2025-2026 ENSO

[jm1220]

3 hours ago, mitchnick said:

Negative PDO may have peaked with notable cooling around Japan over the past 15 days.

Good. Hopefully lots of typhoons in that area this summer. 

[GaWx]

5 hours ago, so_whats_happening said:

Subsurface for June and the first half of July.

Interesting dichotomy between the warm neutral subsurface and the cold neutral surface:

[40/70 Benchmark]

37 minutes ago, bluewave said:

Narrowly focused into the Bay Area in an otherwise sea of warm across the Northern Hemisphere this summer.

 

"40/70....I am your father" (breathes heavily)

[40/70 Benchmark]

22 minutes ago, GaWx said:

Interesting dichotomy between the warm neutral subsurface and the cold neutral surface:

Probably warm ENSO for 2026-2027.

[bluewave]

2 hours ago, 40/70 Benchmark said:

@bluewaveSeriously, though....I do recall telling you last season that I expected your October MJO indicator to be amplified and favorable and I do expect that again, so we'll see.

We have only had one October really amplified MJO 5-6 during each multiyear La Niña going back to 2010. October 2024…+2.76…October 2020…+2.81….October 2017…+3.35….October 2010….+2.88.

The other La Niña years surrounding these in each group had a weaker October MJO in 5-6 like in 2022, 2021, 2016, and 2011. So the more amplified October years featured La Niña +PNA mismatches.

With the exception of last year, these were very snowy winters. But the WPAC pattern and Pacific Jet never relaxed last winter like all the previous mismatch winters did. This is what I was pointing out last year why I mentioned early on that there were competing influences which didn’t exist during the other winters. So not to expect the same type of outcome.  

So this would be a first time occurrence if the MJO 5 peaked again this October in the +2.76 to +3.35 range. Seems like it’s some type of fall forcing event which affects the winter PNA during La Ninas. 

It’s why my guess a few months back that this 25-26 winter will be warmer than last winter was with the PNA averaging less positive than last winter did. But since the snowfall was so low last winter, it wouldn’t take much for one decent snow event to surpass last seasons snowfall totals from around Philly to Boston.

Plus it’s possible that we could get a least one winter month with a decent +PNA like we saw in January 2022 even though the PNA was strongly negative in December tilting the whole winter -PNA. That was largely driven by the MJO 8. January 2022 was the last winter month around NYC which was both cold and snowy especially Long Island.

We have also seen very impressive 500 ridges in Canada since the 2023-2024 El Niño generally boosting the +PNA. So we’ll have to wait for the October verification this year to know if it will be like  past multiyear La Ninas following the mismatch winters like we had in 2024-2025. Always have to leave open the possibility of a first time occurrence with back to back mismatch events. But this hasn’t happened yet since 2010.

[michsnowfreak]

22 hours ago, bluewave said:

The only thing consistent about Ann Arbor’s winter temperatures since the late 1800s has been a steady warming trend. December is up +4.3° with Jan at +3.2° and February +6.1°.

The snowfall has seen a nice increase over this period as you pointed out. This could be a function of the warming winters holding more moisture while still being sufficiently cold enough to increase the snow. It’s one of the benefits of living in a colder region. But this benefit isn’t held by other less warm regions which have seen a steady decrease in snowfall with their rising temperatures.

But relative to other parts of Michigan which can really cash in on lake effect snows, the SE corner was never a particularly wintry  part of the state compared to areas further north. But at least places like Ann Arbor have seen a decent improvement relative to the old days which didn’t see as much snow. 
 

 

Remember this is a coop station. The data is suspiciously cold. One thing that IS good about stations like this (coop stations, unmoved for 140+ years) is that you can still see the "good" and "not good" winters by comparing Ann Arbor to nearby Detroit.

I have noted on multiple occasions that locally December has warmed the most, January not at all, and February slightly over the last 100 years. And Ann Arbor is yet another station that shows this. An increase of 3.4F in Dec, 0.0F in Jan, 1.5F in Feb.

Avg January temperature regression the last 100 years (1926-2025)
Toledo: -0.8F
Detroit: -0.4F
Flint: -0.1F
Ann Arbor: 0.0F
Saginaw: +0.3F
 

 

[40/70 Benchmark]

[40/70 Benchmark]

2 hours ago, bluewave said:

We have only had one October really amplified MJO 5-6 during each multiyear La Niña going back to 2010. October 2024…+2.76…October 2020…+2.81….October 2017…+3.35….October 2010….+2.88.

The other La Niña years surrounding these in each group had a weaker October MJO in 5-6 like in 2022, 2021, 2016, and 2011. So the more amplified October years featured La Niña +PNA mismatches.

With the exception of last year, these were very snowy winters. But the WPAC pattern and Pacific Jet never relaxed last winter like all the previous mismatch winters did. This is what I was pointing out last year why I mentioned early on that there were competing influences which didn’t exist during the other winters. So not to expect the same type of outcome.  

So this would be a first time occurrence if the MJO 5 peaked again this October in the +2.76 to +3.35 range. Seems like it’s some type of fall forcing event which affects the winter PNA during La Ninas. 

It’s why my guess a few months back that this 25-26 winter will be warmer than last winter was with the PNA averaging less positive than last winter did. But since the snowfall was so low last winter, it wouldn’t take much for one decent snow event to surpass last seasons snowfall totals from around Philly to Boston.

Plus it’s possible that we could get a least one winter month with a decent +PNA like we saw in January 2022 even though the PNA was strongly negative in December tilting the whole winter -PNA. That was largely driven by the MJO 8. January 2022 was the last winter month around NYC which was both cold and snowy especially Long Island.

We have also seen very impressive 500 ridges in Canada since the 2023-2024 El Niño generally boosting the +PNA. So we’ll have to wait for the October verification this year to know if it will be like  past multiyear La Ninas following the mismatch winters like we had in 2024-2025. Always have to leave open the possibility of a first time occurrence with back to back mismatch events. But this hasn’t happened yet since 2010.

I could have sworn 2021-2022 was a mismatch season...interesting that we still managed a mismatch period minus the amplified October MJO....I could certainly see something like that occurring this season.

I agree this season will probably be warmer with a more consistently negative PNA, yet still manage more snowfall.

[TheClimateChanger]

1 hour ago, michsnowfreak said:

Remember this is a coop station. The data is suspiciously cold. One thing that IS good about stations like this (coop stations, unmoved for 140+ years) is that you can still see the "good" and "not good" winters by comparing Ann Arbor to nearby Detroit.

I have noted on multiple occasions that locally December has warmed the most, January not at all, and February slightly over the last 100 years. And Ann Arbor is yet another station that shows this. An increase of 3.4F in Dec, 0.0F in Jan, 1.5F in Feb.

Avg January temperature regression the last 100 years (1926-2025)
Toledo: -0.8F
Detroit: -0.4F
Flint: -0.1F
Ann Arbor: 0.0F
Saginaw: +0.3F
 

 

You are something else. I love how you always do what you accuse me and @bluewaveof doing. Let me try to understand this: Co-op data good when it shows less warming, co-op data bad when it shows more warming? Am I doing this right?

And you always claim I ignore data I don't like but are CONSTANTLY doing the same: "The data is suspiciously cold." My man, this is data collected by the University of Michigan. At least the older data collection was probably handled by the earth sciences department. Are you saying the University of Michigan doesn't/didn't know how to properly collect temperatures and precipitation? It's funny because if I doubt an old record high temperature that's way out of line with surrounding observations, you accuse me of doing this. Now, you are out here complaining about old temperature records from a top public research university. In the hierarchy of things, I always give the most weight to records collected by the Weather Bureau/NWS at first-order sites, followed by universities, experimental farms and state/national forests and parks, since these are the sites that were manned by meteorologists, agronomists, or park rangers [i.e., people with an earth science background].

You also like to go on about cherrypicking start dates (1960s & 70s). I like to use these as a baseline since the data is more consistent [i.e., fewer station moves and less biases] and more relevant to current trends. But you say its cherrypicking. But now, you throw out the observations from the 1800s because you don't like the trend, and insist on showing the trend from 1920s? How is that not considered cherrypicking? Warming since 1880, significant warming since 1960/1970 - oh, but we must only consider the trend from exactly the 1920s, where there is little warming? 

[TheClimateChanger]

1 hour ago, michsnowfreak said:

Remember this is a coop station. The data is suspiciously cold. One thing that IS good about stations like this (coop stations, unmoved for 140+ years) is that you can still see the "good" and "not good" winters by comparing Ann Arbor to nearby Detroit.

I have noted on multiple occasions that locally December has warmed the most, January not at all, and February slightly over the last 100 years. And Ann Arbor is yet another station that shows this. An increase of 3.4F in Dec, 0.0F in Jan, 1.5F in Feb.

Avg January temperature regression the last 100 years (1926-2025)
Toledo: -0.8F
Detroit: -0.4F
Flint: -0.1F
Ann Arbor: 0.0F
Saginaw: +0.3F
 

 

For a more evidence-based analysis, NCEI shows January has warmed at a rate of about 2F/century on average in the whole of Southeast Michigan. Why the divergence between your numbers? None of them, except the Ann Arbor data is from a single site. According to your analysis (I didn't independently verify), the Ann Arbor data is exactly flat over the last 100 years. But reviewing the data, the time of observation moves from 5 pm to early morning. There is also a bias from the switch to MMTS, both of which are corrected in the NCEI data. The other sites are all amalgamations of distinct stations. Toledo and Detroit go from rooftop stations in downtown [and, in the case of Toledo, very near Lake Erie, which has a warming influence in the winter] to suburban (DTW) or even downright rural (TOL) airport locations, with proper siting on the ground.

[mitchnick]

6 hours ago, snowman19 said:

As Ray has mentioned a few times, descending solar from a solar max peak is actually more hostile to NAO/AO blocking in winter than it is right at solar max peak

Who referenced Nao or winter? I just pointed out the current solar spike was a short term spike. No more, no less. The irresistible impulse of you and Bluewave to turn everything negative for posters who enjoy cold or snow is strange to say the least. Actually, it's likely much worse than strange. 

[Stormchaserchuck1]

July will likely be the 5th straight month with +AO. 

How does that roll-forward, through March? Here's an animation of the next four 5-month periods.. as you can see there is a -0.3 H5 correlation over the Arctic circle, or a 57-58% chance of the +AO continuing (over the Arctic circle, north of Alaska at least). This is actually a little more +epo than +ao in technical classifications, in the roll-forward. 

[michsnowfreak]

1 hour ago, TheClimateChanger said:

You are something else. I love how you always do what you accuse me and @bluewaveof doing. Let me try to understand this: Co-op data good when it shows less warming, co-op data bad when it shows more warming? Am I doing this right?

And you always claim I ignore data I don't like but are CONSTANTLY doing the same: "The data is suspiciously cold." My man, this is data collected by the University of Michigan. At least the older data collection was probably handled by the earth sciences department. Are you saying the University of Michigan doesn't/didn't know how to properly collect temperatures and precipitation? It's funny because if I doubt an old record high temperature that's way out of line with surrounding observations, you accuse me of doing this. Now, you are out here complaining about old temperature records from a top public research university. In the hierarchy of things, I always give the most weight to records collected by the Weather Bureau/NWS at first-order sites, followed by universities, experimental farms and state/national forests and parks, since these are the sites that were manned by meteorologists, agronomists, or park rangers [i.e., people with an earth science background].

You also like to go on about cherrypicking start dates (1960s & 70s). I like to use these as a baseline since the data is more consistent [i.e., fewer station moves and less biases] and more relevant to current trends. But you say its cherrypicking. But now, you throw out the observations from the 1800s because you don't like the trend, and insist on showing the trend from 1920s? How is that not considered cherrypicking? Warming since 1880, significant warming since 1960/1970 - oh, but we must only consider the trend from exactly the 1920s, where there is little warming? 

No YOU are something else. I picked a round figure- 100 years. Using the same graphs you always use. This is so rich coming from someone who goes into everyones subforum (where most ignore you) with the most random data for the most random starting points. You just pick based on whatever gives you what you want. One minute youll use POR the next you will decry it. I never said I threw out data, I brought up how the 1880s data seemed low at this coop station. Meanwhile, you ALWAYS have a problem with older data you dont like and are always discounting it. See, ive studied my areas climate data for years. You just plug in numbers for anything anywhere to find what you want. Since you got so offended that I questioned anything about the Ann Arbor data, you must LOVE how their winter temps have remained pretty steady for decades and snowfall has MORE THAN DOUBLED since records began.

[Stormchaserchuck1]

Here is US pattern 5 months after 5-straight months of +AO (Aug-Dec)

[donsutherland1]

Although winters are warming in the Great Lakes Region, the initial point made earlier in this thread concerning the coldest temperatures is not unreasonable. I used the Ann Arbor (University of Michigan COOP) to illustrate the point given some of the commentary regarding the Detroit threaded record.

Instead of taking the lowest temperature for each winter, I focused on the lowest 1% of temperatures for a somewhat broader perspective. I also included 30-winter moving averages for the lowest 1% of temperatures. These temperatures have been range-bound (ranging from -10.1° for winters winters 1882-83 through 1911-12 to -3.1° for winters 1930-31 through 1959-60). The most recent value is -5.3°  for the 30-winter period ending with winter 2024-25. The historic value for the entire period of record is -6.8°.

Below are charts for individual winter and 30-winter values (trend line is +0.3° per decade) and also a closer look at 30-winter values on a narrower temperature scale to show the variability in the 30-winter mean value. It should be noticed that the sharpest rise in the 30-year moving average occurred prior to 1960.

The slower rise in extreme cold in the Great Lakes Region likely has to do with how marine heatwaves in the Atlantic and Pacific have increasingly focused the coldest air masses into the Plains/Great Lakes area rather than over a larger stretch of North America through their influence on synoptic patterns. 

If one looks at the warmest 1% of winter minimum temperatures, one has seen a steady rise since the 30-winter period ending in 1959-1960, unlike the roller-coaster with bottom 1% of low temperatures.

What this means is that winter temperatures have grown more volatile allowing for a relatively stable bottom 1% of temperatures even as the top 1% of warmest low temperatures has continued to increase. Below are the 30-year mean difference between the highest and lowest 1% of minimum temperatures.

It is uncertain whether this increased volatility is merely part of a transition in a warming climate (warmer highest 1% of lows but more stable coldest 1% of lows for a short period of time before the coldest 1% lows shift upward), or whether the increased volatility, which allows for continued shots of extreme cold, will become a "new normal" for perhaps several decades. For now, elevated volatility rules.

[Stormchaserchuck1]

15/18 days in July have had +SOI now. 

This is after 9-straight +SOI months, going back to Oct 2024. July will likely be the 10th consecutive month. 

Here's what the next 10 months looks like in the roll-forward - 0.2 correlations is actually pretty decent for such a wide area of time. 

[michsnowfreak]

1 hour ago, donsutherland1 said:

Although winters are warming in the Great Lakes Region, the initial point made earlier in this thread concerning the coldest temperatures is not unreasonable. I used the Ann Arbor (University of Michigan COOP) to illustrate the point given some of the commentary regarding the Detroit threaded record.

Instead of taking the lowest temperature for each winter, I focused on the lowest 1% of temperatures for a somewhat broader perspective. I also included 30-winter moving averages for the lowest 1% of temperatures. These temperatures have been range-bound (ranging from -10.1° for winters winters 1882-83 through 1911-12 to -3.1° for winters 1930-31 through 1959-60). The most recent value is -5.3°  for the 30-winter period ending with winter 2024-25. The historic value for the entire period of record is -6.8°.

Below are charts for individual winter and 30-winter values (trend line is +0.3° per decade) and also a closer look at 30-winter values on a narrower temperature scale to show the variability in the 30-winter mean value. It should be noticed that the sharpest rise in the 30-year moving average occurred prior to 1960.

The slower rise in extreme cold in the Great Lakes Region likely has to do with how marine heatwaves in the Atlantic and Pacific have increasingly focused the coldest air masses into the Plains/Great Lakes area rather than over a larger stretch of North America through their influence on synoptic patterns. 

If one looks at the warmest 1% of winter minimum temperatures, one has seen a steady rise since the 30-winter period ending in 1959-1960, unlike the roller-coaster with bottom 1% of low temperatures.

What this means is that winter temperatures have grown more volatile allowing for a relatively stable bottom 1% of temperatures even as the top 1% of warmest low temperatures has continued to increase. Below are the 30-year mean difference between the highest and lowest 1% of minimum temperatures.

It is uncertain whether this increased volatility is merely part of a transition in a warming climate (warmer highest 1% of lows but more stable coldest 1% of lows for a short period of time before the coldest 1% lows shift upward), or whether the increased volatility, which allows for continued shots of extreme cold, will become a "new normal" for perhaps several decades. For now, elevated volatility rules.

Excellent post Don. Every discussion on climate does not always need to come back to "well the mean temperature rose...". More than one aspect of weather can be discussed. I never said great lakes winters havent warmed since the 1870s. But the warmth is not nearly as extreme other locations. As I've said I've studied Detroits weather climate for decades. Im talking newspaper stories, details in the LCD forms, anecdotes in the old weather books, etc not just the raw numbers on xmacis. These graphs illustrate well the weaker winters of the 1930s-60s locally. No matter what way you slice it or which dataset you use, those decades strongly lacked extreme cold and extreme snow. For instance, let's say i was born in 1930. By my 40th birthday, I would have seen far less severe winter weather than I had seen by being born in 1983 like I actually was. I dont know what the reasoning was or why some are so against discussing it. But it was a multi-decadal regime that really differed from the decades before and since. The radical change to harsh 1970s winters was like a different climate.

Based on what journal data is available, I suspect had official weather records started earlier, say 1830, we wouldve seen a trend of even less snow and more extreme cold low temps. A snow lover is only going to want it so cold in this climate zone.

As for extreme cold? Coldest I've felt in my life was Jan 2014 and Feb 2015. Those winters are infamous for the severity. But in the ensuing milder winters since 2016, there have been multiple shots of severe cold as well. And that was where this discussion started. Talk about warming climate all you want, but dont act like severe cold shots dont still happen.

Lastly, as for the threaded record moving around. That happens in most places. There are obvious differences in different places, but also, you can't compare a place in 2025 and assume it was the exact same in 1935. When the weather bureau/weather service moves a station, all of the bells and whistles go with it. So essentially, since Detroit city airport stopped being the official site in the 1960s its basically a coop station with an unmamned max/min thermometer, an inaccurate tipping gauge and no snow measurements.

 

[Stormchaserchuck1]

^Decadal trends are real. Some people don't get that for some reason. The Earth seems to move in waves, that are decades long. Now with all this data in 2025 we can map out the NAO/PDO and see that the standard deviation of random vs cycles is about 5:1 (5x more likely to be cyclical). The Earth is warming an average of 3-4F. ok. I don't think that applies to everything, and large scale patterns recently. The problem with data-analysis right now is that we might be in the middle of the -PDO/+NAO cycle, seeming like it's a one-way trend (with regards to the most talked about subject - Winter Northeast snowfall). 

[PhiEaglesfan712]

 

[GaWx]

From WB’s Joe D’Aleo today:

That warm water east of Japan will move east in the North Pacific current north of the North Pacific Gyre. How much warmth will be carried east before the winter is something we will need to monitor. If it ends up staying more in mid-ocean, a very different winter story would evolve. This is just a heads up.

When warm water in the Pacific has settled in the northeast Pacific, winters are brutal. Examples include 1976/77, 1977/78, 1993/94 and 1916/17, 1917/18 and 2013/14, 2014/15/

Those back to back cold winters turned the trend negative for 1996 to 2015. The winter snows also set records.

[Stormchaserchuck1]

^lol, he's basically talking about a +PDO. 

[GaWx]

25 minutes ago, Stormchaserchuck1 said:

^lol, he's basically talking about a +PDO. 

But 1916-7 was one of the most -PDO DJFs on record and 1917-8 averaged neutral.

[Stormchaserchuck1]

6 minutes ago, GaWx said:

But 1916-7 was one of the most -PDO DJFs on record and 1917-8 averaged neutral.

I guess 1916-17 is his only one that makes sense. It was a -PNA Winter though, way below average in the West and Upper Midwest, average in the SE and Mid-Atlantic (vs 1900-2000 averages). 

[michsnowfreak]

44 minutes ago, GaWx said:

From WB’s Joe D’Aleo today:

That warm water east of Japan will move east in the North Pacific current north of the North Pacific Gyre. How much warmth will be carried east before the winter is something we will need to monitor. If it ends up staying more in mid-ocean, a very different winter story would evolve. This is just a heads up.

When warm water in the Pacific has settled in the northeast Pacific, winters are brutal. Examples include 1976/77, 1977/78, 1993/94 and 1916/17, 1917/18 and 2013/14, 2014/15/

Those back to back cold winters turned the trend negative for 1996 to 2015. The winter snows also set records.

Sweet Jesus what a dataset lol. He must be trying to get a jump on JBs subscribers.

[michsnowfreak]

2 minutes ago, Stormchaserchuck1 said:

I guess 1916-17 is his only one that makes sense. It was a -PNA Winter though, way below average in the West and Upper Midwest, average in the SE and Mid-Atlantic (vs 1900-2000 averages). 

1916-17 is one of the few winters that I've never looked too deep into. Just glancing it looked cold and white, with snowfall a bit below avg but good enough for lasting deep winter.

[PhiEaglesfan712]

52 minutes ago, michsnowfreak said:

1916-17 is one of the few winters that I've never looked too deep into. Just glancing it looked cold and white, with snowfall a bit below avg but good enough for lasting deep winter.

1916-17 was one of the strongest la ninas of all time, on the level of years like 1973-74, 1988-89, and 2010-11. 1917 was when global temperatures were at their minimum (2024, the warmest year on record, was nearly 2C warmer than 1917). I'd be willing to bet that either 1916-17 or 1917-18 was the coldest winter on record CONUS.

[GaWx]

42 minutes ago, PhiEaglesfan712 said:

1916-17 was one of the strongest la ninas of all time, on the level of years like 1973-74, 1988-89, and 2010-11. 1917 was when global temperatures were at their minimum (2024, the warmest year on record, was nearly 2C warmer than 1917). I'd be willing to bet that either 1916-17 or 1917-18 was the coldest winter on record CONUS.

Back to 1894-5, I have DJ of 1917-8 and 1976-7 as the coldest DJ in the E 1/2 of the US:

[bluewave]

15 hours ago, michsnowfreak said:

Remember this is a coop station. The data is suspiciously cold. One thing that IS good about stations like this (coop stations, unmoved for 140+ years) is that you can still see the "good" and "not good" winters by comparing Ann Arbor to nearby Detroit.

I have noted on multiple occasions that locally December has warmed the most, January not at all, and February slightly over the last 100 years. And Ann Arbor is yet another station that shows this. An increase of 3.4F in Dec, 0.0F in Jan, 1.5F in Feb.

Avg January temperature regression the last 100 years (1926-2025)
Toledo: -0.8F
Detroit: -0.4F
Flint: -0.1F
Ann Arbor: 0.0F
Saginaw: +0.3F
 

 

It’s good that they have a continuous record at Ann Arbor since it shows a similar long term trend is Wayne County. The NCDC takes into account station moves like Detroit compared to the rest of the county. 

 As you said, January has seen the slowest warming since 1895. But December and February have warmed at a faster pace. This makes sense since the falls have also been warming. So it takes longer for winter to get started and the winters across the U.S. are getting shorter with more warming in February. 

December +4.3°/Century

January +2.0°F/Century

February +5.1°F/Century 

[chubbs]

7 hours ago, michsnowfreak said:

Excellent post Don. Every discussion on climate does not always need to come back to "well the mean temperature rose...". More than one aspect of weather can be discussed. I never said great lakes winters havent warmed since the 1870s. But the warmth is not nearly as extreme other locations. As I've said I've studied Detroits weather climate for decades. Im talking newspaper stories, details in the LCD forms, anecdotes in the old weather books, etc not just the raw numbers on xmacis. These graphs illustrate well the weaker winters of the 1930s-60s locally. No matter what way you slice it or which dataset you use, those decades strongly lacked extreme cold and extreme snow. For instance, let's say i was born in 1930. By my 40th birthday, I would have seen far less severe winter weather than I had seen by being born in 1983 like I actually was. I dont know what the reasoning was or why some are so against discussing it. But it was a multi-decadal regime that really differed from the decades before and since. The radical change to harsh 1970s winters was like a different climate.

Based on what journal data is available, I suspect had official weather records started earlier, say 1830, we wouldve seen a trend of even less snow and more extreme cold low temps. A snow lover is only going to want it so cold in this climate zone.

As for extreme cold? Coldest I've felt in my life was Jan 2014 and Feb 2015. Those winters are infamous for the severity. But in the ensuing milder winters since 2016, there have been multiple shots of severe cold as well. And that was where this discussion started. Talk about warming climate all you want, but dont act like severe cold shots dont still happen.

Lastly, as for the threaded record moving around. That happens in most places. There are obvious differences in different places, but also, you can't compare a place in 2025 and assume it was the exact same in 1935. When the weather bureau/weather service moves a station, all of the bells and whistles go with it. So essentially, since Detroit city airport stopped being the official site in the 1960s its basically a coop station with an unmamned max/min thermometer, an inaccurate tipping gauge and no snow measurements.

 

We each have our different perspectives. If you look at Don's charts and compare  today vs 100 years ago, there has been 4-5F warming in both the warmest and coldest days; but, the 100-year path is different. The warmest 1% have had fairly steady warming but the coldest days have been a roller coaster. Despite the roller coaster, the coldest days are still warmer vs 100 years ago. What Don's analysis indicates to me is that we have to be careful in picking our start and end dates when looking at winter cold extremes. 

Metro and City airport have a long overlap period. During that period Metro is roughly 2F colder than the City airport so any analysis that starts with the city airport and ends with Metro will be contain both a station shift and a weather trend. If you don't account for the station change you won't get the correct weather trend.

I would need to see more data to be convinced that the midwest winters are warming at a slower rate. I haven't seen that in any study or apples-to-apples data comparison. The only geographic trend I am aware of is somewhat faster warming further north. Canada is warming faster than the US for instance.

Found one study when googling midwest winters, which looked at midwest winter storm tracks between 1959 and 2021.  The findings aren't surprising. The midwest winter storm track has shifted north and the midwest is getting wetter and warmer storms. Could help explain why some midwest stations are getting more snow.

https://agupubs.onlinelibrary.wiley.com/doi/epdf/10.1029/2024GL109890