Comparison of daily UV doses estimated from Nimbus 7/TOMS measurements andground-based spectroradiometric data

During recent years, methods have been developed for estimating UV irradian ce reaching the Earth's surface using satellite-measured backscattered UV r adiances. The NASA-developed method is based on radiative transfer calculat ions and satellite measurements of parameters affecting UV radiation: extra terrestrial solar irradiance, atmospheric ozone, cloud reflectivity, aeroso l amounts, and ground albedo. In this work a comparison is made between dai ly UV erythemal doses estimated from Nimbus-7/TOMS measurements (from 1991 to May 1993) and those calculated from ground-based spectroradiometer data. Three stations operated by the National Science Foundation were chosen for this comparison: Ushuaia, Argentina (for 573 days), Palmer, Antarctica (fo r 450 days), and San Diego, California, (for 149 days). These stations were selected to illustrate the differences between ground-based measurements u sing the same type of instrument, SUV-100 double monochromator spectroradio meters, and satellite estimates of surface UV irradiance under three differ ent environmental conditions (mountains and snow, nearly continuous snow co ver, and midlatitude urban sea level conditions). Averaging the measured an d TOMS-estimated doses over periods from 1 week to 1 month improves the agr eement. The daily or monthly mean bias increases during months when there i s snow/ice on the surface. TOMS has a larger estimate of the UV irradiance by 25% at San Diego (no snow), in agreement with the summer-month analysis of Toronto irradiances [Herman et al., 1999]. TOMS underestimates the avera ge daily-UV dose at Ushuaia (monthly mean bias of -13%) and at Palmer (-35% ) consistent with snow/ice with cloud effects not being properly accounted for in the TOMS algorithm. When the reflectivity at all three sites is low (no snow), the TOMS irradiance estimate is larger than the SUV-100 measurem ents consistent with previously analyzed Brewer data at Toronto. The effect s of local fog or clouds smaller than the satellite field of view and undet ected UV-absorbing aerosols near the ground are discussed. In addition to u ncertainties in radiometric calibrations of the spectrometers, none of the SUV-100 data are corrected for deviations of diffuser-transmittance from tr ue cosine response.