Flexible pavements are typically designed using historical climate but are challenged by future climate change. Quantifying impacts of climate change on pavement service life can assist road authorities in planning for climate adaptation and, eventually, build climate resilience into road infrastructure design and management. In this study, a novel data-driven methodology is developed in order to quantify impacts of climate change on pavement service life in locations where Falling Weight Deflectometer (FWD) data are continuously measured, by means of: 1) training a supervised model (linear regression or Artificial Neural Networks, ANN) using historical climate data, maintenance, and traffic data as the candidate inputs and pavement layer stiffness back calculated from FWD testing as the outputs; 2) predicting layer stiffness using statistically downscaled future climate projections for three Coupled Model Intercomparison Project Phase 5 global climate models and three greenhouse gas concentration scenarios for four future 20-year periods; and 3) estimating changes in pavement stiffness and service lives due to climate change. A case study performed on a pavement section in Minnesota has shown that pavement layer stiffness will have a long-term reduction under future climate and the investigated pavement will lose up to 22.5% service life at the end of the century (2080–2099) from the 20 years’ service life compared to the baseline climate (1979–1998).

柔性人行道通常使用历史气候进行设计，但会受到未来气候变化的挑战。量化气候变化对路面使用寿命的影响可以帮助道路当局进行气候适应规划，并最终将气候适应力纳入道路基础设施的设计和管理中。在这项研究中，开发了一种新颖的数据驱动方法，以便通过以下方式量化气候变化对连续测量落锤挠度计（FWD）数据的位置的路面使用寿命的影响：1）训练监督模型（线性回归或人工神经网络（ANN），将历史气候数据，维护和交通数据用作候选输入，并从FWD测试计算出的路面层刚度作为输出；2）使用统计学上按比例缩减的未来气候预测来预测层刚度，这三个耦合模型比对项目阶段5的全球气候模型和四个未来20年期间的三种温室气体浓度情景；3）估算由于气候变化导致的路面刚度和使用寿命的变化。在明尼苏达州的一个人行道上进行的案例研究表明，在未来的气候下，人行道的刚度将长期降低，到本世纪末（2080-2099年），被调查的人行道将失去高达22.5％的使用寿命。相较于基准气候（1979-1998年），其20年的使用寿命。