The difference between the mean RH values in the vicinity of thick versus shallow cloudy layers (Δ(RH)), as a function of the overline{mathrm{RH}} values, for seven selected continental (circles) and 6 maritime (squares) stations, for a 1 km layer depth (upper panel) and 2 km (lower panel) above the surface

<p><strong>Figure 2.</strong> The difference between the mean RH values in the vicinity of thick versus shallow cloudy layers (Δ(RH)), as a function of the \overline{\mathrm{RH}} values, for seven selected continental (circles) and 6 maritime (squares) stations, for a 1 km layer depth (upper panel) and 2 km (lower panel) above the surface. The data represents day time measurements during June–July–August, between 2000 and 2011.</p> <p><strong>Abstract</strong></p> <p>The hygroscopic growth of aerosols is controlled by the relative humidity (RH) and changes the aerosols' physical and hence optical properties. Observational studies of aerosol–cloud interactions evaluate the aerosol concentration using optical parameters, such as the aerosol optical depth (AOD), which can be affected by aerosol humidification. In this study we evaluate the RH background and variance values, in the lower cloudy atmosphere, an additional source of variance in AOD values beside the natural changes in aerosol concentration. In addition, we estimate the bias in RH and AOD, related to cloud thickness. This provides the much needed range of RH-related biases in studies of aerosol–cloud interaction.</p><p>Twelve years of radiosonde measurements (June–August) in thirteen globally distributed stations are analyzed. The estimated non-biased AOD variance due to day-to-day changes in RH is found to be around 20% and the biases linked to cloud development around 10%. Such an effect is important and should be considered in direct and indirect aerosol effect estimations but it is inadequate to account for most of the AOD trend found in observational studies of aerosol–cloud interactions.</p>