Construction Weather Delays by City: Weather Workday Index for 56 US Metros

How the index is built

Download the calculations (CSV)

Construction weather delays by city: the 56-metro index ranking

Desert heat and northern cold produce the most weather-limited days; the Pacific coast produces the fewest. The median large metro loses about 41.0 weekdays a year to conditions that limit outdoor work.

The CB&C Weather Workday Index estimates how many days per year at least one of three conditions occurs at a metro's main airport weather station: precipitation of 0.50 inch or more, a daily high of 32F or lower, or a daily high of 95F or higher. Values come from NOAA GHCN-Daily records for 1991 to 2020, the same period as the official U.S. Climate Normals.

Phoenix ranks first at 147.7 days per year. Tucson (120.4) and Las Vegas (112.7) follow. Minneapolis-St. Paul (93.7) is the only top-five metro driven by cold.

San Diego ranks last at 6.8 days per year. Los Angeles (9.7), San Jose (13.5), San Francisco (13.7), Seattle (25.1) and Portland (27.2) complete the bottom six. Tampa (36.5) has the fewest outside the West Coast.

24 of the 56 metros have 60 or more index days per year; 6 have fewer than 30. The median of 57.4 days equals 15.7% of the calendar year, or about 41.0 weekdays if weather is spread evenly across the week.

How the Weather Workday Index is built

Three thresholds tied to published industry and regulatory triggers, counted day by day so that a rainy freezing day counts once.

Weather-limited day: A calendar day at the metro's main airport station with precipitation of 0.50 inch (12.7 mm) or more, a maximum temperature of 32F or lower, or a maximum temperature of 95F or higher.

Cold threshold: the American Concrete Institute's ACI 306R-16 says cold weather concreting conditions exist when air temperature has fallen to, or is expected to fall below, 40F during the protection period. The primary index uses a stricter all-day freeze (high of 32F or lower) because frost alone does not stop most trades; a 40F variant is reported in the sensitivity test.

Heat threshold: California's 8 CCR 3395 requires high-heat procedures when the temperature equals or exceeds 95F, and requires shade above 80F. Washington's WAC 296-62-09547 requires high heat procedures, including mandatory cool-down rest of 10 minutes every 2 hours, at or above 90F, and 15 minutes every hour at or above 100F. The primary index uses 95F; 90F and 100F are tested.

Rain threshold: 0.50 inch in a day is an analytical choice, not a regulatory one. Public owners use other cut-offs: the Tennessee Board of Regents counts precipitation over 0.10 inch, and an Alabama DOT study used rainfall over 0.2 inch. Results at 0.25 inch and 1.00 inch are reported in the sensitivity test.

Double counting: each day is tested once against all three conditions. Summing the three components would overstate the index by an average of 0.8 days per metro, and by at most 1.8 days (Hartford), because heavy rain sometimes falls on a freezing or hot day.

Validation: the GHCN-Daily counts for rain of 0.50 inch or more, highs of 32F or lower, and highs of 90F and 100F match the official NOAA 1991 to 2020 normals for the same stations within 0.4 days per year at every station (mean absolute difference 0.03 days). NOAA's normals do not publish a 95F count or a combined count, which is why daily records are used.

Rain, cold or heat: what drives each city's weather days

Heavy rain is the most common limiting condition nationally, but the ranking is decided by heat in the Southwest and Texas and by cold in the upper Midwest and Great Lakes.

Across the 56 metros, heavy-rain days make up 39.6% of all condition-days, heat days 36.4% and freezing-high days 24.0% (1991 to 2020).

Rain is the largest single cause in 32 metros, from Miami, Atlanta and New York to Seattle and San Diego. Heat is largest in 13 metros, from Phoenix to Oklahoma City. Freezing highs are largest in 11 metros, including Minneapolis-St. Paul, Milwaukee, Grand Rapids, Buffalo, Chicago and Denver.

The mean metro has 23.6 heavy-rain days, 14.4 freezing-high days and 21.7 days at 95F or more. Cold and heat are skewed: the median metro has only 5.5 freezing-high days and 10.2 days at 95F, because a minority of metros has very large counts.

Heavy rain varies least. Miami has the most such days (41.4), followed by Birmingham (38.5) and Memphis (36.9). The desert metros have the fewest: Las Vegas 2.2, Phoenix 4.2.

Month by month: when weather days cluster

The national total has two peaks, July from heat and January from cold, with spring and autumn shoulders that carry far fewer weather-limited days.

The 56-metro average index is 9.1 days in July, 8.2 in August and 7.2 in January. It falls to 2.3 in April, 2.4 in October and 2.3 in November (1991 to 2020).

By season, the average metro has 23.4 index days in June to August, 17.6 in December to February, 9.4 in September to November and 8.6 in March to May.

January is the peak month in 23 metros and July in 21. Minneapolis-St. Paul peaks at 22.8 index days in January, Phoenix at 30.5 in July and Miami at 7.4 in August, when heavy rain is most frequent.

In the median metro, the three worst months hold 52.4% of the year's index days. Concentration is highest in Las Vegas (73.0%) and lowest in Louisville (37.5%), where rain spreads weather days across the year. For how contractor workload and payrolls move through the year, see construction seasonality by month and state.

Florida compared with the national picture

Florida metros rank low on the primary index because they rarely reach 95F or freeze, but they lead the country in heavy-rain days and in 90F days.

Miami (44.7 index days, rank 47 of 56), Orlando (46.9, rank 43), Jacksonville (50.0, rank 37) and Tampa (36.5, rank 50) all sit below the 56-metro median of 57.4 days (1991 to 2020).

Rain is the reason Florida is not lower. The four metros average 34.7 days with 0.50 inch of rain or more, against a median of 25.5. Miami's 41.4 is the highest of all 56 metros. In the official NOAA monthly normals, Miami's June to September months hold 59.2% of its heavy-rain days.

Heat is the swing factor. Florida highs seldom reach 95F (four-metro average 11.1 days), but they pass 90F on 93.9 days a year against a national median of 41.3. Counted at 90F, Orlando rises to rank 8 and Miami to rank 9. Air temperature also understates Florida heat stress because humidity is not included.

Tropical storms and hurricanes are not separated in this index; their rain counts only on the days it falls. For storm exposure by county, see Florida hurricane risk by county.

Sensitivity: how construction weather delay counts change with the thresholds

The level of the index depends heavily on the thresholds; the top and bottom of the ranking do not.

Under a strict set (rain 0.25 inch, any frost, heat 90F) the median metro has 162.1 index days. Under a lenient set (rain 1.00 inch, all-day freeze, heat 100F) it has 27.1. The primary set gives 57.4.

Rank correlation with the primary ranking is 0.96 when rain is counted at 0.25 inch, 0.90 with the ACI 40F cold test, 0.87 for the lenient set, 0.70 for the strict set and 0.62 with heat at 90F. The heat threshold matters most.

San Diego, Los Angeles, San Francisco and San Jose are the four lowest under every variant tested. Phoenix is first under every variant except the strict set, where frost-heavy Denver, Omaha and Minneapolis-St. Paul move to the top.

2021 to 2025 against the 1991 to 2020 normals

Recent years were hotter in Texas, Florida and the Southwest and milder in the Great Lakes, but five years is too short to call a trend.

Using complete station-years from GHCN-Daily, 27 of 56 metros averaged more index days in 2021 to 2025 than in 1991 to 2020; the median change was -0.7 days.

Houston rose from 78.6 to 103.0 days, San Antonio from 84.9 to 104.1, Austin from 89.9 to 105.6 and Orlando from 46.9 to 62.0. In each case more 95F days drove the increase: Orlando's rose from 15.3 to 36.8.

Chicago fell from 69.7 to 54.8 days, Buffalo from 72.2 to 57.4 and Milwaukee from 73.5 to 59.8, as all-day freezes became less frequent. Buffalo's freezing-high days fell from 48.7 to 36.2.

Year-to-year swings are large. Houston's single-year index ranged from 55.0 to 124.0 days within 1991 to 2020, so the 2021 to 2025 average sits inside the historical range.

How owners and contracts treat normal weather

The contract forms and specifications reviewed here expect normal weather to be absorbed in the schedule and treat abnormal weather as possible grounds for more time, which makes a city's normal count the baseline.

AIA Document A201-2017, Section 15.1.6.2, requires a weather claim for more time to be documented with data showing the weather was abnormal for the period, could not have been reasonably anticipated, and had an adverse effect on the scheduled construction (A201-2017 text). Section 8.3.1 lists adverse weather documented that way among causes for which the Contract Time may be extended.

For federal fixed-price construction, FAR 52.249-10 does not terminate a contractor's right to proceed for delays from unforeseeable causes, listing unusually severe weather among them, provided the contractor notifies the Contracting Officer within 10 days of the delay's start.

Some public owners publish their normal. The Tennessee Board of Regents sets a Standard Baseline of 89 calendar days a year (12 in January, 4 in September) that is included in the work and not eligible for extension. Its adverse weather test is broader than this index: precipitation over 0.10 inch, temperatures not above 32F by 10:00 a.m., sustained wind over 25 mph or standing snow over 1.00 inch. Nashville's index under the strict variant is 167.8 days; under the primary set it is 53.2.

An Auburn University study for the Alabama DOT (2023) used rainfall over 0.2 inch and a daily mean below 40F across 88 stations and set annual average available workdays of 185, 193 and 200 for Alabama's North, Central and South zones, after removing weekends, legal holidays, rainy days and cold days.

For how weather fits into project phases and schedules, see the commercial construction process.

Stations used and official NOAA normals

Each metro is represented by its main commercial airport station, which has the longest, most complete record; official normals are listed for reference.

The 56 metros are every metropolitan statistical area with a July 1, 2025 population of 1 million or more in the Census Bureau's Vintage 2025 estimates; together they hold 195,021,830 residents. One airport station per metro was chosen; New York uses LaGuardia, Washington uses Reagan National, Riverside-San Bernardino uses Riverside Municipal Airport and Sacramento uses Sacramento Executive Airport.

NOAA flags a normal as standard when data exist for at least 24 of the 30 years; gaps are filled from neighbor stations. Riverside and San Jose normals rest on 20 to 22 years of data, Austin-Bergstrom on 22 to 24 and Denver International on 25.

What the index cannot tell you

The index is a climate exposure measure at one point per metro, not a forecast or a measure of project delay.

• One station per metro: airports can differ from downtown or suburban sites, especially for coastal fog, urban heat and lake-effect snow.
• Air temperature only: GHCN-Daily has no humidity, so heat index is not measured. Humid metros such as Houston and Miami face more heat stress than the 95F count shows.
• Not included: wind, lightning, snow cover, ground saturation after rain (dry-out days), and hurricanes as distinct events.
• Trade differences: roofing, earthwork, concrete, steel erection and interior trades respond to different conditions; the index applies one rule to all.
• Normals describe 1991 to 2020. The 2021 to 2025 comparison is a check, not a trend estimate.

Key findings

1. The median of the 56 US metros with 1 million or more residents has 57.4 weather-limited days per year under the CB&C Weather Workday Index (1991 to 2020); the mean is 59.0.^[6][8]
2. Phoenix has the most weather-limited days of any large US metro: 147.7 per year, of which 144.6 are days at 95F or hotter (1991 to 2020).^[6]
3. San Diego has the fewest: 6.8 weather-limited days per year, followed by Los Angeles (9.7), San Jose (13.5) and San Francisco (13.7) (1991 to 2020).^[6]
4. The highest metro (Phoenix) has 21.7 times as many weather-limited days as the lowest (San Diego).^[6]
5. Minneapolis-St. Paul is the most cold-limited large metro, with 72.4 days a year on which the high stays at or below 32F, and ranks 4th in the index (93.7 days).^[6][2]
6. Heavy rain of 0.50 inch or more is the largest cause in 32 of 56 metros; heat is largest in 13 and freezing highs in 11 (1991 to 2020).^[6]
7. Miami averages 41.4 days a year with 0.50 inch of rain or more, the most of the 56 metros; Las Vegas averages 2.2, the fewest (1991 to 2020).^[6][2]
8. Heat at 95F or more is the largest cause in 13 metros, including all four Texas metros: Austin (69.9 days at 95F or more), Dallas-Fort Worth (63.1) and San Antonio (66.3) all exceed 60 such days a year.^[6]
9. Across the 56 metros the average index peaks in July at 9.1 days and in January at 7.2 days; April and November are the lowest months at 2.3 days each (1991 to 2020).^[6]
10. Weather days are seasonal: in the median metro the three worst months hold 52.4% of the year's index days; in Las Vegas they hold 73.0%.^[6]
11. Florida's four large metros average 44.5 index days a year, below the 56-metro median of 57.4; with heat counted at 90F instead of 95F, the Florida average rises to 118.0 against a median of 80.8.^[6]
12. Counting 90F instead of 95F as the heat limit moves the median index from 57.4 to 80.8 days and changes the ranking more than any other threshold choice (Spearman rank correlation 0.62).^[6][10][11]
13. In 2021 to 2025, Houston averaged 103.0 index days a year, 24.4 more than its 1991 to 2020 mean of 78.6; 95F days rose from 48.8 to 73.0.^[6]
14. In 2021 to 2025, Chicago averaged 54.8 index days, 14.9 fewer than in 1991 to 2020, as all-day freezes fell from 43.4 to 33.0 a year.^[6]
15. A Tennessee Board of Regents specification treats 89 calendar days a year of weather suspension as normal and not eligible for a time extension, using a 0.10 inch rain threshold and a 32F-by-10 a.m. test.^[16]

Summary for decision makers

• The CB&C Weather Workday Index counts days with 0.50 inch of rain or more, a high of 32F or lower, or a high of 95F or higher. The median of 56 large US metros is 57.4 days per year (1991 to 2020), about 41.0 weekdays.
• Phoenix (147.7), Tucson (120.4) and Las Vegas (112.7) top the index, almost entirely from 95F heat. Minneapolis-St. Paul (93.7) is the highest cold-driven metro.
• Coastal California has the fewest weather-limited days: San Diego 6.8, Los Angeles 9.7, San Jose 13.5 and San Francisco 13.7.
• Heavy rain is the largest single cause in 32 of 56 metros and accounts for 39.6% of all condition-days; heat accounts for 36.4% and freezing highs for 24.0%.
• Florida's four large metros average 44.5 index days, below the 57.4 median, but they have more heavy-rain days (34.7 vs 25.5) and far more 90F days (93.9 vs 41.3).
• Across the 56 metros, July (9.1 days) and January (7.2) carry the most weather-limited days; April and November (2.3 each) carry the fewest.
• In 2021 to 2025, Houston averaged 103.0 index days against 78.6 in 1991 to 2020, while Chicago fell from 69.7 to 54.8: heat days rose in the South and all-day freezes fell in the Great Lakes.

Frequently asked questions

Related guides, services and research

• The Commercial Construction Process, From Feasibility to Closeout
• Commercial Construction Project Checklist: Feasibility to Closeout
• Commercial Builders Near Me: Six Florida Markets We Cover
• Office to residential conversion programs: US registry of incentives, by-right laws and federal guidance (2026)
• Data Center Power Demand vs. Grid Capacity: Interconnection Queues by Region, 2026
• Construction GDP by State: Share of Each Economy, Real Growth and Output per Worker, 2025
• Commercial Construction Spending Tracker: August 2026 Data by Building Type

Methodology

Selection: all 56 US metropolitan statistical areas with a July 1, 2025 population of 1 million or more (Census Vintage 2025). Each is represented by its main commercial airport station in NOAA's GHCN-Daily (station IDs in table t2), the station class with the most complete long records.

Index rule: a day counts if precipitation is 0.50 inch (12.7 mm) or more, the maximum temperature is 32F (0.0C) or lower, or the maximum temperature is 95F (35.0C) or higher. Each day is tested against all three conditions and counted once. Thresholds follow ACI 306R-16 (cold weather begins at 40F, tested as a variant), California 8 CCR 3395 (high-heat procedures at 95F) and Washington WAC 296-62-09547 (high heat at 90F, tested as a variant). The rain cut-off is an analytical choice; 0.25 inch and 1.00 inch are tested.

Computation: GHCN-Daily values with a non-blank quality flag are dropped. A day is valid when both precipitation and maximum temperature are present. For each station, month and year with at least 20 valid days, the share of valid days meeting the rule is multiplied by the number of calendar days in that month. Monthly results are averaged over 1991 to 2020 and summed to an annual figure. The recency check uses only years in which all 12 months qualify. Scripts are in the scripts folder and re-run from the raw files.

Conflicts and validation: NOAA's official normals publish counts for rain of 0.50 inch or more, highs at or below 32F and highs at or above 90F and 100F, but not 95F or a combined count. The GHCN-Daily counts match the official normals within 0.4 days a year at every station for all four published counts. Where they differ, the normals in table t2 are the official figures; the index uses GHCN-Daily so the union and the 95F threshold can be computed consistently.

Limitations: one station per metro; air temperature without humidity; no wind, lightning, snow cover or dry-out days; trade-specific limits ignored. Riverside, San Jose, Austin and Denver have shorter records (20 to 25 years). Weekday equivalents assume weather is independent of the day of the week.

Exclusions: market research and vendor estimates of weather-related costs and productivity losses were reviewed on competing pages but excluded because they did not cite primary data (see excluded.md). Last updated 2026-10-18.

This report presents published data with its limitations. It is not legal, financial or investment advice.

Sources, ranked by quality

Tier 1: primary and official sources (16)

• [1] NOAA National Centers for Environmental Information. U.S. Climate Normals (1991 to 2020 normals, v1.0.1). 1991 to 2020; first released May 2021, 23 sites reissued 2023. Accessed 2026-10-18.
• [2] NOAA National Centers for Environmental Information. 1991 to 2020 Annual/Seasonal Normals, station files (one CSV per station; 56 files used). 1991 to 2020. Accessed 2026-10-18.
• [3] NOAA National Centers for Environmental Information. 1991 to 2020 Monthly Normals, station files (one CSV per station; 56 files used). 1991 to 2020. Accessed 2026-10-18.
• [4] NOAA National Centers for Environmental Information. Annual/Seasonal Normals Documentation, 1991 to 2020. 1991 to 2020. Accessed 2026-10-18.
• [5] NOAA National Centers for Environmental Information. Normals Calculation Methodology 2020. 1991 to 2020. Accessed 2026-10-18.
• [6] NOAA National Centers for Environmental Information. Global Historical Climatology Network Daily (GHCN-Daily), by-station files (56 files used). daily observations 1991 to 2025 used; files updated daily. Accessed 2026-10-18.
• [7] NOAA National Centers for Environmental Information. GHCN-Daily by-station file format readme. current. Accessed 2026-10-18.
• [8] U.S. Census Bureau. Metropolitan and Micropolitan Statistical Areas Population Totals, Vintage 2025 (cbsa-est2025-alldata.csv). July 1, 2025 estimates. Accessed 2026-10-18.
• [9] American Concrete Institute. ACI PRC-306-16 (306R-16) Guide to Cold Weather Concreting, publisher preview. 2016 edition. Accessed 2026-10-18.
• [10] California Department of Industrial Relations. California Code of Regulations, Title 8, Section 3395: Heat Illness Prevention in Outdoor Places of Employment. rule in force October 2026. Accessed 2026-10-18.
• [11] Washington State Legislature. WAC 296-62-09547: High heat procedures. rule in force October 2026. Accessed 2026-10-18.
• [12] U.S. Occupational Safety and Health Administration. Heat: Standards. page as of October 2026. Accessed 2026-10-18.
• [13] U.S. Occupational Safety and Health Administration. Heat Injury and Illness Prevention in Outdoor and Indoor Work Settings Rulemaking. status as of October 2026. Accessed 2026-10-18.
• [14] Acquisition.gov (General Services Administration). FAR 52.249-10 Default (Fixed-Price Construction). current FAR text. Accessed 2026-10-18.
• [16] Tennessee Board of Regents. Section 01 26 20 Weather Delays (Standard Baseline for Tennessee). specification posted 2015. Accessed 2026-10-18.
• [17] Auburn University Highway Research Center for the Alabama Department of Transportation. Determining the Average Available Workdays for ALDOT Construction Projects, Final Report 931-065 (September 15, 2023). weather data from 88 Alabama stations. Accessed 2026-10-18.

Tier 2: trade bodies, market research, company data (1)

• [15] American Institute of Architects (copy hosted by City of Rochester, NH). AIA Document A201-2017, General Conditions of the Contract for Construction (sample, posted with a City of Rochester, NH bid). 2017 edition. Accessed 2026-10-18.