If the sun kicks out roughly the same amount of energy at all times (is there a way to find actual amount at any given time? eg. from satellite data?), and we know where it is in relation to earth (more specifically: in relation to sensor array) as part of heliotrope calculations, and we know how much should land on earth if the sky is clear... it should be possible to estimate aerosols (precipitation, dust, etc) in the atmosphere based on the difference between actual and expected pyrheliometer direct (collimated) readings. The oktameter readings could be used to get some idea of ratio between cloud cover vs. other types of aerosols.