TheClimateChanger

Bit of an understatement. It's the second hottest spring on record to date in the Washington, DC threaded record, and hottest at IAD. A little cooler at Baltimore, I guess. Sixth highest in the threaded record, although the second hottest at the airport. The old Customs House records should have an asterisk attached to them. It was so hot there the NWS stopped taking readings at that site after 1998. DC Threaded IAD Baltimore Thread

Funnel cloud west of Youngstown.

Northern lights are visible to naked eye.

Severe Weather Statement National Weather Service Pittsburgh PA 527 PM EDT Thu May 9 2024 OHC013-WVC051-092145- /O.CON.KPBZ.TO.W.0011.000000T0000Z-240509T2145Z/ Belmont OH-Marshall WV- 527 PM EDT Thu May 9 2024 ...A TORNADO WARNING REMAINS IN EFFECT UNTIL 545 PM EDT FOR SOUTHEASTERN BELMONT AND MARSHALL COUNTIES... At 527 PM EDT, a severe thunderstorm capable of producing a tornado was located over Glen Dale, or over Moundsville, moving east at 25 mph. HAZARD...Tornado. SOURCE...Radar indicated rotation. IMPACT...Flying debris will be dangerous to those caught without shelter. Mobile homes will be damaged or destroyed. Damage to roofs, windows, and vehicles will occur. Tree damage is likely. This dangerous storm will be near... Moundsville, Bellaire, Shadyside, McMechen, and Glen Dale around 530 PM EDT. Other locations impacted by this tornadic thunderstorm include Rocklick, Lone Oak, Grand Vue Park, Dallas, Beelers Station, Sherrard, Glendale, and Benwood. PRECAUTIONARY/PREPAREDNESS ACTIONS... TAKE COVER NOW! Move to a basement or an interior room on the lowest floor of a sturdy building. Avoid windows. If you are outdoors, in a mobile home, or in a vehicle, move to the closest substantial shelter and protect yourself from flying debris. Please report severe weather by calling 412-262-1988, posting to the NWS Pittsburgh Facebook page, or using Twitter @NWSPITTSBURGH. && LAT...LON 3991 8083 4000 8081 4003 8063 4002 8052 3987 8052 TIME...MOT...LOC 2127Z 269DEG 21KT 3995 8078 TORNADO...RADAR INDICATED MAX HAIL SIZE...<.75 IN $$ Milcarek

I don't know if I'd agree that there's been a substantial change in the UHI in Philadelphia. The population was over 1 million by 1890, and over 2 million by 1950.

My question: if I installed a mercury thermometer housed in a Stevenson screen or cotton region shelter on a hot rooftop in center city Philly, how many 90+ days would it register each summer these days?

I just don't buy it. When I look at the annual weather review from 1950, Philadelphia was renowned for its moderate climate - seldom getting too cold or hot. They do acknowledge the humidity adds to the discomfort of the summertime temperature, but I'd be willing to bet humidity has gone way up since then. As for as temperatures were concerned, they report an average of 14 90+ days, but this was from a hot rooftop site in the center city. Today, Philadelphia is a blazing inferno in the summertime where you would die without A/C. Who even knows when the last time there was 14 or fewer 90+ days? Rooftop stations can get way hotter during the day than ground-level stations. What really gets the noggin joggin' is this is how all the old temperatures were taken. So how cold was it really in the past? Like in 1999, NOAA said "SHUT IT DOWN"... yet that's how the temperatures had always been taken at the Baltimore Common House. Wouldn't that bias have been present throughout the history of the site, and then we just graft on the much colder airport records. And, of course, people still kvetch about those BWI temperatures all the time. Source: How not to measure temperature, part 48. NOAA cites errors with Baltimore's Rooftop USHCN Station – Watts Up With That?

This is what I suggested might be the case. Each of the individual component stations with a positive trend, while averaging across the basket of stations [changing over time] manufactures a negative trend. You did not put the Coatesville 2W trend, but it certainly looks positive over the period for which data is available. MQS is probably positive, although it could be flat or slightly down due to starting near a warm set of summers. Chesco's BY data is slightly down, but, like I said yesterday, that's probably just an artifact of him changing equipment.

Hancock County tornado confirmed to be at least EF2.

Got to love that period of overlap. The site with more than twice as many 90+ days drops out after a handful of years of overlap, and replaced with his own weather station. Oh, lookey here, 90+ days are plummeting! At the same time everyone else is sweltering through three scorching summers (2010-2012). Honestly, I'm not even sure if the trend from his own backyard data is "real" because he has changed equipment and siting in that time. I bet there was a change around 2012/2013. For all the hullabaloo about siting, his "official" station [he has three] is right next to a stand of tall trees. Clearly not sited at a distance 4x the length of the obstructions, with considerable shading. Unironically, the rooftop station is probably the best sited in his case.

I'm just not seeing any evidence of this so-called blocking in the actual data, which shows this to be among the hottest springs ever recorded in the northeast and much hotter than anytime in the 1990s? Like wouldn't blocking make it cold in the spring? New York, NY Newark, NJ Hartford, CT Philadelphia, PA Burlington, VT Harrisburg, PA Pittsburgh, PA Buffalo, NY Washington - Dulles, VA

Looks like that was the first or second recorded tornado to touchdown in Hancock County. Not sure whether that 1990 tornado touched down in the county. The NWS PBZ site is ambiguous, with some data showing 0 and some showing 1 prior tornado. There's actually a lot of issues with this, with the map showing Ohio County with 0, and the table showing Brooke County with 0. The map is only updated through 2020, there has been at least 1 tornado in Ohio County since then. And Brooke County has had at least 1 tornado. The table appears to have juxtaposed the data from Brooke & Ohio, so it looks like every county in the CWA has now had at least one tornado. Obviously, the biggest reason the W. Va. panhandle counties have so few reported tornadoes is due to their small area. Although there does seem to be fewer in eastern Ohio there as well. Not sure if it's population or perhaps the rough terrain of the Ohio River Valley causing a relative minimum. Link: NWS Pittsburgh Severe Weather Climatology

How about the nation's "Air Conditioned City" - Bluefield, West Virginia. Nobody lives in Mercer County, West Virginia, so I think it's safe to say this isn't UHI. Looks like the 21st century hasn't been so kind... almost like it's warming or something. Coldest summertime lows Warmest summertime lows

Also, low temperatures in the 1930s and 1940s were much colder than today. You don't need air conditioning when it cools off into the 50s at night. You open the windows. Coldest mean summertime lows Morgantown, WV Wheeling, WV Warmest mean summertime lows Morgantown, WV [some of this 19th century data is suspect, but you get the picture] Wheeling, WV

Looks like mostly nonsense again. Misleading data presented in a way to produce the desired trend. Why basket things into decades when some of the decades don't have full data? Just show the annual trend to avoid overweighting some years. Also the changing basket of station has a huge effect. It looks like the 1890s had a high number of 90+ days, but it's probably based only on these two sites. We can see from the data you presented 90+ days at those sites were actually quite low compared to the average at these stations. The data from Phoenixville in the 30s and 40s is also suspect, inexplicably doubling West Chester many years. Of course, they love to trot out that 111F. Don't even get me started on these state records. Same thing with West Virginia. 112F in Martinsburg, when it was 98F at Hagerstown just up the road on the same date. I will say the deniers have a point to call some of this data into question, but it's mostly in the opposite direction than they claim. Funny the two longest standing coop sites both have a significant positive trend, even in this raw data which is subject to known biases. Yet somehow Chesco's "nature trick" converts that into a declining trend when he switches up the basket of stations.

Also, here is Harrisburg. Just a short drive from Chester County. Weird that Chester would have such a wildly differing trend than Philadelphia or Harrisburg.

Oh yes and as @chubbspoints out, you are still changing the mix of stations over time. It is possible that the individual stations can all have a positive trend [like that for Phoenixville and West Chester] while the average tally for the basket of stations shows a decline due to you varying what stations are included in the basket over time.

Well you know I'm going to have to comment. (1) Data from 1920s - 1960s is likely not directly comparable to more recent data at some of the sites. May co-op stations reset their thermometers at 5 or 6 pm, and thus would register extra 90s on hot days [when the following day may or may not have reached 90]. Would expect this to add at least a couple or few extra 90-degree days on average each year. (2) The trend is minimal from the low elevation sites, and biased due to weighting 2020s equally even though only 4 years are included. It looks like the 2020s have had an unusually low number of 90+ days, but it's based on only 4 years of data. Same thing with the high elevation sites, but you also then extrapolate the trend back to the 1890s even though the data stops in the 1950s. In fact, the trends from 1950s to the present are about the same and we can be pretty confident that general pattern would have persisted into the 1890s. So the big drop off implied in the regression is clearly not accurate. If the remainder of the 2020s have a high number of 90+ readings, the trends could be significantly different. (3) These trends do not agree with Philadelphia data [from a county neighboring Chester County]. These data were collected by trained meteorologists from the Weather Bureau and later National Weather Service, and thus are less likely to contain errors and are unaffected by any changes in observation time. I can anticipate you will claim the UHI effect is the culprit, but there's no explanation for why center city Philadelphia would have fewer 90-degree days than Chester County, which would have been very rural, in the 19th century and early 20th century.

Clearly, GPT is confused as well because it comes to the same conclusion as me.