One of the interesting products that was generated by lowering the restrictions to the source of the data was an increased resolution on sampling. Also a increased frequency on sampling was noted. This is not a higher accuracy but just more points to plot and make a interesting graph. I am sure more significant studies are being preformed to figure if this is even worth looking at from a meteorological standpoint. As a demonstration of the need to move data collection and storage into the cloud this dataset opens a new resource. Below is plotted a single timestep temperature contour with wind overlaid as arrows. Stations are plotted in blue and areas of poor coverage(the lake) tend to have poor interpolations.
This contour plot was then preformed over a 48 hour period.
The animation is provided as a TimeLapse.
A slice was created through the city to try and identify any increase heating effect.
Again this is not attempting to construct conclusions but only to look at the raw data.
These three products were able to be constructed within seconds of the data
being available in the NetCDF files. For complex queries data was loaded in to
a sql server and plotting was preformed directly against the optimized sql server.