What Austin Weather Actually Does to a Construction Schedule

Your baseline says August in Austin is a full production month. NOAA says August has averaged 19 days at or above 100°F for the last fifteen years.

I pulled every daily weather record from the Camp Mabry station - 1938 through last week - and counted the days that hurt a crew: measurable rain, triple-digit heat, freezing mornings. Not modeled. Counted.

Here's the part most out-of-state baselines get wrong: Austin doesn't lose its schedule to rain. Rain days hold steady at three to five a month all year, and real washouts - a half inch or more - run about one or two. The month that kills you is the one that looks driest. June through September stack up roughly 54 impacted days, almost all of it heat. August alone averages 20.

And 100° is the blunt threshold. Count days at 95°F or better - where afternoon production quietly dies even when nobody calls it - and August averages 27 of them. That's the whole month.

The winter risk is smaller but it's real and it's specific: January averages five freezing mornings, February four. That's a pour schedule problem, not a lost-month problem. Plan the concrete around the mornings and it goes away.

One more thing the data says that your rule of thumb doesn't. If you learned your Texas weather assumptions in the 1990s, the ground moved. The ten-year average for 100° days has gone from about 15 to about 44. 2011 hit 90. 2023 hit 80. A calendar carrying 1990s heat assumptions is roughly 30 days a year optimistic.

And the heat doesn't just cost production. It puts crews in the hospital. We laid OSHA's severe-injury reports over the same weather record: 657 heat-exposure hospitalizations reported in Texas since 2015, and construction is the number-one industry on that list. Month by month, injuries track the thermometer - the correlation across 120 months of data is 0.74.

But look closer at June. It carries nearly the injuries of August on a third of the hundred-degree days - 2.3 hospitalizations per 100° day against August's 1.0. The crews aren't acclimatized yet. June heat catches bodies still calibrated for May. The most dangerous heat of the year isn't the hottest. It's the first.

What we'd do with this in P6: build a heat calendar for exterior and concrete crews, separate from the interior calendar, with reduced capacity June through September instead of a fiction of full production - and treat June as the ramp-in month the injury data says it is, not a milder August. Put the winter pour risk on the calendar as restricted mornings in January and February. Load the heavy exterior push into October through April - the data shows those months average four impacted days, the cleanest working weather on the chart. And when a baseline shows suspended-slab pours running full rate through an Austin August, reject it before it becomes your problem.

All told, Austin averages about 97 weather-impacted days a year. Weather isn't a surprise. It's a dataset. If August sits on your critical path at full production rates, the schedule was wrong the day it was baselined.

Methodology

Data: NOAA GHCN-Daily, Austin Camp Mabry station (USW00013958), complete daily records 1938-present; monthly figures are 2010-2025 averages with zero missing precipitation or temperature days in the window. Quality-flagged observations excluded. Thresholds are planning proxies, not work rules: rain day = precipitation of 0.10 in or more; washout = 0.50 in or more; heat day = high of 100°F or more (95°F shown for the productivity lens); freeze = low of 32°F or less. Impacted days counts a day once even when two conditions overlap. Validation: computed 100°F-day counts match the documented Austin records for 2011 (90) and 2023 (80). Injury overlay: OSHA Severe Injury Reports (employer-reported hospitalizations, federal-jurisdiction states, 2015-2024), event = exposure to environmental heat, Texas statewide (657 cases; 157 in construction), correlated against Camp Mabry 100°F-day counts over the same 120 months (Pearson r = 0.74). Severe injuries are a proxy for heat stress, not sick time - no public dataset reports crew absence hours at monthly resolution. Site microclimates and work-hour timing vary; treat these as calendar planning inputs, not guarantees.