Temperature is one of the most important factors affecting functional as well as structural performance of asphalt pavements with thick asphalt layer (>30 cm). For a successful pavement design, it is vital to accurately predict the pavement temperatures at various depths. However, most previous researches focused on the temperature predictions for conventional asphalt pavements, of which the asphalt thickness is less than 30 cm. This suggests their proposed models are applicable in top layers, but may not be so effective for temperature predictions at deeper depths. As a result, the primary objective of this research was to develop a statistical model to predict temperatures at deep depths. Three test sites were selected, and they were instrumented with a number of sensors and a data logger to record the pavement temperature hourly. Also, all test sections can provide meteorological monitoring to collect hourly air temperatures and hourly total solar radiation. The recorded meteorological conditions were found to have cumulative effect on the measured pavement temperatures at various depths. On basis of their relationship, a statistical regression was performed, and the temperature prediction model was determined as a function of depth, average air temperature and total solar radiation calculated in the cumulative time. For an improvement in applicability, historical mean monthly air temperatures were also incorporated into the mode. The accuracy and applicability of the improved model were validated by applying it to additional sites for which the measured pavement temperatures and meteorological data were available. Also, by comparing with existing models, the developed model was testified to be more effective for asphalt pavements with thick asphalt layer, promising the model’s potential use.

温度是影响具有较厚沥青层（> 30厘米）的沥青路面功能和结构性能的最重要因素之一。对于成功的路面设计，至关重要的是准确预测各种深度的路面温度。但是，大多数先前的研究都集中在常规沥青路面的温度预测上，其中沥青厚度小于30 cm。这表明他们提出的模型适用于顶层，但对于较深深度的温度预测可能没有那么有效。因此，这项研究的主要目的是建立一个统计模型来预测深处的温度。选择了三个测试地点，并为它们配备了多个传感器和一个数据记录仪，以每小时记录一次路面温度。还，所有测试部分都可以提供气象监测，以收集每小时的气温和每小时的总太阳辐射。发现记录的气象条件对不同深度的路面温度具有累积影响。根据它们之间的关系，进行统计回归，并根据深度，平均气温和在累计时间内计算出的总太阳辐射确定温度预测模型。为了提高适用性，该模式还包含了历史平均每月气温。改进模型的准确性和适用性通过将其应用到可获得测量的路面温度和气象数据的其他地点进行了验证。此外，通过与现有模型进行比较，