Spatial and temporal variability of martian water-ice cloud effective radius in EMIRS thermal infrared observations

New analyses of thermal infrared spectra obtained from the Emirates Mars Infrared Spectrometer (EMIRS) allow for retrieval of water-ice cloud effective radius, r_eff,ice, in the lower atmosphere of Mars across spatial, seasonal, and daytime diurnal scales. Using a one Martian-year observational dataset the mean retrieved r_eff,ice is 3.1 μm (standard deviation, σ = 1.5 μm) for an assumed effective variance, ν_eff,ice = 0.1, and with most individual values between 1 and 10 μm. Sensitivity analyses show the uncertainty associated with r_eff,ice is largely inversely proportional to water-ice cloud optical depth, τ_ice. As a result, our retrieval is generally applicable to water-ice clouds with τ_ice of at least ∼0.02 to 0.08 at 825 cm⁻¹. Seasonal results show mean r_eff,ice values of 4.3 μm (σ = 1.9 μm) during aphelion season (Ls = 40°–140°), associated with larger particle sizes in the aphelion cloud belt (ACB), and 2.5 μm (σ = 0.8 μm) during perihelion season (Ls = 225°–360°). Diurnal analyses between roughly 06:00 and 20:00 local true solar time indicate the largest variability occurs in ACB clouds, with effective radii tending to be smaller during midday (zonal-mean r_eff,ice of 4–5 μm) as compared to the mornings and afternoons (zonal-mean r_eff,ice of 6–7 μm). Relatively large spatial gradients in r_eff,ice within the ACB are observed, including a seasonal mean r_eff,ice of ∼8 μm in the Tharsis region and regions in the southern mid-latitudes with seasonal mean r_eff,ice of ∼2 μm. Spatial variability in r_eff,ice tends to be smaller throughout the rest of the year, though with some similar patterns. This includes several time periods between regional dust storms when a low-latitude band of water-ice clouds formed in the afternoons near Tharsis, which grew in both optical depth and effective radius to maxima in the early evenings. Lastly, changes in r_eff,ice are found to be generally positively correlated with changes in τ_ice and negatively correlated with estimated cloud height, though r_eff,ice distributions also show this to be a more complex relationship than simple averages might suggest.