Has Warming Changed the Asphalt Binder Grade? Superpave Design Temperatures at 20 US Airports, 1966-1995 Versus 1996-2025, the Reliability Actually Achieved by Old Designs, and a Simple Law for Grade Changes

Superpave asphalt binders are chosen from two climate statistics: the average daytime high over the hottest 7 days of the year and the coldest temperature of the year, each taken at 98% reliability. Using GHCN-Daily records for 20 US airports, I recompute the SHRP design temperatures for 1966-1995 and 1996-2025. The annual minimum temperature rose at all 20 stations, by a median 2.3 C, and its trend over 1951-2025 is positive and significant at every station (median 0.54 C per decade). The hottest-week high barely moved (median change 0.14 C; trend 0.07 C per decade, significant at 7 stations). As a result the low-temperature grade became one step warmer at 8 stations and one step colder at 1 (Boston, where the year-to-year spread rose), while the high-temperature grade rose at only one (Seattle, from PG 52 to PG 58). I prove that if a design value's position within its 6 C grade band is uniform, a shift of d degrees changes the grade with probability d/6. This predicts 10.1 low-grade changes and 2.0 high-grade changes, against 9 and 1 observed. Old 98% high-temperature designs still deliver a median 97.8% reliability in the new climate, but only 79.9% at Houston and 90.3% at Dallas-Fort Worth. I also show that a trend inside a 30-year window biases the window's 98% design value by -beta(n-1)/2 plus a small spread term, which at the median low-temperature trend makes the low design temperature 0.86 C colder than the end-of-window climate requires.