Where our load data comes from

Current status

Dataset status: verified, against ASCE 7-22, on 2026-08-22. All 50 records were replaced with per-point lookups at Risk Category II. Every record changed: 43 wind speeds, 49 ground snow loads, and 29 seismic design categories. The result pages no longer carry the unverified banner.

Which edition, and exactly which layer

ASCE 7-22, Risk Category II. The values come from the same layers that the ASCE Hazard Tool draws, queried one point at a time rather than read off a contour:

• Wind — ASCE's w2022_mri700 layer: the basic wind speed at a 700-year MRI, which is Risk Category II. 3-second gust at 33 ft, Exposure C, to the nearest mph.
• Ground snow — ASCE's s2022_RiskCategory2 layer, which ASCE labels “Snow Load (lb/ft^2)”, to the nearest psf. This is p_g, not roof snow load.
• Seismic — the USGS ASCE 7-22 service, at Risk Category II and Site Class D. It returns the category itself, derived from full-precision S_DS and S_D1, so no rounding of ours sits between the source and the letter.

Each record stores the latitude and longitude actually queried — the principal city's civic centre — so any figure here can be re-derived rather than taken on trust. Two metros needed the standard's special provisions rather than the national layers: Anchorage takes its ASCE 7-22 Alaska zone value, which is only valid below 500 ft, and Oahu sits in a special wind region where the mapped speed varies across the island and the state code governs.

We report individual numeric values with attribution. We do not reproduce ASCE's hazard maps — those are ASCE's work. Hazard data © American Society of Civil Engineers.

ASCE 7-22 ground snow is not ASCE 7-16 ground snow

This is the trap in the whole dataset, and it is worth more of your attention than any single number. ASCE 7-22 replaced the old 50-year ground snow map with reliability-targeted values that are strength-level, and in the same edition dropped the snow load factor in the strength combinations from 1.6 to 1.0. So the mapped number got substantially bigger while the factor applied to it got smaller.

Washington DC reads 25 psf under 7-16 and 62 psf here. Boston goes 40 to 65. The ratio is not a constant you can scale by: it is widest in marginal-snow climates, where year-to-year variance is largest, and narrowest where snow is heavy and dependable — Minneapolis moves only 50 to 58.

Do not carry a value from this page into a 7-16 workflow that still applies a 1.6 factor. That double-counts, and over-designs the roof by roughly 1.6×. Plenty of jurisdictions still enforce ASCE 7-16 through IBC 2018 or 2021, so check which edition governs at your site before you use any snow number, from us or anyone else.

Get the authoritative values here

• ASCE Hazard Tool — the primary source. Free, no login, and it will produce a downloadable report you can hand to a manufacturer. This is where our wind and snow figures come from.
• USGS building-code seismic service — free and needs no key. Give it a latitude, longitude, Risk Category and Site Class and it returns S_DS, S_D1 and the design category.
• Your local building department. Non-negotiable, and the step people skip. Jurisdictions adopt code editions on their own schedule and frequently amend them, so the governing values at your site may not match the current national standard at all.

Four ways a single number misleads

1. Metro resolution hides real variation. This is now the largest source of error left in the dataset, and it is not small. Sampling the same ASCE 7-22 snow layer a few miles out from each city centre: Boise reads 32 psf downtown and 161 in the foothills, Salt Lake City 38 against 121 on the Wasatch benches, Baltimore 59 against 65. Elevation does this across the whole Mountain West, and lake effect does it across Erie County, New York. A metro-level figure is a starting point, not a site value.
2. Ground snow is not roof snow. Roof load is derived from ground snow, and then drift against a wall, a valley, or a roof step routinely governs over the balanced case. On multi-gable roofs this is where the design problem actually lives.
3. Seismic Design Category moves with the soil. Our figures assume Site Class D. A geotechnical report can shift the category and change connection and bracing design after the frame has been priced.
4. Risk Category changes everything at once. These are Category II values. An assembly occupancy is III and an essential facility is IV, and moving up raises wind, snow, and seismic demand together.

How to help

If you design or permit buildings in one of these metros and a value here is wrong, tell us which metro, which value, and what the Hazard Tool returns. Corrections go into includes/loads.php, which is the single source for every page that shows a load figure, and which records the coordinate each value was read at so you can check our working.

The dataset is pinned to ASCE 7-22. When the next edition lands the whole set has to be re-queried, not patched: the ground snow basis already changed once between editions, and that is exactly the kind of change that does not survive a partial update.

Back to the design loads tool