Abstract. In this paper we use a novel observational approach to investigate MODIS satellite retrieval biases of τ and r_e (using three different MODIS bands: 1.6, 2.1 and 3.7 μm, denoted as r_e1.6, r_e2.1 and r_e3.7, respectively) that occur at high solar zenith angles (θ₀) and how they affect retrievals of cloud droplet concentration (N_d). Utilizing the large number of overpasses for polar regions and the diurnal variation of θ₀ we estimate biases in the above quantities for an open ocean region that is dominated by low level stratiform clouds.

We find that the mean τ is fairly constant between θ₀ = 50° and ~65–70°, but then increases rapidly with an increase of over 70 % between the lowest and highest θ₀. The r_e2.1 and r_e3.7 decrease with θ₀, with effects also starting at around θ₀ = 65–70°. At low θ₀, the r_e values from the three different MODIS bands agree to within around 0.2 μm, whereas at high θ₀ the spread is closer to 1 μm. The percentage changes of r_e with θ₀ are considerably lower than those for τ, being around 5 % and 7% for r_e2.1 and r_e3.7. For r_e1.6 there was very little change with θ₀. Evidence is provided that these changes are unlikely to be due to any physical diurnal cycle.

The increase in τ and decrease in r_e both contribute to an overall increase in N_d of 40–70% between low and high θ₀. Whilst the overall r_e changes are quite small, they are not insignificant for the calculation of N_d; we find that the contributions to N_d biases from the τ and r_e biases were roughly comparable for r_e3.7, although for the other r_e bands the τ changes were considerably more important. Also, when considering only the clouds with the more heterogeneous tops, the importance of the r_e biases was considerably enhanced for both r_e2.1 and r_e3.7.

When using the variability of 1 km resolution τ data (γ_τ) as a heterogeneity parameter we obtained the expected result of increasing differences in τ between high and low θ₀ as heterogeneity increased, which was not the case when using the variability of 5 km resolution cloud top temperature (σ_CTT), suggesting that γ_τ is a better predictor of τ biases at high θ₀ than σ_CTT. For a given θ₀, large decreases in r_e were observed as the cloud top heterogeneity changed from low to high values, although it is possible that physical changes to the clouds associated with cloud heterogeneity variation may account for some of this. However, for a given cloud top heterogeneity we find that the value of θ₀ affects the sign and magnitude of the relative differences between r_e1.6, r_e2.1 and r_e3.7, which has implications for attempts to retrieve vertical cloud information using the different MODIS bands. The relatively larger decrease in r_e3.7 and the lack of change of r_e1.6 with both θ₀ and cloud top heterogeneity suggest that r_e3.7 is more prone to retrieval biases due to high θ₀ than the other bands. We discuss some possible reasons for this.

Our results have important implications for individual MODIS swaths at high θ₀, which may be used for case studies for example. θ₀ values > 65° can occur at latitudes as low as 28° in mid-winter and for higher latitudes the problem will be more acute. Also, Level-3 daily averaged MODIS cloud property data consist of the averages of several overpasses for the high latitudes, which will occur at a range of θ₀ values. Thus, some biased data are likely to be included. It is also likely that some of the θ₀ effects described here would apply to τ and r_e retrievals from satellite instruments that use visible light at similar wavelengths along with forward retrieval models that assume plane parallel clouds, such as the GOES imagers, SEVIRI, etc.