In an attempt to find out why the all sky photometer can show spikes before and after local noon, I tried another experiment today. This time the photometer was tilted and oriented to place it's DB9 connector towards the south.
The results look much better, inspite of the intermittent cloud cover this afternoon. In addition, the sky has grown hazy again.
I will need to create a consistent angle of tilt before making this a standard procedure.