July Heat Set U.S. Record

gollum123 sends this excerpt from CNN: "The July heat wave that wilted crops, shriveled rivers and fueled wildfires officially went into the books Wednesday as the hottest single month on record for the continental United States. The average temperature across the Lower 48 was 77.6 degrees Fahrenheit, 3.3 degrees above the 20th-century average, the National Oceanographic and Atmospheric Administration reported. That edged out the previous high mark, set in 1936, by two-tenths of a degree, NOAA said. In addition, the seven months of 2012 to date are the warmest of any year on record and were drier than average as well, NOAA said. U.S. forecasters started keeping records in 1895. And the past 12 months have been the warmest of any such period on record, topping a mark set between July 2011 and this past June. Every U.S. state except Washington experienced warmer-than-average temperatures, NOAA reported."

Comments

Re:Hopefully it's an outlier

[Internetuser1248]

No it is god punishing you for using fahrenheit.
25.33 degrees for those that both care and didn't already type '77.6f in c' into google.

Re:Hopefully it's an outlier

[AlienIntelligence]

I love it.... "Take the 1936 Texas below normal temperature out of the mix and there goes your 0.2F record making difference with July 2012." Of course, if you randomly take out data points you don't like, you're going to get the result you're looking for. Not to mention that their entire post focuses on the fact that not all states all linearly increased in temperatures, which betrays a complete lack of understanding of how temperatures are come about.

FWIW, a graph tends to be of more value if you evaluate and potentially take out outlier points.
If you are looking for trends. Also, some toss lowest and highest as well.

Just saying.

-AI

Re:Hopefully it's an outlier

[jcoy42]

No, poster is correct. It's called winsorising. It's common to toss out the top and bottom 5% just to discount anomalies.

But you don't discount it after you see the data because you don't like it, you plan to discount it before you collect the data and more importantly you do it indiscriminately and equally on both sides of the data set. Not just points you don't like after you see the data.