Mismatch between variations of solar indices, stratospheric ozone and UV-B observed at ground

In solar cycles 22–23, all solar indices showed maxima near 1990 and 2000 and minima in 1996. The maximum to minimum variation was only 1–2% in the UV range 240–350 nm. Dobson ozone intensities did not show any clear relationship with solar cycle and ozone variations were less than 10%. The UV-B (295–325 nm) observed at ground by Brewer spectrophotometers at some locations had variations of 50–100% for 295–300 nm, and 20–50% for 305–325 nm. The maxima were in different years at different locations (even with separations of only 300 km), did not match with the solar cycle, and were far too large to be explained on the basis of ozone changes (1% decrease of ozone is expected to cause 2% increase of UV-B). Thus, if the data are not bad, the UV-B changes do not match with solar activity or ozone changes and must be mostly due to other local effects (clouds, etc.?). When data are averaged over wide geographical regions, UV-B variation ranges are smaller (10–20%, probably because localised, highly varying cloud effects get filtered out), and are roughly as expected from ozone variations.