Occasional Thoughts on Climate Change

[bluewave]

Really extreme to see parts of France go +10° F for the first half of summer using the warmest 1991-2020 climate normals.

With a +4.2°C deviation from the already warmed 1991-2020 baseline, the tally for the first half of the meteorological summer is eye-opening. The 2026 anomaly forces us to scale up the graph, to the point of literally compressing 2003 (+2.8°C "only"). ➡️From the Center-West to Burgundy, we're seeing unprecedented values of +5 to +6°C. ➡️The rainfall deficit is just as stark, ranging from -30% in the north to -100% in the Var. The Center-West, where the vegetation is already completely scorched, hovers between -60 and -80%. ➡️Over the past month, nearly every day has been spent above the heatwave threshold, with only a 4-day pause at the start of July. ➡️The third heatwave should wrap up this weekend. But in the southwest, temperatures will remain heatwave-level (34 to 38°C) for an indeterminate period. Given the hot air mass south of Europe, we're far from safe from a 4th heatwave later in the summer.

 

France is dry and scorched. After already surpassing 2003 with the hottest day ever observed in France, the year 2026 now exceeds the droughts of 1976 and 2022. The soil moisture index reaches its lowest level ever measured for a July 9. And the situation continues to worsen. A new hairdryer effect is expected in the central west starting Sunday. From the sky, France appears literally burned: parched meadows, defoliation of forests, summer crops in great distress (corn, soy, sunflower), and early winter harvests contribute to this color. As if that weren't enough, we have just broken the NATIONAL RECORD for the hottest night ever observed with 30.6°C in Céret in the Pyrénées-Orientales at the "coolest" point of the morning. Photo from the NAOO-21 / VIIRS satellite (true color).

 

[frontranger8]

3 hours ago, bluewave said:

Really extreme to see parts of France go +10° F for the first half of summer using the warmest 1991-2020 climate normals.

With a +4.2°C deviation from the already warmed 1991-2020 baseline, the tally for the first half of the meteorological summer is eye-opening. The 2026 anomaly forces us to scale up the graph, to the point of literally compressing 2003 (+2.8°C "only"). ➡️From the Center-West to Burgundy, we're seeing unprecedented values of +5 to +6°C. ➡️The rainfall deficit is just as stark, ranging from -30% in the north to -100% in the Var. The Center-West, where the vegetation is already completely scorched, hovers between -60 and -80%. ➡️Over the past month, nearly every day has been spent above the heatwave threshold, with only a 4-day pause at the start of July. ➡️The third heatwave should wrap up this weekend. But in the southwest, temperatures will remain heatwave-level (34 to 38°C) for an indeterminate period. Given the hot air mass south of Europe, we're far from safe from a 4th heatwave later in the summer.

 

France is dry and scorched. After already surpassing 2003 with the hottest day ever observed in France, the year 2026 now exceeds the droughts of 1976 and 2022. The soil moisture index reaches its lowest level ever measured for a July 9. And the situation continues to worsen. A new hairdryer effect is expected in the central west starting Sunday. From the sky, France appears literally burned: parched meadows, defoliation of forests, summer crops in great distress (corn, soy, sunflower), and early winter harvests contribute to this color. As if that weren't enough, we have just broken the NATIONAL RECORD for the hottest night ever observed with 30.6°C in Céret in the Pyrénées-Orientales at the "coolest" point of the morning. Photo from the NAOO-21 / VIIRS satellite (true color).

 

Thankfully, it appears some relief is imminent.

[chubbs]

Good 2-part blog from last year on Canadian wildfire history outlining the role of forest management and climate. Canadian forest fires are less numerous than in the past; but, larger and burning a wider area. The cause of fires has transitioned from mostly man-made in the past to mostly ignited by lightning today.

https://thetradeoff.substack.com/p/north-americas-forests-used-to-burn

https://thetradeoff.substack.com/p/part-2-many-of-north-americas-forests

[gallopinggertie]

Current SST anomalies and contour chart. The western Mediterranean Sea is 5C above normal, with temperatures over 30C south of Sicily! Even the Black Sea is pushing into the range of temps warm enough to support tropical cyclones. 

[donsutherland1]

A “must read” piece on climate, data, and data manipulation:

https://www.theclimatebrink.com/p/hot-days-cold-thermometers

[GaWx]

8 hours ago, donsutherland1 said:

A “must read” piece on climate, data, and data manipulation:

https://www.theclimatebrink.com/p/hot-days-cold-thermometers

Thanks, Don! 

 This is the first time I can recall becoming aware of this observation time bias potentially causing overcounting of high max temps due to double counting of these highs as a result of observers getting 24 hour max prior to 1940s during late afternoon, near the time of most very hot highs, as opposed to mainly doing it near sunrise 1940s+.

 Do you have a source with more details on this? How widespread was this practice of getting 24 hour max late in afternoon? Do you by chance know how often this would result in double counting hot highs?

[donsutherland1]

17 minutes ago, GaWx said:

Thanks, Don! 

 This is the first time I can recall becoming aware of this observation time bias potentially causing overcounting of high max temps due to double counting of these highs as a result of observers getting 24 hour max prior to 1940s during late afternoon, near the time of most very hot highs, as opposed to mainly doing it near sunrise 1940s+.

 Do you have a source with more details on this? How widespread was this practice of getting 24 hour max late in afternoon? Do you by chance know how often this would result in double counting hot highs?

This paper has some information on time of observation bias. Hundreds of stations once recorded afternoon resets.

https://www.ncei.noaa.gov/pub/data/ushcn/papers/menne-etal2009.pdf

[chubbs]

7 hours ago, GaWx said:

Thanks, Don! 

 This is the first time I can recall becoming aware of this observation time bias potentially causing overcounting of high max temps due to double counting of these highs as a result of observers getting 24 hour max prior to 1940s during late afternoon, near the time of most very hot highs, as opposed to mainly doing it near sunrise 1940s+.

 Do you have a source with more details on this? How widespread was this practice of getting 24 hour max late in afternoon? Do you by chance know how often this would result in double counting hot highs?

We have a good example in Chester County PA. Here are high temperatures for July 1934 and 1936 for 4 local stations. Phoenixville ran much warmer than the other stations probably due to poor shelter placement. Phoenixville also carried over hot temperatures to the next day when the other stations cooled. This shows that the Phoenixville max/min thermometer was flipped near the time of peak temperature while the other stations were not. Phoenixville completely dominates the # of 95 and 100F days in Chester County in the 1930-40s, but the data is spurious. This is why I don't trust plots of historic max temperatures  in the US, unless the data has been scrutinized for bad data. Only takes one or two  bad apple stations to create a spurious result. 

[donsutherland1]

2 hours ago, chubbs said:

We have a good example in Chester County PA. Here are high temperatures for July 1934 and 1936 for 4 local stations. Phoenixville ran much warmer than the other stations probably due to poor shelter placement. Phoenixville also carried over hot temperatures to the next day when the other stations cooled. This shows that the Phoenixville max/min thermometer was flipped near the time of peak temperature while the other stations were not. Phoenixville completely dominates the # of 95 and 100F days in Chester County in the 1930-40s, but the data is spurious. This is why I don't trust plots of historic max temperatures  in the US, unless the data has been scrutinized for bad data. Only takes one or two  bad apple stations to create a spurious result. 

Yes, Phoenixville had a 4 pm time of observation: