chubbs

Agree, probably shelter-related. Easy with a non-aspirated thermometer. The whole region gets the same day-to-day weather so there's a wealth of information in the coop data that's revealed by inter-comparing sites. Been telling Chesco that for years.

I've looked at a couple of the individual months, July 1934 shown below. In addition to an overall warm afternoon bias. Phoenixville also must have flipped the max/min thermometer in the late afternoon, allowing hot days to be double counted. When the weather tuned cooler, Phoenixville often missed or lagged the cooling observed at other sites. Note that the difference vs neighboring sites was largest on the coolest days. A good illustration of time of day bias. Despite the obvious warm season bias at Phoenixville in the mid-20'th Century, this June was the 6'th warmest ever at Phoenixville and, as shown below, the year to-date is running #2 behind 2012. Strong evidence that Phoenixville (and Chester County) is warmer now than the 20'th century.

You didn't understand my point. The best way to evaluate whether NOAA the Chesco long-term temperature trend correct is to use reliable stations with long-term records, your house included. The stations I trust, including your house, both inside and outside the county, have similar trends as NOAA. I don't care about a one month temperature in your network. I care about the long-term trend. The problem with comparing your average and NOAA. Is that your average is not the County average temperature. Its the average of the group of weather stations. You are making no attempt to match your station network to the characteristics of the county. There are two problems. The older stations are not representative of the county as a whole and the network is changing in time. As illustrated upthread, If the network changes in time a simple average will skew the trend. Your current network appears to be more representative of the county, but that means its a poor match to the older stations. The current network is also largely made up of stations that aren't used for climate analysis. When you can't match the NOAAcounty average or trend, its due in large part to: 1) changes in your station network, 2) other method differences between you and NOAA, and 3) differences in the stations you are using and not due to NOAA's bias adjustments. In any case the bias adjustments for each station are all published. Not sure why you feel a need to overestimate the bias adjustmets and misrepresent their true nature.. The best way to evaluate the bias adjustments that are given to individual stations is to compare the station to other stations in the region. When I do that I don't see a problem with NOAA's adjustments. Phoenxiville, West Chester and Coatesville 1SW all deserve the bias adjustments that I have checked.

And you just complained about me cherry-picking one month. For all your complaints, NOAA matches the recent trend at your house closely. How is that? Simple, your house doesn't change from year-to-year. Making your house a much better platform to judge if temperatures are changing with time than a network of stations that is constantly changing.

You aren't paying attention. The charts I posted upthread show that Phoenixville ran warm for decades.

As usual you are carefully cherry-picking your "facts" to preserve your worldview. Using weather data outside the county makes perfect sense. Weather data is correlated for hundreds of miles and the historic data within the county is very limited. Bottom-line: You don't like NOAA's result; and you'll dismiss/deny any evidence that supports NOAA. As I said above, you live in a different Chester County than I do. The Newark Ag farm is less than 10 miles away from me. I want that data reflected in my climate history. Its just as relevant as any station in Chester County. Not that you couldn't do a better job analyzing the Chester County data by itself. For instance, per chart above, you don't need data from outside the county to bias adjust the Coatesville wartime warm spike. You also don't need any data outside the county to identify the excessive heat at Phoenixville in the 30s, 40s, and 50s, just increases the confidence in doing so. Not surprising that you are way off for Chesco in that period.

September 10 in Caribbean https://x.com/BMcNoldy/status/1807047443676340225

A little surprised by the quote below. This year seems like a close call vs 2023 to me. Its warm this year; but, last year was super warm for a developing nino. "By the end of 2024, global mean temperature will have declined significantly, but the annual 2024 global temperature should readily exceed the prior (2023) record."

Easy explanation. The older Chesco stations aren't representative of the County as a whole: low elevation, in the warmer part of the county, and they tend to be biased warm; and, in addition, NOAA is using weather stations that are outside of Chester County. NOAA takes advantage of the fact that weather station data is correlated for hundreds of miles. Instead of four stations, tens to hundreds of stations can provide relevant data on Chester County temperatures. Per chart below NOAA has Chester County right where it should be: closer to Allentown than Philadelphia. Note however that Coatesville is very close to NOAA in 1940 and after 1947, further confirming the NOAA the result. In contrast, by just averaging the Chesco stations, you think Chester County is as warm as the Philadelphia Airport or the Newark Delaware Ag farm during the war years. A silly result.

No, that's not what I am saying at all. We have plenty of weather stations outside of Chester County to help. The first thing scientists did when stating to develop climate information was investigate how well weather stations were correlated. Turns out they are very well correlated and the correlation extends to roughly 1000 km depending on the region and season. NOAA is looking at much more data when estimating Chester County temperatures and bias adjustments than you are. You are also looking at the adjustments in the wrong way. The adjustments are station-specific and are completely independent of the county average. The adjustments are based in the monthly change in temperature from one year to the next, not on the absolute value. As an example if Phoenixville warms by 1F, while the stations relevant to Phoenixville don't warm, Phoenixville's adjustment will change by 1F reflecting a probable change at the Phoenixville site. The shelter could have been moved for instance. Has nothing to do with the absolute temperature level at Phoenixville, or any other site, or the County average. The county average is only calculated as the last step after all the stations have been bias adjusted, and the individual station values has been mapped to a 5 by 5 km grid covering the entire country. Furthermore Phoenixville, West Chester and Kennett Square are all warmer locations than the county on average due to their location and elevation. Not surprising that they are warmer than the NOAA county average before or after adjustment. The "adjustments" that you are coming up with are much larger than NOAA's, not even close in many cases. You don't understand what NOAA is doing and are way off-base in your criticism.

Here are the annual average temperatures for the 3 main historic coop sites in Chester County. Further evidence that Phoenixville ran warm in the mid-20'th century. Not that the other sites don't have potential data inconsistencies. Need to include other stations to fully resolve. When I look at the Chesco coop data I don't see any reason to worry about the NOAA bias adjustments. Every one that I have checked in detail is justified.

Remember the big spike in 95F days in Phoenixville in the mid-20'th century that didn't occur at other regional sites (posted upthread). Per chart below, the July diurnal temperature range also increased in that period. The increase in diurnal range isn't matched at West Chester or Coatesville. Probably shelter-related as a non-aspirated thermometer will be more prone to shelter problems. Resetting the max/min thermometer in the late afternoon and evening as was the usual practice back in the day would compound the problem. In any case its clear that the raw data need a bias correction.

Could be mixing of dryer air from aloft. As the ground warms the depth of mixing increases.

Its on a flat earth with a green sky

LOL. Weather station data is highly correlated over hundreds of miles. All the stations in Chesco should move the same way. If they don't there is a problem with the measurements or analysis or both. You live in a different Chesco than I do. All I see is steady warming just like the rest of the world. Your own Chescowx data says so.

So your whole worldview depends on Chesco not warming. No wonder you ignore all the cracks in the edifice.

LOL so the low elevation stations cooled dramatically in one decade 1950s--->1960s, while the high elevation stations warmed.

Instead of "apples to apples" I see a lot of station changes that you aren't analyzing properly. The chart I posted previously illustrates the point. There are other differences between these stations besides elevation. For one thing the older data is from Coop stations, while the modern data is predominantly non-Coop. As an example Glenmoore Coop and your house are very similar in location and elevation, but differ by almost 2F on average. Simply taking an average of a group of stations skews the results if the stations operated over different periods. Per chart below all the "elevated" stations are warming but the average is flat.

Saw this tweet on the paper

Yes, a hot week and the models were even hotter fueling the hype. Funny thing is, in our current climate it probably wasn't that hot.

You haven't provided any evidence that you are comparing "apples" to "apples". No maps of station locations. No plots of station data. None. All you are providing is meaningless rhetoric like "machine based adjustments". All the evidence I have seen in this thread says you are way off: your own Chescowx series, The Mt Holly climate sites. the obviously excessive number of 95+ days at Phoenixville, etc. You have Chester County warmer than the Philadelphia airport at times in the past. No wonder you can't find much warming.

All that chart shows is that you don't understand how to analyze weather station data. Below is a recent blog article that provides good background. As described in the article, Its very difficult to determine the average temperature of a geographic area. Move a non-aspirated shelter closer to a house or in the shade of a tree and you get a different answer. Yet that is exactly what you are trying to do, determine the average temperature in Chester County. Furthermore you have collected a station population that changes dramatically with time making it unsuited for your task. In particular the older station population is very limited and clearly too warm: south and east, low elevation, in towns, and with older equipment. Much different than your modern stations. NOAA on-the-other-hand focuses on the change in temperature with time. Much easier to determine, as the changes with time are well correlated over hundreds of miles if the stations don't change. If the stations do change, the changes are easy to spot by inter-comparing stations. NOAA takes advantage of the dense station network in the US. If it is a relatively warm month in Chester County, it is also relatively warm in the surrounding counties and states. Hundreds of stations can provide information on the year-to-year change in Chester County's temperature. In short there is a night and day difference between NOAA's methods and your own. NOAA is informed by science and focusing on the easy with proven methods and a large dataset. You are uninformed by science and trying to the hard with a much smaller and poorly constructed dataset. It shows in the results. https://diagrammonkey.wordpress.com/2024/06/16/sticking-em-where-the-sun-dont-shine/

Yes, Phoenixville clearly was too warm in the middle of the 20'th century. Other stations don't have the spike in 95 degree days that Phoenixville does. You get the wrong answer about our local climate if you don't bias adjust the Phoenixville data. Same with many of the other Chesco coops. Fortunately, we have a dense regional network to catch the mistakes. As I said above thank goodness for NOAA.

Long video (10 min), but a good depiction of on-the-ground conditions in Broward County Florida during ytday's flash flooding. Note that several flooded homeowners had damaging flooding last year. https://twitter.com/AstuteGaba/status/1801253815121015183

Finally got a chance to look at some of the Chesco coop high temperature data. Chescowx's averaging hides a lot of sins. There is a big disparity among the two main long-term stations that I checked: West Chester, and Phoenixville, Most noticeable, the Phoenixville data in the 1930s and 1940s has many more 95F days than West Chester and the other Chesco stations. Phoenixville is also much warmer than Philadelphia in this period, with over 3x the 95 degree days as Philadelphia in the 1930s and 40s. With only a handful of Chesco stations, Phoenixville skews the county average, giving Chester County many more 95F days than Philadelphia in the 1930s and 40s. A silly result. Of course this isn't the only problem. The shifting station population also skews the results. Many of the modern stations, at higher elevation like Chesco's house, rarely reach 95. This data is a good advertisement for bias adjustment. Problematic data, like Phoenixville's in the 30s and 40s, is easy to spot and correct by inte-rcomparing stations. That is the benefit of a dense US station network. A benefit completely lost on climate deniers. Thank goodness we have NOAA to provide unbiased estimates of our past climate.