Yeah, I picked A because that first sentence suggests that New Yorkers are taking their observation of seasons and generalizing them to be true for all places. And I also failed to realize that the extreme temperature could mean unusual temperatures for the season. i was looking at it like this.
for the three years: trend of temperature.
<----winter temp____________0 degrees___________summer temp------>
for the two years: temp trends
_____winter temp----------------->0 degrees________summer temp------>
(E) it is.
(A): There is no relation drawn in the question between New York and other locations. Incorrect.
(B): There is no previous argument here. Incorrect.
(C): This is true (just five data points do not constitute conclusive evidence), but it does not present a fatal flaw in the author's reasoning. Incorrect.
(D): There is no causal relationship being assumed, just correlations. Incorrect.
(E): CORRECT. The coldest winters were followed by the hottest summers, and the hottest winters were also followed by the hottest summers. In all cases, extreme temperatures in winter were followed by extreme temperatures in summer. The author ends up supporting the claim he/she wants to refute.
The catch in option (E) is to realize that extreme temperature in winters need not mean extreme cold - it could also mean unnaturally hot weather in winter, uncharacteristic of the season, and so considered extreme.