Current Mie files for water clouds might be not accurate enough. It should be investigated better, what size distribution is required to calculate accurate phase functions without wiggles. The standard resolution of 0.03 maximum bin size for size parameter is not sufficient foc calculations in the visible. With size parameter 0.003 the results look smooth but take extreme computation time (appr. 8 days for wavelengths below 300 nm).
Calculations for dx=0.03 and dx=0.003 are in /home/users/emde/optical_properties/wc/mie_wc_20100917/, in the directory “plots” there is a plotting script for plotting the phase functions for various effective radii.
Sometimes the miev0 function produces NaN values (for individual input parameters), if this happens for one size in the size distribution, the final result is one. For the fine size distributions this happened for 260, 270, 350, 390, 400, 500, 580 nm (reproducable).