meehl? >> choose to add to that. this analysis we did we're looking at basically temperature records the second half of the 20 century from weather stations that had good data records. this is a bigger problem than just a thermometer problem. have stations reporting their daily high to mature and get no temperature, every day so you have a lot of data records. this ratio which is now to the one which we thought was kind of odd, without initially in fact this came from a guy at the weather channel because he was noticing this, keeping track of records on his own as a meteorologist. and i was down there, heidi invited me down to, and he said what's with his two to one ratio? i said i do know that he says is that some kind of a unique thing about climate change? i said i have no idea. he said let's look at it. it turns out this ratio, we just happen to be at about two to one right now, a decade ago it was a little less than two to one. a decade before it was a little less than that. if you had