![](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjmh6O2RRNyzm8asb1qb3CGcsTMk9fjrEjQLENBZYFlh6nd4wwew4lxQPwodLnLoXvEHnd3kctuTeZNJ0YJEd0ezQh37fQg6MUUg5GPjjzG34d7scmUXb1CzTDrLR7pnwe8dThRaRTnjZE/s320/temp.1108.jpg)
![](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgCY4bYtvX-j1o67gYyID51xaXJdBXBLRulOMCY3VdzS-hiC4oGC7x8iZc5ebeq_GO513zrrOfNip61iA0J0pVwRBm0VSibGrU82ynNCLS7JKAgalZnLBeyr2O2xODVVBSu-b0HwAAT9EM/s280/nov.djf.2008.jpg)
Does a relatively cold November predict a cold winter? The lower chart to the right shows the relationship between the November average temperature and the following December-January-February average since 1871. The upward sloping trend line indicates that cold Novembers tend to be associated with relatively colder winters, but the relationship is not very strong. The equation of the trend line indicates that for every 1° change in the November average, the winter average moves in the same direction by about 0.4°. On the other hand, there is a lot of scatter in the data. The "R-squared" says that only 19% of the variability of the winter average temperature is accounted for by the November average.
For more details on regression analysis and a discussion of the relationships between temperature and precipitation, see these blasts from the past:
- As December Goes . . .
- Winter Marches On
- Round Rodent vs. Regression
- Who'll Start the Rain?
- Chilly Still on the Menu
Images: CapitalClimate charts from NWS data, photos © Kevin Ambrose
No comments:
Post a Comment