WaPo Cherry Blossom Claims Refuted: “Nothing But Lies And Statistical Manipulations”!

via NoTricksZonehttp://notrickszone.com

Reader David Reich left a comment  in response to Kenneth Richard’s post on grape harvests and climate.

I’ve decided to upgrade it as a post below. Both stories show that today’s climate is well within the range of our climate’s natural variability over the past 100 and 1000 years, and that today’s weather events aren’t unusual.

By USDA photo by Scott Bauer, United States Department of Agriculture, Public Domain

================================================
Washington DC Cherry Blossoms, By David Reich

This [Grape Harvest Date Evidence: No Significant Modern Warmth] reminds me of the Washington Post (WaPo) story that ran a story a few days ago about the earlier than normal Washington DC cherry blossoms hitting their “peak day”. It was stated by the National Park Service that the blossoms are now blooming “on average about 5 days earlier than normal” since records have been kept by NPS.

So, I went back and checked the data. Turns out that the data have been kept for 96 years going back to 1921. The average day of hitting “peak” during the decade of the 1920’s was day 93 into the year – April 3 in a non-leap year, April 2 in a leap year. The average in this decade so far is………92 days. So, how does the Park Service come up with “about 5 days”?

If you average all 96 days, you do get close to 94 days largely because during the 1950’s and 1960’s, the peak date in those 2 decades was over 97 days. So, the recent data is only 2 days earlier than the average. BUT, since the 50’s and 60’s decades which many people remember, the average has in fact gone down about 5 days. Of course no mention that the average went up from 93 days in the 1920’s to 97 in the 1950’s and 1960’s.

Equally egregious with both the WaPo article and also the NPS claim is the complete ignoring of the fact that the standard deviation of all 96 data points is well over 7 days (1990 was the earliest peak at day 74 while 1958 was the latest at day 108) so the data varies wildly. So, even if a 5 day trend were valid, such a trend is well within the normally expected variation in the “peak” day and is thus no trend. No change.

Nothing but lies and statistical manipulation by the WaPo and the NPS.”

via NoTricksZone http://notrickszone.com

March 7, 2017 at 05:21AM