Compare the heating and cooling calendar in 24 US cities. We analyzed NOAA's monthly climate normals and kept each airport station, measurement period and quality flag attached to the numbers.
24
city reference stations 12 months · 65°F base 1991–2020 climate normals
Analysis by Lonta AI · NOAA source files checked October 9, 2026 · Version 1.0
Put two local calendars side by side
Choose airport stations near your markets. Both charts use the same scale and start at zero. These are historical climate normals, not this year's weather or a forecast of leads.
Cooling degree days · 65°F base · 1991–2020 normals · Dallas–Fort Worth and Phoenix
First market
Dallas–Fort Worth, TX
DALLAS FT WORTH AP, TX US NOAA station USW00003927
* NOAA flag X: a nonzero value rounded to zero. The three-month share describes weather concentration, not a share of annual calls, jobs or advertising spend. Station-year coverage and completeness flags are in the records below.
One summer calendar misses local differences
Even Dallas and Phoenix peak in different months.
The highest cooling normal is 643.0 °F-days in August at Dallas–Fort Worth Airport, and 945.4 °F-days in July at Phoenix Sky Harbor. A single “July is the peak” calendar loses that distinction.
Three summer months represent different shares of the year.
The three largest cooling months contain 83.3% of Boston Logan's 12-month cooling total and 37.0% at Miami International. This is weather concentration. It does not say what fraction of each contractor's jobs happens then.
Coastal stations add another complication: San Francisco International's largest cooling normal occurs in September. Los Angeles International peaks in August. An airport can also be cooler than an inland service area, so match the reference point to the addresses your crews actually reach.
Use a local calendar to choose what to check: staffed phone hours, technician availability, maintenance outreach and landing-page promises. Your own completed jobs, equipment mix and service-area conditions determine the business plan.
The 24 station calendars at a glance
Peak means the largest published monthly normal at this station. The three-month share uses the three largest cooling months, which need not be June–August. This selection is not a ranking or a representative sample of all US markets.
NOAA 1991–2020 normals · heating and cooling base 65°F
Open a station record below for all twelve months and the underlying completeness/year flags. A monthly total also reflects month length; it is not a daily intensity measure.
Every month, with its source
S = Standard; R = Representative. The number beside the flag is NOAA's years field for that variable and month. X marks a nonzero value rounded to zero. The downloads retain these fields separately.
Atlanta, GA · USW00013874
ATLANTA HARTSFIELD INTL AP, GA US Latitude 33.63 · Longitude -84.4417 · Elevation 307.8 m
Lowest combined heating + cooling normal: April. This is an unweighted weather calculation, not a prediction of a slow business month.
Monthly degree days (°F-days) · 65°F base · 1991–2020
We selected 24 US city references to show different regional patterns, using one identifiable airport station for each. This is a purposive selection, not a population-weighted national estimate, a list of the biggest markets or a sample of our clients. Station names and coordinates are NOAA's, including historical airport names in the source files.
On October 9, 2026, we downloaded each station's original monthly file from NOAA NCEI's 1991–2020 U.S. Climate Normals. We extracted MLY-HTDD-NORMAL and MLY-CLDD-NORMAL, their measurement and completeness flags, and their years fields. Both variables use a 65°F base. There are 288 station-month records and 576 degree-day values; months and variables are not independent weather stations.
What a degree day means
A degree day accumulates temperature departure from a base. In the National Weather Service's daily explanation, a 78°F daily mean gives 13 cooling degree days above 65°F; a 29°F mean gives 36 heating degree days below it. A monthly total combines days. It does not count days with broken air conditioners or households ready to buy.
We use NOAA's published monthly degree-day variables directly. We do not estimate them by subtracting 65°F from a monthly mean temperature. NOAA's constrained daily temperatures and normals methodology supply the original calculations.
What we calculated
Peak month: the month with the largest heating or cooling normal. All tied peaks are retained.
12-month total: the sum of the twelve published monthly values, rather than a separately downloaded annual normal.
Three-month concentration: the sum of the three largest monthly values divided by that 12-month total. Equal values at the third-month cutoff use calendar order; the share is unchanged by that tie choice.
Lowest combined month: the minimum monthly heating + cooling total, without business weights. All tied minima are retained.
Coverage and limits
All 576 selected values are present. NOAA marks 526 Standard and 50 Representative; 13 have an X measurement flag for rounding to zero. These flags are copied from the source, not reclassified from their year counts. NOAA can estimate missing years using surrounding stations. A 30-year normal therefore does not mean every value has 30 complete observed years.
The period is 1991–2020. It describes a historical baseline, not current weather, recent warming or an upcoming heat wave. Airport siting, coastlines, elevation and urban conditions can make your service area differ. Humidity, equipment stock, household income, outages, maintenance plans, auction competition and your capacity are outside this dataset.
No jobs, leads, search volumes, ad-account records or revenue observations are included. The calculations do not estimate contractor demand or advertising lift. The checks establish that the selected records and calculations match the published NOAA files; they are not an independent audit of NOAA's underlying climate model.
Download the data. Keep the station attached.
Free, without registration. The 288-row CSV and JSON keep coordinates, period, units, flags, year counts, original station URLs and source-file hashes. The chart summarizes the peak months; the complete records remain the reference.
Lonta AI. HVAC Seasonality Calendar: 24 US City Reference Stations. Version 1.0. Analysis of NOAA NCEI's 1991–2020 monthly climate normals. Source files retrieved October 9, 2026. lonta.io/research/hvac-seasonality/.
Cite NOAA for the published climate values and Lonta for the selection, comparison and analysis. Preserve the station, base temperature and normals period when reusing a figure.
Primary sources and definitions
Checked October 9, 2026. Each station's original CSV is linked in its record.
Heating and cooling exposure has a local calendar. In this selected NOAA dataset, the largest monthly cooling normal occurs in July at Phoenix Sky Harbor, August at Dallas–Fort Worth Airport and September at San Francisco International. Those station normals do not measure the months with the most service calls or jobs.
Are these current weather data?
No. These are NOAA's monthly climate normals for 1991–2020, downloaded October 9, 2026. A normal describes a historical seasonal pattern. It is not a 2026 observation or a forecast.
Can I allocate my advertising budget using these percentages?
The percentages describe the concentration of degree days across months. They do not describe shares of annual leads, repairs, replacements or revenue. Set a budget from your own job economics, available capacity and measured demand, with climate as one planning input.
Why use an airport station instead of a city average?
These are identifiable NOAA station records with monthly degree-day normals, source files and quality flags. One airport is an explicit reference point, not a city-wide average. Coastal neighborhoods, elevation and inland service areas can differ from it.
Bring the calendar into your next account review
Choose a station near your actual service addresses. Check elevation and coastal or inland differences.
Compare the monthly pattern with your own repair, replacement and maintenance jobs. Keep new, returning and unknown customers separate.
Write staffed phone hours and available technician hours beside the busiest months.
Check the current forecast and actual weather before changing a live campaign. A climate normal is a planning baseline.
Use your own retained job contribution and capacity to set a spending ceiling.