State by State Climate Series - Actual raw data vs. NCEI 1895 through 2025

[ChescoWx]

Thanks to all of you for the many suggestions to broaden the studies performed by the team at ChescoWx beyond a single county as we have done for Chester County PA! Based on this feedback we are starting a state-by-state series to illustrate how the actual historical climate records with a focus on how much the annual average temperatures have been systematically adjusted by NCEI with the launch of the USHCN project back in the later 1980's and 1990's. To be transparent at least for Chester County PA which was the original focus of the study the gap between the raw county record and NCEI’s Chester County series, we could only find support for about 1/3 of the Chester County adjustments based on documents we found, roughly another quarter could be the observation-time correction (not settled), and about 40% has no independent check. The actual Chester County climate data showed a slight but clearly not statistical warming trend from 1895 through 2025. While the revised NCEI data showed a clear statistically significant warming trend.

If any of you are able to provide detailed support for any of the statistical and methodological data estimates which are what the homogenized NCEI adjustments represent... please provide this detail for these estimates by state if you can - thanks!

We are undertaking this study following a just completed detailed review of 48 randomly selected based on similar size counties to Chester County PA. We selected 1 County in each of the 48 contiguous US States. The findings in those 48 US Counties were consistent with what we saw here in Chester County PA. We found a clear consistent pattern across all 48 stations across the United States of clear cooling of the past records and general warming adjustments to more recent decades. Below are the consolidated findings and the detail of each of those selected 48 US Counties across the entire United States. Based on that finding (see below) of the share of counties where NCEI's adjusted value is cooler than the raw readings by decade....we will start a state-by-state review to see how many states show significant warming or not based on both the actual and adjusted NCEI data. First up will be my home state of Pennsylvania! I hope you all enjoy this upcoming study series!!

All of these studies will also be presented and will be available for download at www.chescowx.com 

[ChescoWx]

Our first state of the Nationwide Study is now complete. First up is the state of Pennsylvania. We will release one state per day over the next 47 days. For much more detail including study design, by decade breakdowns, county by county maps and breakdowns for all State Counties that compare actual vs NCEI trends for each state please go to https://chescowx.com/raw-vs-ncei-state-by-state/ 

If interested in seeing your state next up on the list please let me know and we will do our best to release that state within 2 days. Thanks again!

The main finding for Pennsylvania is the actual readings show no warming at all since 1895!
NCEI's published series shows significant warming. Averaging what Pennsylvania's observers actually recorded, county by county and weighted by county area, gives a flat statewide trend from 1895 to 2025 (-0.02 °F per decade, not significant). NCEI's published statewide series warms +0.17 °F per decade, about 2.2 °F in total. The two diverge at +0.19 °F per decade (p < 0.001). NCEI's series runs 1.5–1.8 °F below the actual readings in every decade from 1895 through the 1940s, about 1.0 °F below in the 1950s and 0.7 °F in the 1960s. It is within about 0.35 °F from the 1970s on.

The short version

1. NCEI's 67 counties add up to its statewide series. Weighting NCEI's county values by area reproduces its Pennsylvania statewide series within 0.03 °F in every year. Building the state from counties is therefore a like-for-like comparison.
2. The actual readings show no significant statewide trend. They give -0.02 °F/decade (95% range -0.10 to +0.06). NCEI's series gives +0.17 (range +0.10 to +0.24). The ranges do not overlap.
3. The two trend lines are significantly different. The gap between NCEI's statewide series and the actual readings grows at +0.19 °F per decade (95% range +0.15 to +0.22, p < 0.001), so NCEI's adjustments add a statistically significant amount of warming. County by county, NCEI's adjustments add significant warming to the trend in 25 of 43 counties (58%) and significant cooling in 0.
4. NCEI cools the early record in nearly every county. From 1895 through the 1940s, NCEI's county series is cooler than the actual readings in 90–100% of counties in each decade, by a median of about −0.9 to −1.7 °F. The share is 77–81% in the 1950s–70s, about 53% in the 1980s and 1990s, and about 31–33% after 2000.
5. Recent decades match. From 2000 on, the median county difference is about 0.0 °F. NCEI is cooler than the actual readings in 32% of county-decades and warmer in 41%.
6. County verdicts flip toward warming. 43 counties have long enough raw records for a trend. In the actual readings, 8 (19%) show significant warming, 31 (72%) no significant trend, and 4 (9%) significant cooling. In NCEI's county series, 41 (95%) show significant warming and none show cooling. NCEI's trend is steeper in 42 of 43 counties; Philadelphia is the only exception.
7. The 1930s lose their warmth. In the actual readings the 1930s average +0.9 °F against the 1991–2020 normal, the warmest decade before the 2020s. NCEI puts the 1930s at -0.7 °F.
8. Chester County fits the pattern. The actual readings from Chester County's GHCN-Daily stations give -0.01 °F/decade (not significant), against +0.24 °F/decade (significant) for NCEI's Chester County series over the same years.

[ChescoWx]

A certain local PA MET just messaged me on X and said the number that stunned him most in the study was the flip from only 19% of counties showing any statistical warming in PA to with the NCEI adjustments it jumps to 95% of all counties. To quote him "things that make you go hmmm!"