Exposure simulations using radiative transfer modelling

All simulations were done using the libRadtran software (Mayer and Kylling, 2005).  This model has been extensively validated in the past both with measurements (Mayer et al., 1997) and against other models (Van Weele et al., 2000).  In general, the model provides good agreement with measurements under cloudless skies (within 5%).

Exposure indicator

The annual erythemal doses and average daily erythemal doses were selected as exposure indicators, in order to match the metrics used in published exposure-response functions.  The erythemal dose is based on the action spectrum for erythema (sunburn). (The software can be modified to handle other dose types as well.)

Exposure distribution

Measurements of ambient UVR only give an indication of 'possible' UVR exposure of a population; the relationship between an outcome and the risk factor occurs at an individual level.  However, ambient UVR does not easily translate to actual population exposure distribution, because of differences in local UV levels, individual time-activity patterns and dress behaviour.  Detailed information on personal exposure and exposure distribution was not applied here; ambient UVR is thus used as the best available proxy of population exposure, and is also in accord with the available ERFs. 

Current day exposure data

Data on current daily exposure were required both to define baseline conditions, and as a means of validating the models.  

For this purpose, daily erythemal doses for years 1997-2003 were estimated from the TOMS satellites.  Daily total ozone data were likewise obtained from TOMS for the same period. 

Modification factors needed to be estimated to allow for the effects of non-clear sky conditions: for erythemal doses, the application of modification factors has been validated (e.g. Schwander et al. 2002).  Daily clear-sky erythemal doses were therefore estimated for the total ozone columns, assuming cloudless conditions and a black non-reflecting surface.  Cloud/aerosol/albedo modification factors were then estimated by:

• adding all the matching daily erythemal doses from the TOMS satellite data;
• likewise adding all the matching clear-sky daily erythemal doses;
• taking the ratio of the two cumulated doses over the seven year period (1997-2003).

Provisionally, modification factors were assumed to remain constant for the next century.

Future UV predictions

The same approach was used to estimate future clear sky erythemal doses, for the years 2000, 2030, 2050 under the IPCC 2001 scenarios A2 and B1.  Future global ozone fields for these conditions were derived from the Oslo Chemical Transport Model (Rummukainen et al., 1999).   The clear sky erythemal daily doses were then rescaled, using the fixed modification factors to obtain realistic estimates of ambient UV exposure.