TheClimateChanger

Decided to quantify the changes in early spring temperatures by comparing the average March high temperatures in the late 19th century to the most recent 13 years (2010-2022). I used average high temperatures to somewhat minimize the impacts of urban heat island effect. What we can see are absolutely prodigious changes. Large cities in the lower Great Lakes (Cleveland, Detroit, Chicago, Buffalo) experienced 19th century early spring temperatures which were generally cooler than Green Bay, Wisconsin in the most recent 13 years (41.6F). In the case of Buffalo, as well as cities like Minneapolis and Milwaukee, March high temperatures in the 19th century are about equal to (or in the case of MSP) cooler than recent average March high temperatures at places such as Duluth (37.1F) and International Falls (37.4F), and only a couple of degrees warmer than average March high temperatures at Caribou, Maine in the most recent 13 years (34.8F). March high temperatures in New York City during the late 19th century are cooler than recent readings in places such as Milwaukee, Minneapolis, Buffalo and Rochester, and only a couple degrees warmer than recent readings in Green Bay, Wisconsin. By comparison, recent March high temperatures in New York City are about a degree warmer than what was observed in Washington, D.C. in the 19th century. March highs in Pittsburgh, as well, are now 0.6F warmer than D.C. in the late 19th century. Washington, D.C. on the other hand is seeing March high temperatures today only a couple of degrees cooler than places such as Raleigh/Durham and Charlotte, North Carolina experienced in the 19th century. Philadelphia had cooler March highs in the 19th century than places such as Cleveland, Detroit and Chicago. Today, Philadelphia sees March high temperatures which exceed those measured in Washington, D.C. in the 19th century by several degrees. Generally, northern locations have warmed from 5-8F over that timeframe, while southern locations have warmed from 4-5.5F. These temperatures correspond to climate shifts of hundreds of miles. I think if you teleported someone in from the 19th century, they would be very confused by these changes - maybe even moreso than they would be of all the technological changes. Here is what I found: Milwaukee, Wisconsin 1871-1900: 37.1F 2010-2022: 44.1F (+7.0F) Detroit, Michigan 1874-1900: 39.7F 2010-2022: 47.4F (+7.7F) Chicago, Illinois 1873-1900: 40.9F 2010-2022: 47.6F (+6.7F) Cleveland, Ohio 1872-1900: 40.9F 2010-2022: 48.2F (+7.3F) Minneapolis, Minnesota 1873-1900: 36.7F 2010-2022: 43.3F (+6.6F) Buffalo, New York 1874-1900: 37.3F 2010-2022: 43.4F (+6.1F) Rochester, New York 1872-1900: 37.6F 2010-2022: 44.6F (+7.0F) Erie, Pennsylvania 1874-1900: 40.3F 2010-2022: 45.2F (+4.9F) Pittsburgh, Pennsylvania 1875-1900: 47.6F 2010-2022: 50.8F (+3.2F) Keep in mind at Pittsburgh, this is comparing downtown records (elevation: 780-800 feet) to airport records (elevation: 1200 feet), and about 15 miles northwest, which hides a lot of the warming trend. Washington, D.C. 1872-1900: 50.2F 2010-2022: 57.6F (+7.4F) New York, New York 1869-1900: 43.4F 2010-2022: 51.0F (+7.6F) Boston, Massachusetts 1872-1900: 42.0F 2010-2022: 46.9F (+4.9F) Philadelphia, Pennsylvania 1874-1900: 47.2F 2010-2022: 53.8F (+6.6F) Atlanta, Georgia 1879-1900: 61.1F 2010-2022: 66.5F (+5.4F) Memphis, Tennessee 1879-1900: 60.6F 2010-2022: 64.8F (+4.2F) Raleigh/Durham, North Carolina 1887-1900: 59.1F 2010-2022: 63.5F (+4.4F) Charlotte, North Carolina 1879-1900: 59.9F 2010-2022: 65.1F (+5.2F)

Another way to look at this is to compare locations today to places in the south in the past. Pittsburgh's average March high from 2010-2022 at 50.8F, is 0.6F warmer than Washington, D.C.'s average March high from 1872-1900 [50.2F]. By comparison, Washington, D.C.'s average March high in the most recent 13 years has been 57.6F.

Always interesting to do this exercise... looked at changes in March average high temperatures from 1871-1900, compared to the most recent complete 13 years (2010-2022) in the Great Lakes subforum. Decided to do the same to a few spots in our subforum. Buffalo, New York 1874-1900: 37.3F 2010-2022: 43.4F (+6.1F) Rochester, New York 1872-1900: 37.6F 2010-2022: 44.6F (+7.0F) Erie, Pennsylvania 1874-1900: 40.3F 2010-2022: 45.2F (+4.9F) Pittsburgh, Pennsylvania 1875-1900: 47.6F 2010-2022: 50.8F (+3.2F) Keep in mind at Pittsburgh, this is comparing downtown records (elevation: 780-800 feet) to airport records (elevation: 1200 feet), and about 15 miles northwest, which hides a lot of the warming trend. Based on the numbers, the March climate in Buffalo and Rochester during the late 19th century is about on par with recent (2010-2022) March weather in International Falls [37.4F mean] and Duluth, Minnesota [37.1F mean] and just about 2.5-3F warmer than Caribou, Maine [34.8F] - which is to say the 19th century March climate in Buffalo and Rochester is significantly closer to the modern March climate in Caribou, Maine than it is to the current climate at either location. I think this last point kind of serves to bring home the scope of the changes we are creating. Often times, this reality is lost in the focus on the numbers. But when you point to a concrete modern equivalent for those numbers, it really showcases how dramatically things have changed. Places like Duluth, International Falls and Caribou, Maine are thought of as unbelievably cold. Yet, many cities in in the lower Great Lakes saw comparable early spring temperatures to what those locations experience today.

Urban heat island effect can explain some of the local warming. But how can Urban Heat Island effect explain that Minneapolis had a colder average high in March from 1873-1900 [1900 population: 202,000] than International Falls [a town of 5800 on the Canadian border known as the "icebox of the nation] or Duluth, Minnesota [a city of 86,000 on the shores of Lake Superior]? Or that the average March high in Chicago [1900 population: 1.7M], Detroit [1900 population: 286,000], and Cleveland [1900 population: 382,000] in the same period is less than Green Bay's [pop: 107,000] high in the most recent 13 years?

The 19th century March high temperatures for these big cities, mostly in the southern Great Lakes, are comparable to modern averages at places like International Falls, Duluth, and Green Bay! How can the urban heat island effect cause that? This is an incredible shift, considering how cold we consider those locations today. Cleveland, Ohio 1872-1900: 40.9F Milwaukee, Wisconsin 1874-1900: 37.1F Detroit, Michigan 1874-1900: 39.7F Chicago, Illinois 1873-1900: 40.9F Minneapolis, Minnesota 1873-1900: 36.7F Modern Equivalents Green Bay, Wisconsin 2010-2022: 41.6F [warmer than Cleveland, Detroit and Chicago, and significantly warmer than Minneapolis and Milwaukee in the 19th century] Duluth, Minnesota 2010-2022: 37.1F [same as Milwaukee, warmer than Minneapolis, only a couple/few degrees cooler than lower Lakes] International Falls, Minnesota 2010-2022: 37.4F [warmer than 19th century Milwaukee and Minneapolis, a couple/few degrees cooler than lower Lakes]

Rockford only has sporadic 19th century data. I can offer Toledo, Ohio, as a somewhat smaller city. The airport is in a mostly rural part of Lucas County, southwest of the city. It does show the same general pattern with a bit less warming, with no change in January, nearly 6 degrees of warming in March, and between 3 and 4 degrees of warming in July & September. January 1874-1900: 33.4F 2010-2023: 33.4F (no change) March 1873-1900: 42.5F 2010-2022: 48.3F (+5.8F) July 1873-1900: 82.1F 2010-2022: 85.9F (+3.8F) September 1873-1900: 73.2F 2010-2022: 76.7F (+3.5F)

Well, this was a March thread. But here's what I got for those months. Generally March shows the most warming, which I assume is due to changes in snow and ice cover. July & September an intermediate level of warming, and January the least warming (although northern sites show more warming than southern sites). January Cleveland, Ohio 1871-1900: 33.5F 2010-2023: 35.4F (+1.9F) Milwaukee, Wisconsin 1874-1900: 27.1F 2010-2023: 30.7F (+3.6F) Detroit, Michigan 1874-1900: 31.1F 2010-2023: 32.9F (+1.8F) Chicago, Illinois 1873-1900: 31.1F 2010-2023: 31.6F (+0.5F) Minneapolis, Minnesota 1873-1900: 26.8F 2010-2023: 29.9F (+3.1F) July Cleveland, Ohio 1871-1900: 79.3F 2010-2023: 84.1F (+4.8F) Milwaukee, Wisconsin 1874-1900: 77.7F 2010-2022: 82.3F (+4.6F) Detroit, Michigan 1874-1900: 81.2F 2010-2022: 85.2F (+4.0F) Chicago, Illinois 1873-1900: 79.3F 2010-2022: 84.5F (+5.2F) Minneapolis, Minnesota 1873-1900: 82.8F 2010-2022: 84.9F (+2.1F) September Cleveland, Ohio 1871-1900: 72.2F 2010-2023: 76.1F (+3.9F) Milwaukee, Wisconsin 1874-1900: 69.8F 2010-2022: 73.5F (+3.7F) Detroit, Michigan 1874-1900: 72.0F 2010-2022: 75.5F (+3.5F) Chicago, Illinois 1873-1900: 71.5F 2010-2022: 76.0 (+4.5F) Minneapolis, Minnesota 1873-1900: 71.1F 2010-2022: 74.2F (+3.1F)

Mean March maxima for various cities in the Midwestern U.S. Using high temperatures here to minimize impacts of urban heat island effect. Any thoughts? Seems a lot more of an increase than I would have expected, considering the warming is said to be about 1C. Cleveland, Ohio 1872-1900: 40.9F 2010-2022: 48.2F (+7.3F) Milwaukee, Wisconsin 1874-1900: 37.1F 2010-2022: 44.1F (+7.0F) Detroit, Michigan 1874-1900: 39.7F 2010-2022: 47.4F (+7.7F) Chicago, Illinois 1873-1900: 40.9F 2010-2022: 47.6F (+6.7F) Minneapolis, Minnesota 1873-1900: 36.7F 2010-2022: 43.3F (+6.6F)

The Ohio Weather Review reported phenological data ever year for many years around 1900. In northern Ohio, at that time, the first leaf did not appear on soft maple trees until May 1 to May 10, with elm and birch occurring in mid to late May. Not sure about the latter two species, but I know the first leaf on soft maple species probably averages mid April now. Maybe late April to May 1 in a particularly cold spring. So on average, I'd say it's moved up 2-3 weeks in a little over 100 years. And in 2012, there were leaves in March in Toledo. A full six weeks earlier. I don't know how this can be disputed. Springtime temperatures have warmed nearly 5 degrees in Cleveland comparing the 1871-1900 mean [45.98F] to the 2010-2022 mean [50.62F], and more than 4 degrees at Detroit when comparing the 1874-1900 mean [45.50F] to the 2010-2022 mean [49.61F]. These are massive changes. The 1871-1900 (1874-1900 at DTW) means are comparable to the 2010-2022 means at Traverse City, MI [46.62F] and Green Bay, Wisconsin [44.80F], meaning a late 19th century spring in Cleveland and Detroit saw temperatures that are only slightly warmer than what is now a typical Green Bay spring, and about a degree cooler than modern springs in Traverse City, in northern Michigan. You may not like the data but that's what the data shows. You're welcome to perform your own analysis on the data. And this is all on topic - this is a banter thread, observations are about spring, and relevant to the Great Lakes/Ohio Valley weather discussion.

Very heavy rain here. Lots of thunder and lightning. Water everywhere.

Well, every year is an early greenup compared to the start of the 19th century. If you transported someone from the 1800s, they would think even the "late" greenups you cite were exceptionally early. This is not my opinion either, but backed up by real phenological evidence.

It will be interesting to see what kind of temperatures we achieve in the upcoming El Nino. Looking at the records, it's clear that the type of heat we see every single year now only occurred during significant El Nino events (1973, 1991 & 1998) when many of us were youngsters. Of course, 2015 & 2016 are both on the list as well. I suspect we'll see temperatures climb to new record-breaking heights with the next strong El Nino.

This just blew my mind. Shared this in the Ohio thread, but thought this was a statistic worthy of sharing to the entire subforum. Warmest start to the year at Akron-Canton Regional Airport, but more shockingly 8 of the 14 warmest years on record have occurred in the last nine years, dating back to 1887. Just absolutely astounding.

Very impressive stretch of heat in Ohio. At Akron-Canton Regional Airport, this has been the warmest start to the year on record (see below). Perhaps even more shocking, 8 of the 14 warmest years on record dating back to the 1800s have occurred in the last 9 years, and 9 since 2012! Outstanding, awe-inspiring statistic.

This is an objective index of multiple variables. Looking at Buffalo, I can certainly see support for a moderate designation. Temperatures were very mild, and length of period with snow on ground was fairly low [except at figures of greater than 6"] despite the heavy snowfall. All dating back to 1950, when the index begins: Second fewest highs less than 32 [16] Ninth warmest mean high Dec-Feb [37.9F] Second warmest mean minimum, Dec-Feb [26.7F] Fourth warmest mean average, Dec-Feb [32.3F] Coldest low temperature of 0F, T-18th warmest [of 74 years] Coldest high temperature of 14F, T-15th warmest [of 74 years] 23rd least [of 74 years] for days with 1"+ snow depth 27th least [of 74 years] for days with 3"+ snow depth 37th [of 74 years], or right in middle of pack, for days with 6"+ snow depth

Mild winter regionwide per the Accumulated Winter Season Severity Index (AWSSI), with several locations observing a record mild winter thus far. Seems to be a bit of disconnect in Michigan, where some folks are reporting more snowfall than usual, but this objective measure is showing this past winter to have been very mild.

https://www.washington.edu/news/2022/10/03/study-suggests-la-nina-winters-could-keep-on-coming/

Anyways, can we lock this in?

There were only 5 colder high temperatures since January 1, that I could find - and none by more than 5 degrees. 27 on the 1-14; 29 on 1-31; 28 on 2-1; 31 on 2-3; and 28 on 3-14. Plus, yesterday's 32F reading wasn't hit until nearly 6 pm, aided by the lengthy daylight. Early afternoon was still in the 20s.

Yeah, it is quite cold. My car was reading 25 just now and it’s rather brisk.

Not really feeling the snow squall potential out there. Looks more like light snow flurries given the minimal humidity to work with.

I did a quick "sanity check" of the annual snow numbers - I call it the good, the bad and the ugly. The good: CLE's data looks solid. Again ignoring any changes in site location and measuring techniques over time. No missing data with the exception, of course, being this year, which is incomplete. CAK's data looks solid, with the exception of 1947-1948. Snowfall records do not begin in the threaded record until 1/1/1948, so it's missing everything that fell on or before 12/31/1947. Looking at the recorded precipitation and temperatures, it's clear at least several inches fell in November and December 1947 if measured. TOL's data looks pretty good, with the exception of 1997-1998. Not sure what was going on there - it was a warm, snow-free winter. But there is a ton of missing data. I know there was a period at a lot of sites in the mid 90s where they did away with snowfall measurements for a few years, not sure if that's the problem here. As for the other years with missing data, when the missing data is a multiple of 30/31 in the early years of the record, it's usually just warm weather months where nothing was reported being interpreted as missing data. But as in the case noted above, there are circumstances where that might not be the case. The bad: Erie's data is mixed. Something weird happened in the 30s and 40s, where a lot of snowfall was observed but not transcribed and shows up as missing data. The years with missing data as a multiple of 30/31 are probably just warm weather months where nothing was reported and so it shows up as missing data, but obviously no snow fell. The years with unusual number of days missing are suffering from data quality issues. In 1943-44, 48.5 inches were actually observed, but 25 snowfall days are blanked out as missing resulting in the 12.1 inch count. 1942-1943, only monthly numbers were available on the NCDC/NCEI IPS site for the LCDs [and December was missing altogether]. But there were 31" in the months available [less December]. December shows 6.6" of snow on Climod 2, but there are a bunch of missing days - including a stretch where 0.76" of precipitation was recorded with temperatures mostly at or below freezing. Ascribing a 10:1 estimate on those days would suggest 45.1" season total, but this would need to be verified with the actual daily observations or monthly total for December 1942. 1931-1932 shows only one missing day, but that one missing day was March 31, 1932, when 3.8" of snow fell. So the seasonal count should be 30.3 inches. I did not check 1928-1929, but it shows 7 missing days and is in the same general era, so it's probably an undercount as well. The bad, bordering on ugly: YNG's data is pretty good from the late 40s on, but the 30s and 40s have tons of missing data. The first two years do not show up on your lists, as they were properly excluded due to the missing data, so I've shown 12 years here to encompass the 10 years that are on the list. As you can see, lots of missing data, which mostly is days where no snow fell. However, there are significant data quality issues in some of these years. I looked at 1941-1942 and 1943-1944, and only liquid equivalence of snowfall was reported. The algorithm, or whatever captured this data, interprets 0.04" as 0.4" thereby prescribing a 10:1 ratio to smaller snowfall amounts. However, bigger storms are substantially undercounted. A day with 0.52" showing a depth of 4" is interpreted as 0.5" of snow, so there is a huge undercount. A day with 0.29" of liquid, is reported in Climod 2 / XMacis as 0.3" of snow [when in reality is was probably 3" or more]. Like I said, I only looked at 2 years, so I don't know if this was going on in the other years. The downright ugly: MFD's data quality is really bad. 1919-1920 shows up as the least snowy with 3.3", but this was based only on December 1919 [and there were several days with snowfall that month that were not recorded by the observer, so it's not even a complete month of data] - no records were observed in November 1919, or at all in the entire year of 1920. So obviously 1920-1921 is also incomplete. 1903-1904 also reports 335 missing days. This was a weird case. The coop site observed only precipitation and snowfall at that time. It was actually a rather snowy winter with over 50" of snow observed, but none of that data is transcribed. Like at Toledo, 1997-1998 also shows lots of missing data at MFD as well. While 1918-1919 looks bad with 213 days missing, it actually appears to be a legitimate reading at least based on what was reported. Also lines up with the historically bad numbers reported elsewhere, but a lot of the other years look to be incomplete [1919-20, 1920-21, 1903-04, probably 1997-98]. I didn't look at every year on here, but given the missing data, likely other issues.

Pittsburgh just missed an even worse snowstorm later in April 1901 that dropped 20-40+ inches of snow in eastern Ohio, but dumped about 4" of rain in the city with the worst flood until 1936. Parts of West Virginia also saw more than two feet of snow with the first storm that did hit Pittsburgh. If the atmosphere was capable of producing 40" of snow in Ohio in late April of 1901, surely a foot and a half is possible in Pittsburgh in March even with climate change?

I'm not talking about the quality of data, but problems with the digitization of the data resulting in a lot of missing data that actually was observed and/or data improperly transcribed. Here's what NRRC Climod2 site shows for lowest seasonal snowfall in Erie, PA: Which matches with what NWS CLE shows (see: https://www.weather.gov/cle/Top10Annual): 1943-44 is shown to be the least snowiest in both data sets with 12.1 inches; however, there was actually 48.5 inches reported on the raw LCDs. There are 25 days with missing data on the digitized figures - all of them had observed snow that are missing from the total snowfall count. For example: The raw LCD data for Erie, PA in October 1943 shows 8.2 inches of snow fell, which would be the 4th most on record for the month of October. There appears to have been a mid-month storm system which gave way to lake effect/lake enhanced rain and then heavy, wet snow, judging by the massive precipitation totals. In the digitized data, from which the 12.1" seasonal figure is drawn, the snow from that storm is completely missing. The only missing data in October is the snowfall from the three days in which snow was observed - everything else matches the LCD exactly. This is only one example, it's like the entire winter in Erie. The bulk of the reported snow is not reflected in the digitized data. and reported as missing data. As noted, there is actually 48.5 inches of snow recorded in the official LCDs, which are accessible here: https://www.ncdc.noaa.gov/IPS/lcd/lcd.html

This is just a ridiculously bad snowfall season.