rainycloudthermoQMIn the mad lead-up to the latest UNFCCC 21st Conference Of Partners (COP21), I thought I’d look at French weather station records to see how quickly France has been progressing toward climatological catastrophe. In particular, I wanted to see long-term trends that extend to the present.

Looking through the inventory of GHCN-Daily datasets for stations that began recording both TMAX and TMIN prior to 1930 and are still recording today, I found only 8 stations that qualified, including 3 of France’s 6 Global Climate Observing System Surface Network (GSN) stations.

I downloaded the “raw” daily data for those stations from NOAA and ran it through an Excel VBA macro that extracts annual averages and yearly extremes as JavaScript variables for an online viewer I’ve been writing (work in progress). After incorporating all this exciting data into my viewer page, I looked for recent trends.

Disappointingly, there were no recent trends to be seen.

Except for breaks during the World Wars, these 8 stations displayed near-continuous daily records…but only up to a point. Four of the stations’ records ended with the year 2000 and four with 2004. After that…blank.

By default, my viewer ignores years with more than 7 days of data missing from any month. When I opted to include years with up to 28 days missing data in any month, lo and behold, all 14 years missing from Perpigan’s 21st century record  appeared:


The other 7 stations displayed from 10 to 13 of the missing years.

How much data is missing? I created an Excel macro to count missing-data days for all 8 stations from January 1, 1930 to November 13, 2015 and graphed them:


Except for some shutdowns during WWII, these records are amazingly complete until 2001. No station had more than two days missing data in any year from 1946 through 2000. Then, for some in 2001 and others in 2005, the data suddenly becomes very sparse.

I then examined the 3 French GSN stations that began recording after 1929 and found the same issue. Data from Bourges Aerodrome became sparse in 2001, after which there were 74 to 246 missing days per year. Data from Mont Aigoual and Strasbourg Entzheim became sparse in 2005, after which there were 107 to 258 missing days per year.

Quite bizarre. Are the French waging war on climate records?