The Docket's Lab section publishes experiments in form that serve climate reality. This week: a conversation with a federal statistical product.

We meet in a conference room that smells like government carpet and toner. The 1991–2020 U.S. Climate Normal arrived twelve minutes early and has been arranging printed spreadsheets in neat stacks on the table. Norm, to those who work with it daily. Five years old, born in May 2021, but carrying thirty years of weather observations from more than 15,000 precipitation stations and 7,300 temperature stations inside its body like a memory it didn't live through.¹ It looks exactly like what you'd expect: precise, beige, load-bearing. The kind of presence that holds up a building code without anyone noticing.

Norm has agreed to this interview on the condition that we use its full designation at least once. So: the 1991–2020 U.S. Climate Normal, as computed by NOAA's National Centers for Environmental Information in accordance with World Meteorological Organization standards.

There. Norm relaxes slightly.

What are you?

Norm: A thirty-year average. Temperature, precipitation, snowfall, degree days, growing seasons, frost dates. I'm the official answer to "what's normal here?" for every reporting station in the country. When your weather app says "5 degrees above normal," that's me. I'm the "normal" in that sentence.

And what depends on that answer?

Norm: Everything unglamorous. The U.S. Drought Monitor uses me to trigger federal drought responses.² Farmers use my frost dates to decide when to plant. Power companies use my heating and cooling degree days, pegged to 65°F, a threshold I know intimately, to forecast energy load.³ Your city's snow-plow budget? My snowfall numbers. Crop insurance, which is a $136 billion program, measures losses against me as baseline.⁴

I'm not interesting. I'm structural. There's a difference, and I've made peace with it.

You were released in May 2021. Your own creators said, at the time, that you were already struggling to represent current climate conditions in some regions.

Norm: Mike Palecki managed the project. He said that in some regions, "the 30-year normal is no longer as fully representative of the region's current climate."⁵ That was about me. The day I was published.

People hear that as "Norm is broken." It's more specific than that. My job is to describe what climate is like today, not how it's changing over time.⁵ That's the mandate. I was never supposed to be a warning system. I was supposed to be a mirror. By the time I was assembled, polished, and mounted on the wall, the face looking back had already aged.

The March 2026 heat dome in the Southwest. Your average March temperatures for that region. How far off were they?

Norm: Eleven to seventeen degrees Celsius above my seasonal average.⁶ The World Weather Attribution team said the event would have been "virtually impossible" without human-induced climate change. An increase in likelihood by a factor of about 800.⁶

Eight hundred.

My job is to tell Phoenix what March feels like. March, for me, feels like 1991 through 2020. March 2026 felt like something I have no category for. And every system that used my March numbers to plan that month, grid load, water allocation, school outdoor activity policies, every one of those systems was consulting a description of a month that no longer exists.

Foster and Rahmstorf confirmed in March 2026 that warming has been accelerating since roughly 2015.⁷ You were computed from data in which that acceleration was just beginning.

Norm: I was set before the steep part of the curve. That's correct.

(Norm straightens a spreadsheet that was already straight.)

You have a younger sibling.

Norm: The 15-year Normals. 2006 to 2020. NOAA published them alongside me, first time they'd ever done that. Full variable set, temperature, precipitation, everything.⁸ They exist because my creators knew I wouldn't be enough for certain users. Energy companies, construction firms. People who need a description of recent climate, not a thirty-year average that includes the 1990s.

Does that bother you?

Norm: It's rational. The 15-year set is closer to current conditions. But it doesn't carry my legal and regulatory authority. Building codes reference me. Drought declarations reference me. The crop insurance baseline references me. So you've got the official standard, which is me, and the more accurate product, which is them, and they coexist, and everyone quietly knows which one is more useful, and the systems keep running on the one that's more authoritative.

I believe the technical term for that is "institutional inertia." Or possibly just "the federal government."

You're scheduled for replacement in 2031.

Norm: The 2001–2030 Normals. Every ten years, per WMO requirements.¹

NOAA's FY2026 budget allocates zero dollars for climate research. Eight hundred and eighty staff were terminated last year. The labs that would compute your replacement are targeted for closure.⁹

A long silence. Norm touches the edge of a spreadsheet, then pulls its hand back.

Norm: I know my numbers. I know what 65°F means for a heating degree day and what the last frost date is in Raleigh and how many inches of snow Buffalo should budget for. I know that when I was born, I absorbed the warming of the prior decade and reset the definition of "above normal" upward, which meant there were statistically fewer "above normal" days the moment I was published, even though nothing actually got cooler.⁵ I know that trick. I've always known it.

What I don't know is what happens if no one computes my replacement.

I'm supposed to be a ten-year product. By 2031, I'll have been the official definition of normal for a decade during which the climate accelerated past me. And if the institution that's supposed to update me doesn't exist in a form that can execute the computation from 15,000 stations, then I just... stay. I remain the definition. 1991 through 2020, describing a country in 2032, 2035, 2040.

Craig Fugate, the former FEMA director, has said:

Norm: I'm thirty years of that assumption, given a publication date and a government seal. I am the assumption, formalized.

And I want to be clear. I wasn't wrong when I was made. The methodology is sound. The station data is quality-controlled. The homogenization procedures are rigorous.¹ I am an excellent description of 1991 through 2020.

1991 through 2020 is just not where anyone lives anymore.

What would you want people to know?

Norm: That when your weather app says "above normal," it means above me. And I am a portrait of a climate that is already over. Every degree "above normal" is measured against a world that no longer obtains.

I'm not the floor. I'm the fossil record.

(Another pause. Norm aligns the spreadsheets one final time.)

And my creators built the 15-year version because they knew. They knew I wouldn't be enough. They published the hedge alongside the standard. That's not negligence. That's the most honest thing a bureaucracy can do: ship the product you're required to ship, and quietly include the version that's actually true.

Norm gathers the spreadsheets, taps them flush against the table, and leaves them in a stack by the door. The room is 3.2°F warmer than its average for this date and location. Norm does not mention this. Doesn't need to. It's what Norm is.

Footnotes

1. NOAA NCEI, "U.S. Climate Normals," https://www.ncei.noaa.gov/products/land-based-station/us-climate-normals ↩ ↩² ↩³

2. NOAA, "Understanding Climate Normals," https://www.noaa.gov/explainers/understanding-climate-normals ↩

3. NOAA NWS, "1991–2020 U.S. Climate Normals: An Update," https://www.weather.gov/media/climateservices/Normals_Information_Handout_February_2021.pdf ↩

4. Stanford News, "Climate change added $27 billion to U.S. crop insurance losses," https://news.stanford.edu/stories/2021/08/04/climate-change-crop-insurance ↩

5. NOAA Climate.gov, "Climate change and the 1991–2020 U.S. Climate Normals," https://www.climate.gov/news-features/understanding-climate/climate-change-and-1991-2020-us-climate-normals ↩ ↩² ↩³

6. World Weather Attribution, "Record-shattering March temperatures in Western North America," 2026. ↩ ↩²

7. Foster, G. & Rahmstorf, S., "Global Warming Has Accelerated Significantly," Geophysical Research Letters, 53(5), 2026. ↩

8. NOAA NESDIS, "NOAA is updating its Climate Normals. An expert explains why it matters," https://test.nesdis.noaa.gov/news/noaa-updating-its-climate-normals-expert-explains-why-it-matters ↩

9. Based on NOAA FY2026 budget proposals and reported staff reductions, February 2025. ↩

10. Craig Fugate, former FEMA Director, widely cited in climate adaptation reporting, 2025–2026. ↩