Some weird warm bits of New Hampshire out along the western hills

(See Update at the bottom, with explanation from the paper's weather guru.)

I'm doing a story for the Telegraph about the updated plant hardiness map from the U.S. Department of Agriculture, which reflects the fact that average lowest temperatures during teh winter are warmer than they used to be. The picture shows NH's listings by plant zone ... the weird thing are those green splotches running north-south, roughly from Keene to Sunapee, green signifying about areas with average lows about 5 degrees warmer than the surrounding areas.

Why are there slightly warmer areas (that is, areas with a higher average-low-winter-temperature) along there? It looks like they're associated with the line of low mountains/high hills that end up at the Sunapee Ski Area, which includes the ridgeline holding Lempster Mountain Wind Farm.

I suppose the mountains funnel air currents to the point that they create microclimates, but it's interesting to see.

UPDATE: Doug Webster, a meteorologist who writes our monthly weather column, explains. The map measures the average of low temperatures, not overall averages, and for places without tall mountains (the mountains in southwestern NH rarely top 3,000 ft), the coldest spots are valleys where cold air collects during cloudless winter nights. On average, ridgetops won't have temperatures as low as valleys (this doesn't count wind chill, of course).

In the White Mountains, it's a different matter: They're tall enough that they're seriously affected by the fact that temperatures decline 5.5 degrees per thousand feet elevation in dry air).