This research presents a computational framework for predicting temperatures in flax-reinforced composite sandwich footbridges. The case study is a 15m-span footbridge situated in Almere, the Netherlands, instrumented with 82 fiber Bragg grating (FBG) sensors and 8 thermocouples within an IoT-based structural health monitoring (SHM) system, enabling real-time evaluation. The main focus of this article is to enhance the understanding of structural temperature distributions in FRP sandwich cross-sections for bridge engineering applications. Furthermore, providing a comprehensive and accurate framework for temperature normalization of FBG strain measurements. The methodology involves environmental data, point-by-point solar radiation analysis with sun sheltering, and thermal properties characterization, culminating in a numerical analysis using Abaqus. A k-d tree binary search is performed previous to the numerical analysis to solve compatibility between Abaqus and Rhino's meshing algorithms. The non-uniform heat flux is used as input for the numerical analysis and handled through UEXTERNALDB and DFLUX subroutines. The use of a sub-modeling technique, allows the isolation, in correspondence of the sensor locations, of one critical section of the bridge for detailed analysis. Verification against in-situ measurements demonstrates the framework's efficacy. It was concluded that the framework restitutes errors in the range of ±1 ÷ 3 °C across varying weather conditions, during different seasons, and consistently for multiple sensor locations. Finally, the results are used to compensate strain measurements of nearby selected FBG sensors from the temperature counterparts.