This article is from the Ozone Depletion: UV Radiation and its Effects FAQ, by Robert Parson email@example.com with numerous contributions by
06 How does UV-B vary from place to place?
A great deal. It is strongest at low latitudes and high altitudes.
At higher latitudes, the sun is always low in the sky so that it takes
a longer path through the atmosphere and more of the UV-B is absorbed.
For this reason, ozone depletion is likely to have a greater impact on
_local_ ecosystems, such as terrestrial plants and the Antarctic
marine phytoplankton, than on humans or their livestock. UV also
varies with altitude and local cloud cover. These trends can be seen
in the following list of annually-averaged UV indices for several US
cities [Roach] (units are arbitrary - I don't know precisely how this
index is defined though I assume it is proportional to some integral
over the UV-b region of the spectrum)
Minneapolis, Minnesota 570
Chicago, Illinois 637
Washington, DC 683
San Francisco, California 715
Los Angeles, California 824
Denver, Colorado 951
Miami, Florida 1028
Honolulu, Hawaii 1147
The effect of clouds on local UV-B irradiance is not straightforward
to determine. While the body of a cloud attenuates the radiation,
scattering from the sides of a cumulus cloud can actually enhance it.
[Mims and Frederick 1994.]
In comparing UV-B estimates, one must pay careful attention to
exactly what is being reported. One wants to know not just whether
there is an increase, but how much increase there is at a particular
wavelength, since the shorter wavelengths are more dangerous.
Different measuring instruments have different spectral responses,
and are more or less sensitive to various spectral regions. [Wayne,
Rowland 1991]. Wavelength-resolving instruments, such as the
spectroradiometers being used in Antarctica, Argentina, and Toronto,
are particularly informative, as they allow one to distinguish the
effects of ozone trends from those due to clouds and aerosols.
[Madronich 1993] [Kerr and McElroy]. When wavelength-resolved
data are available, they are frequently convolved with an "action
spectrum" that is relevant for a particular biological influence.
Thus the "erythemal action spectrum", designed to estimate the
tendency of UV radiation to redden human skin, places less emphasis
on short wavelengths that the action spectrum designed to estimate
the tendency of UV to damage DNA. When the ozone column overhead
decreases by 1%, erythemal UV increases by about 1% while DNA-damaging
UV increases by about 2.5%. [Madronich 1993] The widely-used broadband
Robertson-Berger meter has a spectral response that is close to
the erythemal action spectrum.