AzSz88, Nick,
I have a graphics programme that can stack and in which I can change opacity, but that's not the same as averaging.
I guess the thing to do would be to use the native concentration files and work out something like a past 7 day average, this would be a very primitive first step towards the sort of persistence model that Andrew Slater does. I'd have to code up the image drawing*, but the rest of the code I have already got, the averaging operation is trivial.
Yes, now I check ImageJ my memory does serve me well, there is an average option in ImageJ, Process>Image Calculator - set option average. But you can only do average of two images and the result isn't of much use with Bremen, might be better with a true monocolour bitmap.