“Weather” art thy climate? Climate representativeness in rigid pavement design

Climatic conditions significantly influence the performance of Jointed Plain Concrete Pavements (JPCPs). Variations in temperature, wind speed, solar radiation and other factors generate non-linear temperature distributions. These distributions result in curling stresses and eigenstresses that increase the fatigue damage and thus reduce the expected service life of the section. Mechanistic-Empirical JPCP design relies on historical weather data of several years to generate temperature profiles for performance prediction. However, determining the optimal duration of historical data needed for reliable predictions remains unknown. Limited datasets may overlook long-term trends and variability, while an overly extensive dataset may increase computational time without significantly improving accuracy. This study aims to investigate the climatological representativeness of various weather data sources for assessing JPCP performance. Weather data spanning 30 years, which is the accepted definition of climate representativeness, was examined in comparison with 15 years of data and a statistical dataset called a Typical Meteorological Year (TMY) for both periods. These datasets were used to numerically evaluate the temperature profiles through various JPCP sections, and the performance of those sections was assessed based on curling and eigenstresses developed. In particular, the thickness and the solar reflectance (albedo) of the JPCP sections were varied to understand the extent to which climate representativeness depends on these properties. The analysis revealed that the distributions of temperature differences and eigenstresses for 15 years, 30 years and their respective TMYs align closely. However, the 30 years of data consists of more extremes than 15 years period. The TMYs for both 30 years and 15 years period show similar results. Hence, for design, if the extreme events are not considered, then 15 years' TMY can be used. However, if extremes are to be considered, then 30 years of historical should be considered.