Total ozone/UVB monitoring and forecasting: Impact of clouds and the horizontal resolution of satellite retrievals

Abstract

This study compares the horizontal resolution of solar backscatter ultraviolet 2 (SBUV2) total ozone (Ω) fields with those from the new NASA earth probe (EP) and Advanced Earth Observing Satellite (ADEOS) total ozone mapping spectrometer (TOMS) side‐scanning photometers. The latter instruments provide high resolution, easily resolving the medium‐scale waves (4–7 wavelengths around the Earth at a fixed latitude) that dominate day‐to‐day midlatitude Ω fluctuations. In contrast, SBUV2 instruments do not, since these devices measure only at nadir (straight downward), yielding ∼14 measurements daily at a given latitude. This method has consequences not only for global monitoring of Ω and ultraviolet B (UVB, 290–320 nm), but also for short‐timescale Ω and UVB predictions in summer because timescales of a few days are coupled to medium horizontal scales (several thousand kilometers) by baroclinic waves that typically force the observed Ω variations. We use a simple Ω prediction model to test the use of Ω fields from TOMS and SBUV instruments and show that the higher zonal resolution from side‐scanning TOMS instruments results in sizeable reductions in Ω prediction errors, whereas predictions using SBUV2 Ω are no better than persistence (where tomorrow's Ω is taken to be today's) in the biologically important summer months. Daily variabilities (equivalent to errors in 24‐hour persistence forecasting of Ω) in high‐resolution TOMS midlatitude ozone during summer are shown to sometimes exceed 50 Dobson units, producing daily changes of 20% or greater in computed ground‐level clear‐sky UV index. This study demonstrates that even these large daily changes in measured or predicted clear‐sky UV are usually smaller than daily UV changes associated with transient clouds. While surface UVB variability is dominated by local cloudiness variations, Ω forecasts can enhance UVB prediction in relatively cloud free regions such as the U.S. desert southwest and in stagnant high‐pressure regimes that can persist for 1–2 weeks in summer. Furthermore, as weather forecast models increase in accuracy of forecasted cloudiness, accurate predictions will allow more accurate UVB forecasts for cloud free regions, the locations where they are most needed. Results from the present paper show, however, that high‐resolution TOMS‐like side‐scanning Ω measurements are required for ozone and UVB monitoring and prediction, rather than SBUV‐type nadir observations.