temperature seasonal patterns? Stats method?

I have collected 6 years of data and last few months written some basic weather statistics display and calculating modules in SciLab (similar to MathLab and similar tools).

I start to see that there may be some “patterns” recurring each year, at least when it comes to temperatures.

Already from the beginning I decided to be pretty careful about what I did, and since it had to do with temperatures I decided that I would have winter and summer solstices and the spring and autumn equinoxes as zero points for offsets - just because they change by date over the years, and also to eliminate problems around 29th of February.

I think one can see that there may be some patterns if looking at the two graphs I have attached - and this even though this winter was extra ordinary cold in December and perhaps first half of January, the patterns still may be there.

So, my question is, what would be the best STATISTICAL MERTHOD to use to find patterns, in this case temperature patterns?

Regards, Ingemar


overlaid_spring.png

overlaid_winter.png