Asphalt Graded for Yesterday's Weather

Asphalt pavement binder grades (PG XX-YY) are selected based on historical 7-day maximum and minimum pavement temperatures at a given location. Climate change is pushing actual temperatures outside the design envelope, causing premature rutting (high-temperature failure) and cracking (rapid-cycling failure). A FHWA study found that by mid-century, 35% of U.S. roads may need higher-grade binders than currently specified. Each 1°C increase in average temperature accelerates asphalt aging by 3–4%, increasing maintenance costs and shortening pavement life. Current Superpave design specifications use 20-year historical weather data that systematically underestimates future thermal exposure.

Pavement maintenance is the single largest infrastructure expenditure for most transportation agencies. The US alone spends $40+ billion annually on road maintenance. If binder grades are systematically under-specified for future climate, pavement lifespans will shorten and maintenance budgets will be overwhelmed. Early adoption of climate-adjusted specifications could save billions in avoided premature rehabilitation.

AASHTO's Superpave system (developed in the 1990s) was a major advance in performance-graded binder selection but hard-codes historical climate data as the design input. Researchers have published "climate-adjusted" PG grade maps, but no state DOT has adopted them into standard practice. The fundamental mismatch: climate models provide probabilistic temperature projections (distributions with uncertainty ranges), while pavement design codes require deterministic inputs (a single PG grade). Translating from one to the other requires reliability-based design methods that the pavement engineering community hasn't adopted. Polymer-modified binders can handle wider temperature ranges but cost 30–60% more, making specification changes politically difficult without rigorous justification.

A reliability-based pavement design framework that directly ingests regional climate projections (including uncertainty bands) and outputs binder grade recommendations with explicit risk levels — e.g., "PG 70-22 provides 90% reliability through 2060 at this location." This requires coupling downscaled climate models with pavement thermal models and rutting/cracking prediction models, validated against actual field performance data from the FHWA Long-Term Pavement Performance (LTPP) database. The data and models exist separately; the gap is integration and validation.

A team could select a specific state DOT region, obtain downscaled climate projections from CMIP6, run the LTPPBind software to determine current vs. projected binder grade requirements, and quantify the economic impact of the mismatch. Civil engineering, climate science, and data analysis skills would be most relevant. The LTPP InfoPave database provides decades of pavement performance data for validation.