TheClimateChanger

The high resolution, short term modeling did a good job with depicting that stationary band of thunderstorms over north central Ohio associated with the cold front and interaction with PTC Debby’s cyclonic flow. Radar estimates up to 5-6” in spots.

I feel like when these people make these ludicrous claims [i.e., they are manipulating recent temperatures upward], they forget the US has had an independent, parallel network of high quality stations for two decades now.

Anyways, anyone with half a brain knows recent years have been warm. Why would they be manipulating the data by adding to it? Moreover, NCEI and NOAA generate their anomalies from the nClimDiv dataset with way more stations, not USHCN. There is also a parallel network known as the Climate Reference Network (CRN) which has been operational since 2005. CRN sites are carefully monitored, mostly rural locations, with highly accurate equipment and quality control. The skeptics were the ones who lauded this effort. The recent trends from nClimDiv, USHCN and the raw and undadjusted U.S. CRN data are nearly identical. In fact, CRN has a slightly greater positive trend in the period 2005-2024. We can see this from May, himself, who created this graphic in 2020. Since that time, USCRN has continued to warm slightly faster than nClimDiv and USHCN. So obviously the homogenization and infilling are not creating spurious warming. If anything, those datasets have a slight cooling bias still. Comparing USCRN and nClimDiv to USCHN – Andy May Petrophysicist

I don't think he shares his sources... probably something Heller cooked up. Anyways, it piqued my curiosity enough to look into. I came across this "petrophysicist" who I would characterize as somewhat of a more cerebral skeptic than Heller. His conclusions are clearly wrong, but at least he provides context to determine what is being shown in this graphic. @chubbscan correct me if I'm misinterpreting this. Recent USHCN Final v Raw Temperature differences – Andy May Petrophysicist This image appears similar to the one presented above: According to Mays, this shows the difference between the mean of the raw data and the mean of the final data by year. The orange trace shows the number of stations over time with raw data. Many stations with missing data are estimated using pair-wise homogenization. If you look only at stations with raw data, the graphic is much different. There is substantially less difference in recent years from ~0 to ~0.2C [ending in 2020]. This purports to show the difference between the final average of all data [including estimated data] and the average of the station data for which actual raw data is available. I think you can see the problem with this - if the stations where the estimated data is from locations where the mean temperature is generally warmer than the US average, then it's going to look like there's some big discrepancy. And that appears to be what's going on. May appears to anticipate this argument, implicitly, by claiming the stations dropped off are widely scattered. Although he does acknowledge Oklahoma appears to have a disproportionate number. Looking at this, yes, Oklahoma and the Delmarva region appear to have the most station dropoff [assuming this is accurate]. Oklahoma has a yearly mean of nearly 60F. The Delmarva is like 54-55F. The national average is 52F. In order to accurately gauge the impact of infilling, you would need to compare it to the trend generated only from the stations with raw data for the entire period. Mays never does this step. Instead, he just concludes without evidence that is causing a massive warming trend.

Thank you for the update. I saw mention of the flash flood warning for that on my X (formerly Twitter) feed this morning, but wasn't sure if it was unusual or something that happens frequently there. Has this ever occurred before?

Will be interesting to see whether the NAM 12k or 3k performs better? Or if they are both out to the proverbial sea. The 3K brings over 5" of rain to parts of eastern Ohio, while the 12K has minimal rain there. 3K 12K

The 5-minute obs could be rounded up or down depending on the temperature reported. A 5-minute ob of 91 means the rounded Celsius value was 33C, which means the actual value could be between 32.5 and 33.4C. An actual max temperature between 32.5-33C would all result in a rounded high of 91F, while a temperature of 33.1C to 33.4C would result in a high of 92F. An observation of 90F (i.e. 32C) is the opposite - it can be either 89F or 90F, depending on the exact reading in Celsius.

Up to 3.5” of rain on the day IMBY. Currently raining at a rate of about 2” per hour.

79 mph wind gust at Burke Lakefront Airport in Cleveland.

Pushing 3” at my home gauge. Wow.

Made it up to 91F just ahead of the line of approaching thunderstorms.

I think I may be the only one still keeping track, but yesterday was indeed the 18th 90+ reading of the year. First to top off at exactly 90F. Has reached 89F so far today, which typically would probably guarantee a 90F or better high temperature, but clouds are quickly on the increase with showers and thunderstorms to the north. Clearly, I was wrong when I indicated last week might be the last chance for awhile. But, at this point, the only day with any chance of 90F for the foreseeable future [after today] is Thursday. At this point, mid to upper 80s looks like the better bet. Debby's remnants move north into Pennsylvania on Friday, cooling things off, and helping to pull down a cooler and drier airmass in its wake.

Interesting. Hopefully, the NAM is off its rocker. Could definitely do without over 4" of rain in 12 hours from the remnants of Debby.

Clearfield Lawrence AP (1516' elev.) Dubois Jefferson AP (elev. 1804') That's 1.6F difference over less than 300 feet. So my estimate is based on actual data. I guess I need to go to flight school to learn how the actual data is wrong.

So yeah, I said what I said and I believe it to be accurate.

Pick any two sites that are nearby are you'll see the actual difference is greater than 4F per 1000 feet. Here's 3.4F difference on 800' elevation gain. You say you need 1,000 feet of gain for this change. Actually, you claimed even less in a humid climate like Pennsylvania. Cambria County Airport (2274') Altoona Airport (1467')

Near surface lapse rate is greater than open air lapse rate, due to downsloping effects. Only one is relevant to this discussion.

Going to be a close one today, either 89 or 90F so far. With the clouds rolling in, might be done for the day. I did see the current temperature was 31.7C (89.1F), which is also what the 6-hourly high was at 1:51 pm. Needs to reach [or have reached] 32.0C (89.6F) to go in the books as 90F.

NCEI reports August 2022 was 3.3F warmer than August 1975 in the Southeastern Piedmont region of Pennsylvania. Is this reasonable? One way of looking at this would be to compare stations for which data is available in both periods - unfortunately, I see only two such stations - Philadelphia International Airport and Phoenixville. Looking at these two, PHL was 4.7F warmer in August 2022 relative to August 1975. Phoenixville was 4.4F warmer, albeit with some missing data in 2022. Additionally, the Franklin Institute site (2022) is very near to the 23rd & Market site (1975) in Philadelphia. It was 4.3F warmer in 2022 versus 1975 at those two sites. Overall, if anything, the NCEI figures look cooler than the comparative data might suggest. This looks like more hullabaloo over nothing.

This is actually surprisingly easy to debunk. Chester County is part of what is known as the Southeastern Piedmont climatological region of Pennsylvania. In August 1975, the mean of 19 stations was 74.6F. In August 2022, the mean of 21 stations was 77.2F. If you pull up the current divisional data from NCEI, you'll see August 1975 is 73.0F. Looks pretty damning - that's a drop of 1.6F over what was reported. But then turn to August 2022, the value is 76.3F, or 0.9F cooler than reported. It's important to note that the values reported in the publication are simply the arithmetic mean of the station data, while the means shown on NCEI are areal averages factoring in topography/elevation etc. As long as processed in the same manner, the suite of stations can change and still produce a comparable value. In any event, there is no systematic warming of recent data like these guys frequently show - the opposite, in fact. You're welcome to do this with any time period, any month, any region since they started using the current climatological divisions and the result. That's why it's important, as @ChescoWx always says, to look at the raw data!

Not too surprised the "ROU" is cooler than Harrisburg International Airport. One is at ~300' on the Susquehanna River, and the other is 750-800 feet, with some residential areas on the neighboring hills at 1,000 to 1,200 feet. Not sure how much of that is attributable to an urban heat island effect. That level of elevation gain is sufficient for daytime maxima to be around 3F cooler in town, and upwards of 5F cooler in the hills, relative to the airport.

Some of this data is wrong for the record-breaking winter of 1935-1936. While many of the lowest hourly dewpoints on record at Des Moines did occur during that winter (including the all-time low of -40F on January 29, 1936 & February 14, 1936), it was not lower than -60F with those huge daily swings shown.

Also has daily minimum and maximum dewpoints, which can be downloaded as a .csv file.