Previous: Validation of the method for retrieving UV irradiance variations Next: Conclusions Up: Ext. Abst.

4. EW irradiance variations due to ozone and cloudiness

Using the historical data of cloud observations as well as ground ozone measurements we reconstructed variations in EW irradiance due to clouds and ozone in different geographical regions during warm period. Fig.5 illustrates variability in EW irradiance inferred from ground and satellite measurements for several Eurasian sites since the 1920-30s. It is clearly seen that there are different tendencies in CQ_g variations at different sites. We observe a significant CQ_g growth (about 10%) at the end of 30s over the Northern part of Eastern Europe, a strong CQ_g decrease over continental region of Western Siberia at the end of 40s, and in the 50s -over the Monsoon Far East area, where the CQ_g growth (higher than 10%) was observed in the middle of 80s. EW irradiance variations due to ozone changes have slightly less amplitude at the sites analyzed. But even over Arosa, where the ground ozone measurements were available since the middle of 1920x , the maximum changes of EW irradiance due to ozone comprise about ± 7%. It is necessary to emphasize that in several regions interannual changes in ozone and cloudiness are correlated (i.e. R²>30% for Arosa, Nikolaevsk-na-Amure, Irkutsk, etc): i.e. the minimum in ozone values corresponds to the higher cloud transmittance. The same picture was observed over San Diego (R² >55%) and, to some extent, over Ushuaia.(see Fig 4). This effect may significantly enhance the EW growth in conditions of ozone loss.

In order to reveal the main cause of the cloud fluctuation we analyzed relative changes in CQ_g for the sites within the Atlantic Continental region of Eastern Europe together with the examination of synoptic processes according to classification of Klimenko [1999], which has been developed for this region. The classification is based on the revealing of typical cyclone and anticyclone trajectories at the Russian Plane and was applied for 100 year period. Fig. 6 shows interannual variation of the processes with cyclonic origin (P), which contribute to cloud formation and relative CQ_g changes within the analyzed territory. The correlation coefficients between the CQ_g at different sites are significant at 95% level and lie in the range of r=0.4-0.65, while there is strong inverse correlation of CQ_g against P variability with r =-0.68. Therefore we may reliably reveal the periods of relatively high CQ_g values in the past which cover the middle of 30s and the end of 60s due to attenuation of cyclonic activity over the Russian Plane. At the same time after 1976 there was a significant decreasing of CQ_g values almost at all sites over this region. The application of Fourier spectral analysis has revealed the periods of approximately 2, 4 and 9 years in variability of CQ_g at three sites, except Samara, similar to those in variations of P processes (Fig. 6b). The Samara peculiarity can be explained by its boundary location at the eastern part of the Russian Plane.

In order to compare the mean cloud and ozone effects on EW irradiance we use the following parameters:

where A_i(X ) was determined in (1), s _cl and s (Ai(X)) are the standard deviations of CQ transmittance and A(X). Thereby, V_cl and V_X characterize the normalized EW irradiance variations due to cloudiness and ozone within a 95% significant interval. D is the difference between the cloud and ozone effects on EW irradiance.

Fig.7 shows the EW variations due to ozone and cloud changes obtained from satellite and ground observations for different time scales over Eurasia. For most sites the EW variability due to ozone evaluated from long-term ground measurements (V_X ) lies within few percents with V_X obtained from TOMS. The maximum V_X is observed for ground ozone measurements and comprises 10-11%. EW changes due to cloud effects retrieved from ground cloud observations since the middle of 30s can reach 14-16%, and are mainly in agreement with TOMS V_cl values obtained for much less period. The sites with positive UV trend can be noticed over Southern Europe (42-44° N, the Mediterranean), over the local areas in Central Asia, where the effects of ozone and cloudiness are comparable, and the areas in Central Europe (50-52° N), which are characterized by high cloud variations. The observed EW increase over Eurasia is the combination of variability in ozone and cloudiness impacts.

Fig.6. a/ Interannual variations in cloud transmittance (CQ_g ) and in synoptic process responsible for cloud formation (P) according to [Klimenko, 1999] within the Atlantic- continental region of Eastern Europe, in percents; b/ Periodogram obtained by Fourier (spectral) analysis. May-September. Periodogram was set for the 1936-1990 period.

Fig. 8 shows the difference D=V_cl-V_X between the effects of clouds and ozone on EW variability. The cloud effects on EW variability can be significantly higher than ozone effects, but this phenomenon takes place only in local areas: over the Northern part of Europe, Northern Canada, the South-Eastern monsoon regions of Asia, etc., which are characterized by temporally erratic but intensive circulation processes. On the whole, the effects of ozone and cloud on EW variability are comparable and lie within ± 5% for most of the Earth’s territory. The effects of ozone slightly dominate in Southern Hemisphere(D<0%). During May-September period they also prevail over Northern Africa, Mediterranean region, and local areas inside the continents (see Fig.8a) covering low latitude zones, where EW doses are large (see the distribution of EW irradiance in Fig.7). During November-March period the area of prevailing ozone effects on EW irradiance decreases but still it covers a large territory.

Fig.7. EW irradiance over Eurasia (summer solstice, W/m2 ) and its changes due to cloud variability obtained from ground (light blue columns), from TOMS (blue columns) and due to ozone variations inferred from ground (light yellow columns), from TOMS (yellow columns). The sites with significant positive trend according to TOMS data are marked with pink background. May-September period.