Chester County PA - Analytical Battle of Actual vs. Altered Climate Data

[ChescoWx]

Below is how much faster the warming is just since 1960 at our local PHL Airport with their growing UHI problem. That warming slope is a wee bit different just around 25 to 35 miles west of the concrete and river warm oasis that is PHL Airport! 

[chubbs]

On 3/22/2026 at 4:52 PM, ChescoWx said:

Below is how much faster the warming is just since 1960 at our local PHL Airport with their growing UHI problem. That warming slope is a wee bit different just around 25 to 35 miles west of the concrete and river warm oasis that is PHL Airport! 

Reposting some of the charts I posted previously. Your line doesn't look anything like the raw data from individual stations. There is no significant difference in warming between individual Chester County stations and the Philadelphia Airport. Of course cooling station moves should be excluded. That's why the West Chester plot ends in 1969.

Per the table below, there are big changes in the Chester County station population that you aren't accounting for. In comparison, the Philadelphia airport heat island is mature and isn't changing much from decade to decade. If heat island is important, why ignore the movement of Chester County stations out of towns after World War II? 

[ChescoWx]

20 hours ago, chubbs said:

Reposting some of the charts I posted previously. Your line doesn't look anything like the raw data from individual stations. There is no significant difference in warming between individual Chester County stations and the Philadelphia Airport. Of course cooling station moves should be excluded. That's why the West Chester plot ends in 1969.

Per the table below, there are big changes in the Chester County station population that you aren't accounting for. In comparison, the Philadelphia airport heat island is mature and isn't changing much from decade to decade. If heat island is important, why ignore the movement of Chester County stations out of towns after World War II? 

Try consolidating your data you will find the story matches my above analytics.

[A-L-E-K]

denier board

[ChescoWx]

46 minutes ago, A-L-E-K said:

denier board

Climate realist board....

[ChescoWx]

Many climate alarmist like to point to the Time of Observation Bias (TOB) as solid "scientific adjustments" required to correct that bias. NOAA/NCEI in fact chose to chill every year at Coatesville from 1895 thru 1982. However, if anything when we look at the facts of when these observations were taken (see the below of COOP observation time/years) we see that with the exception of 11 years.... temperatures were in reality recorded in the morning. So if their rationale is correct the bias for Coatesville should in fact be too cool with all of those AM minimum temperature reports. Yet they chose to not warm those years - they actually made additional chilling adjustments.

[chubbs]

21 hours ago, ChescoWx said:

Try consolidating your data you will find the story matches my above analytics.

OK I took an initial stab at consolidating the data using Chester County's 3 long-term COOP stations. My consolidation doesn't look at all like your "analytics". Why? I only use periods without major station moves: 1949-2025 for Coatesville and Phoenixville and 1894-1969 for West Chester. I also use the 1949-1969 overlap period to take out the temperature difference between the 3 stations. While it doesn't look like your "analytics", my consolidation is a good match to the data collected at individual Chester County stations, posted above. That's gives my confidence in this approach and I plan to extend this method to the rest of the data.

 

[ChescoWx]

3 hours ago, chubbs said:

OK I took an initial stab at consolidating the data using Chester County's 3 long-term COOP stations. My consolidation doesn't look at all like your "analytics". Why? I only use periods without major station moves: 1949-2025 for Coatesville and Phoenixville and 1894-1969 for West Chester. I also use the 1949-1969 overlap period to take out the temperature difference between the 3 stations. While it doesn't look like your "analytics", my consolidation is a good match to the data collected at individual Chester County stations, posted above. That's gives my confidence in this approach and I plan to extend this method to the rest of the data.

 

 

Cherry picking always gives the desired answer.....

[chubbs]

1 hour ago, ChescoWx said:

Cherry picking always gives the desired answer.....

Yes, here's a good example of cherry picking. Do you have any specific technical complaints? I'll be adding other stations; but, why would the results change?. The other stations all have much shorter record lengths. Plus the modern stations are all warming rapidly in complete agreement with the Coatesville and Phoenixville data. 

[Cobalt]

5 hours ago, chubbs said:

Yes, here's a good example of cherry picking. Do you have any specific technical complaints? I'll be adding other stations; but, why would the results change?. The other stations all have much shorter record lengths. Plus the modern stations are all warming rapidly in complete agreement with the Coatesville and Phoenixville data. 

Unironically posting a graph like that should get you banned from any scientific community ever. The absolute failure of even Stats 101-level thinking is abhorrent to a degree where it feels like Chesco is playing a persona. 

[bdgwx]

On 3/26/2026 at 3:17 PM, Cobalt said:

Unironically posting a graph like that should get you banned from any scientific community ever. The absolute failure of even Stats 101-level thinking is abhorrent to a degree where it feels like Chesco is playing a persona. 

Yes. The @ChescoWx method is statistically useless. This can be demonstrated quantitatively by shifting the starting point (1998/04) and ending point (2026/01) of the trend multiple times and seeing how sensitive it is to cherry-picking the start and end points. For example, shifting the line to 1996/06 to 2024/04 yields a warming rate of +0.41 C.decade-1 using this method.

Using a more robust linear regression from 1998/04 to 2026/01 we get 0.1851 ± 0.092 C.decade-1 k=2 with an acceleration of an 0.1516 ± 0.035 C.decade-2 k=2. And this is starting from an El Nino and ending with a La Nina with a 5m lagged RONI of +2.4 and -0.8 respectively (or using the deprecated ONI it is +2.4 and -0.4 respectively). Note that uncertainties here are computed using a Heteroscedasticity and Autocorrelation Consistent model [Zeileis 2006].

 

[ChescoWx]

On 3/26/2026 at 4:17 PM, Cobalt said:

Unironically posting a graph like that should get you banned from any scientific community ever. The absolute failure of even Stats 101-level thinking is abhorrent to a degree where it feels like Chesco is playing a persona. 

Poor Cobalt!! Little does he seem to know.....

[ChescoWx]

On 3/27/2026 at 5:07 PM, bdgwx said:

Yes. The @ChescoWx method is statistically useless. This can be demonstrated quantitatively by shifting the starting point (1998/04) and ending point (2026/01) of the trend multiple times and seeing how sensitive it is to cherry-picking the start and end points. For example, shifting the line to 1996/06 to 2024/04 yields a warming rate of +0.41 C.decade-1 using this method.

Using a more robust linear regression from 1998/04 to 2026/01 we get 0.1851 ± 0.092 C.decade-1 k=2 with an acceleration of an 0.1516 ± 0.035 C.decade-2 k=2. And this is starting from an El Nino and ending with a La Nina with a 5m lagged RONI of +2.4 and -0.8 respectively (or using the deprecated ONI it is +2.4 and -0.4 respectively). Note that uncertainties here are computed using a Heteroscedasticity and Autocorrelation Consistent model [Zeileis 2006].

 

bdgwx just keep on cherry picking - you will always get the anwer you want!!

[Cobalt]

18 minutes ago, ChescoWx said:

Poor Cobalt!! Little does he seem to know.....

You have to be playing a persona, right? Because you can't just go on to say this

 

17 minutes ago, ChescoWx said:

bdgwx just keep on cherry picking - you will always get the anwer you want!!

and then go on to IMMEDIATELY post this

On 3/26/2026 at 10:33 AM, chubbs said:

 

it is beyond parody. 

[bdgwx]

On 3/29/2026 at 10:50 AM, ChescoWx said:

bdgwx just keep on cherry picking - you will always get the anwer you want!!

So choosing 2 data points out of a possible 335 to make a trend without any uncertainty analysis is good while taking all 335 with a robust uncertainty analysis is cherry-picking? Make that make sense.

And let me preempt your gaslighting. I didn't pick the 335 data point subset or claim that it would be sufficient to draw conclusions about whether the planet was warming or not. You did.

[ChescoWx]

8 minutes ago, bdgwx said:

So choosing 2 data points out of a possible 335 to make a trend without any uncertainty analysis is good while taking all 335 with a robust uncertainty analysis is cherry-picking? Make that make sense.

And let me preempt your gaslighting. I didn't pick the 335 data point subset or claim that it would be sufficient to draw conclusions about whether the planet was warming or not. You did.

No warming or cooling is my point you miss - just typical cyclical normal climate changes FTW!!

[ChescoWx]

On X today ABC Chief Meteorologist Ginger Zee answered a message I sent to her regarding the controversial NCEI temperature adjustments that have been made to the actual historical raw NWS Cooperative Data. Below is her response and a link with her "deep dive" report on the controversy. To be honest her report for the most part simply repeated the standard response saying "the science supports it" which is what you will always hear. For many folks with that answer from a professional meteorologist, they will simply nod their heads and say well if the science says it must be so. This in most cases will often stop any further questioning of the data. To Ginger's credit she agrees with my long-held stance in the video that science should always be questioned and evaluated. I certainly have questions and have to date not found any solid support for the consistent 2-to-3-degree chilling of the old data for every single year from 1895 through 2000. So, I will continue to question these adjustments. Science is always about questioning data not blindly following!

https://t.co/e5CFYdO803

Below is my response to Ginger with the data to try and support any adjustments to the raw data. Thanks Ginger but you didn't get into the deeper detail as to the explanation for the changes made to not only the ASOS sites at Airports you mentioned but to the NWS Cooperative Observer Data. A case in point is the long running Coatesville 1SW NWS Cooperative Station data for the philly burbs of Chester County PA with data from 1894 through 1983. There are many reasons given for these post observation ad hoc adjustments. The most common are station moves and time of observation adjustments. Below shows that for this station NCEI chilled 86 of the 89 years between 1894 through 1982. Below are the station moves grouped by year. These were all within a couple miles in each case with annual clear consistent cooling adjustments applied to the raw data. The time of observation adjustment also is not relevant in this case as only for 11 years (1910-21) was the daily observation taken only in the evening. So how and why exactly was this particular station chilled so consistently across 97% of all years?