TheClimateChanger

It takes decades to get the full panoply of possible climatological outcomes in a stable climate. We have only surpassed 1.5C for a couple of years. Even if temperatures were to stabilize at this point for several decades, we would likely eventually see a summer heat wave exceeding anything of record as we started to "feel" the full impact of variability upon a new climatological state.

While less impressive from a historical standpoint, EWR saw 10 straight 90+ days in early August 2022. They also had streaks of 14 days (2010) and 12 days (2012), since 2002.

As recently as 2022, a 6-day heatwave at JFK matched its third longest string on record. A completely separate 8-day heatwave at LaGuardia later in the same summer also matched its 3 longest streak.

There's nearly 100 years with, more or less, complete data for May in the threaded record. As for that claim, possible, but unlikely. Fort Frances, Ontario is just across the border from International Falls and shows 0 days of 90+ that month. The second closest site (Littlefork Ranger Station) recorded 1 day.

It's observed from space. Less reflectivity would make it appear somewhat dimmer.

I would say a monthly record high maximum of 96F and monthly record high minimum of 70F are potentially deadly in a location with limited air conditioning. But this is a sparsely populated region, so one or two deaths over the baseline wouldn't even register as a blip.

A record-breaking 4th 90+ day at International Falls. Never before observed in the month of May. For some context, Cleveland has only seen one May with 4 or more days of 90+ (1962, 4); Detroit, 3 (1962, 6; 2018, 4; and 1895, 4); and Bristol, TN, 2 (1962 & 1941, 6). INL has twice as many days at or above 90+ than Dayton, Ohio has seen in the last 13 years! International Falls working hard to shed that the "icebox of the nation" title. I think we are really going to need to reconsider that title moving forward in a location that sees this kind of heat in the month of May.

International Falls has seen twice as many 90+ days in the month of May over the last 4 days than Dayton, Ohio has seen in the last 13 years! Just unreal. @Spartman

To put that into perspective, Cleveland has only seen one May with 4 90+ days (1962); Detroit, 3 times: 1962 (6), 2018 (4), and 1895 (4); and Bristol/Johnson City, Tennessee, twice (1962 & 1941, 6 each). And for those who would blame the latter on elevation, Tri-Cities Airport in Tennessee is actually closer in elevation to International Falls than the other two cities. International Falls looking to shed its title as "icebox of the nation" in a BIG way!

Incredible stuff. Record breaking 4th day at or above 90F yesterday at International Falls. That's the most ever recorded in the month of May, with 17 days remaining in the month.

He also never posts the record lows which would invariably show a much greater bias towards early year than record highs, especially using his convention of ignoring subsequent ties.

Those maps are always just a bunch of nonsense anyways. Martz only lists the first occurrence of a record high so it biases it toward earlier years. With that said, even using his data, 10 of 50 states have set monthly record highs since 2000. That's 20 percent of all states. Keep in mind, the data this is drawn from actually extends beyond 1895, with scattered data all the way back to 1870 [and limited data before then]. Limiting to 1895, that's 20 percent in 19.3 percent of the total years. Extending back to 1870, that becomes a fairly hefty overweighting of recent record highs. Again, especially considering that Martz only shows the first occurrence, so there are probably more recent ties not reflected on the map. Data quality on some of these records are also highly questionable. He includes even highly questioned observations. I saw one record shown in March for the State of Missouri that was 10+ degrees warmer than any other observation from the same month. When I went back to the original document, the Weather Bureau had actually drawn a question mark by a few of the observed highs. And even some of the recognized highs are questionable by today's standards. For instance, the 112F at Martinsburg, West Virginia in July 1936. It was 105F in Kearneysville, and 103F in Hagerstown on the same day. Is that level of discrepancy plausible for a daytime high temperature? @FPizzand @ChescoWxwould be having a conniption fit. People are constantly complaining about ASOS readings a couple of degrees too high on here.

This is such a big anomaly, I wonder if it's just some sort of data entry error.

Here's the difference for each site. The rankings have 1978 as 0.4F warmer in this District. Idk. Coatesville 1 SW: 0.7F cooler in 1978 Devault 1 W: 0.6F warmer in 1978 Drexel University: 2.2F cooler in 1978 Ephrata: 2.5F cooler in 1978 George School: 2.3F cooler in 1978 Graterford 1 E: 2.3F cooler in 1978 Holtwood: 1.8F cooler in 1978 Landisville 2 NW: 2.0F cooler in 1978 Lebanon 2 W: 2.2F cooler in 1978 Marcus Hook: 2.2F cooler in 1978 Morgantown: 1.0F cooler in 1978 Philadelphia (City): 3.8F cooler in 1978 Philadelphia (AP): 1.0F cooler in 1978 Phoenixville 1 E: 0.8F cooler in 1978 Reading 3 N: 1.6F cooler in 1978 West Chester 1 W: 1.1F cooler in 1978

I will say the adjustments are not perfect. There was one thing that had me scratching my head. I was confused that I thought I had made a mistake and was looking at the wrong data! 1976 is shown as the fifth coldest July on record for Pennsylvania's Division 3, with a reported mean of 71.2F on Climate at a Glance. This appears to be entirely an artifact of whatever adjustments they are making. The raw data reflects a cooler than normal July, but only by a little bit. The reported mean is 74.6F. By comparison, 1978 was much colder with a reported mean of 72.7F, yet Climate at a Glance has it at 71.6F or 0.4F warmer than 1976! Looking at the station composition, they add Blue Marsh Lake, Octoraro Lake, and get rid of Philadelphia 2304 Mkt [inner city] and West Grove 1 SE. So the 1978 set is more rural and not directly comparable, but what's odd is if you compare every site that it's common to both datasets 1978 is still significantly cooler across the board. Yet Climate at a Glance shows it as warmer. I highly doubt 1976 was actually the 5th coldest July on record for this climate division. I'm not really sure what's going on here, and whether it affects other districts. 1976 was the only year that had anywhere near that deviation from the mean. Wonder if it's just some sort of data error?

Anyways, I went through the raw published figures for the month of July for Pennsylvania's Division 3 [Southeastern Piedmont] back to 1957, and the raw numbers were warming at 3.3F/century. Climate at a Glance shows 5.8F/century. This isn't evidence of any error or nefarious intent, however; the "raw" trend that I calculated is just the change in the simple arithmetic mean. I did not make any adjustments/homogenization for changes in site location, elevation and urban character. The early years were "Philadelphia heavy" so they would be inflated relative to the recent means, with up to 4 of the stations being in the city. For comparison, PHL shows a warming trend of 7.8F/century over the same time frame. Given the changing composition of the stations, I suspect the reported trend is closer to the actual trend, although I can't rule out it being slightly overstated.

Is there any evidence these are actually stations used by NCEI? That seems like a lot for one county. Why would they add 16 stations in the past two decades?

What are the stations in each year?

I didn't look at Pennsylvania's Division 3 [Southeastern Piedmont] but I did notice the most recent Julys were showing lower on NCEI's Climate at a Glance than the published figures.

Well, at least for Pennsylvania's Division 9, I don't see that trend. Since 1957, 54 years show a cooling adjustment, 10 warming, and 2 years are exactly the same as published. To be honest, the warming of the 1980s appears to be an artifact of a pairwise homogenization adjustment - perhaps some sort of preferential treatment is given to first order stations. It looks like PIT had a warm bias, so the numbers were adjusted upwards a bit to fall more in line with PIT. The adjustment probably should be in the opposite direction, as the mean value is probably inflated from the apparent warm bias. Which is likely partially due to the HO-83, although the warm bias seems to predate its installation. Some of the recent Julys (2020 & 2021) have among the largest "chilling" adjustments.

I pulled all of the published numbers for Division 9 since 1957 for the month of July, which is when they began publishing in the current divisional format, and there is basically no difference in trend from the published numbers and the values reported on NCEI's Climate at a Glance. I don't see the big warming of recent data that you claim in these values, or in the Chester County May temperature figures you gave above. The earlier years are subject to some adjustment, most notably for Time of Observation bias, but I don't see a big nefarious warming of recent years? Climate at a Glance - Pennsylvania's Division 9 Original Published Values in Climatological Data - Pennsylvania's Division 9 Pittsburgh International Airport

Looks like they mostly resolved the UHI at MSP. It's been running middle of the pack at least for daytime temperatures compared to the region. Used to be 2-3F warmer than most sites.

It looks like @ChescoWx's claim is only true when you include data from personal weather stations and recently opened mesonet sites in the mix, but if you compare the actual divisional means from the reported observation sites, I'm having a hard time finding these big warming adjustments that are often claimed?

Compare this to 1988. The reported value is chilled only 0.3F, versus 1.3F for 2020. And we can see the impact of faulty thermometers (HO-83) with PIT showing a huge warm bias in that era, with temperatures matched only a couple of lock & dam sites [at the lowest elevations of the region].