Is the surge working if casualties always drop in winter?

Numerous blog posts — from Attytood to dkos — have questioned the validity of right wing talking points claiming the surge is lowering violence and US casualties in Iraq and that we should allow additional time for ‘the plan’ to work.

Few comments have mentioned that according to statistics available through icasualty.org casualty rates in Iraq have always dropped in the winter.

I recall this point being made when the idea of a surge was first proposed, and the thought that the surge might have been proposed to take advantage of such a decrease in violence and casualties, both from tactical and public relations stand points.
A graph with only forty-eight data points isn’t going to be the smoothest curve, but if you’ll looks at these two representations of US deaths and US wounded in Iraq, you’ll see the downward trend every November through March:

http://icasualties.org/oif/US_chart.aspx
http://icasualties.org/oif/woundedchart.aspx

So an important qualifier of the effectiveness of the surge would have to include the relationship of how much violence and casualties have dropped the previous three winters as compared to this winter.

A quick glance shows that US fatalities remain consistant, if not slightly higher this winter.  US wounded numbers may not be complete, as the February number seems abnormally low.  (If that number is correct, if would indictate a vast improvement, which would be good news indeed.)