A Statistical Principal Component Regression-Based Approach to Modeling the Degradative Effects of Local Climate and Traffic on Airfield Pavement Performance

Abstract

Airfield pavement systems support the global economy, passenger travel, and national defense. Accurate pavement degradation predictions are critical inputs for maintenance and repair decisions, and when skillful, they may reduce the need for time-intensive, costly physical inspections that disrupt airfield operations. Existing airport pavement management systems (APMS) compute expected degradation as a function of pavement type and age, but they do not account for local climate and traffic conditions and they are not built to adapt to future changes in either mode of variability. This paper implements a bias-reduced statistical model that reveals the effects of local conditions using observed historical climatic and aircraft traffic data. Model performance is evaluated using a diverse data set from nine Air Force installations, encompassing three major Köppen-Geiger climate zones in the contiguous United States and representing a wide range of aircraft. Environmental factors are more impactful on pavement degradation than aircraft traffic; a climate-only model produces R2 values as high as 0.84, while traffic improves explained variance across installations (R2 = 0.86–0.97) for the most heavily trafficked pavement family. This work illustrates the impactful role of climate in pavement degradation and demands implementation into the current APMS. Airfield asset managers can use this adaptable framework to more accurately determine sources of local degradation and inform sustainable pavement design and management practices.