displaying a significant correlation for global averages after removal of the solar cycle. However, when limiting the analysis to data after 1978 this correlation disappeared. Low clouds displayed positive bands of correlations at mid to high latitudes similar to UV. The response with UV and GCR was generally very similar and it was difficult to distinguish between these two parameters. However, after removal of the solar cycle the globally averaged correlations were definitely weaker for GCR. Zonally averaged temperatures under solar max(min) conditions displayed a consistent negative(positive) correlation in the troposphere. A similar, but weaker, signal was observed for UV.

25. Climate Correlation review Meeting, ESTEC | March 22, 2006 | Freddy Christiansen | page 25 Summary Solar activity (SSN) is responsible for a large part of the correlations observed with all of the climate parameters, except aerosols. This means that any correlation observed for one solar modulated forcing parameter may be found in the others, particularly TSI, UV, and GCR. In the analysis of the fluctuations on top of the solar activity (SSN) GCR and UV more often showed a significant respons compared to the other solar modulated parameters.