Field data acquisition campaign study for physically based temperature modeling in complex urban environments

Modeling the spatial and temporal distributions of urban temperature is critical for understanding complex urban environments and developing adaptive countermeasures for intensifying urban heat. Statistical model and physically based model using data from existing weather observation networks, which often suffer from spatial representativeness, lead to unexplained discrepancies to those for real-world systems. The urban temperature forecasting accuracy can be enhanced through a custom-built mobile meteorological measurement system combined with an optimized field campaign. By integrating mobile data into physically based models, biases inherent in traditional observation networks can be minimized. In order to overcome the high costs and inefficiencies of mobile data collection, this study proposes an optimization method for field data acquisition campaign using multiple datasets under a complex topological and landuse environment. The proposed protocol improves high-resolution temperature prediction accuracy while minimizing operational costs.