Using artificial neural networks to model climate data to adapt transport infrastructure to climate change

  • 1 University of Architecture, Civil Engineering and Geodesy (UACEG) – Sofia, Bulgaria


This paper examines an innovative approach for modeling the influence of climate parameters on transport infrastructure using artificial neural networks. Through them, detailed climatic data are generated by geographical positions and monthly and annual maps are created for Bulgaria’s territory using the following parameters: surface temperature, diffuse fraction, horizontal solar irradiation, and average albedo of the terrain. Average ground temperature and monthly solar irradiation are essential for maintenance planning and developing strategies to adapt to extreme weather conditions, such as heat waves or frost, which can affect the condition and performance of the road surface. Average monthly temperatures can be used to design effective systems to prevent icing of road surfaces and improve drainage systems. By demonstrating the capabilities of accurate modeling and analysis, this paper highlights the importance of applying artificial neural networks in planning and improving the resilience of transport infrastructure against climate change.



  1. Republic of Bulgaria, National Strategy for Adaptation to Climate Change and Action Plan, 2019.
  2. St. Ivanova. Artificial neural networks for energy efficiency, UACEG – Sofia, 2022.
  3. Photovoltaic Geographical Information System (PVGIS):
  4. NASA POWER (Prediction of Worldwide Energy Resources):
  5. NASA Earth Observations:
  6. Home page:
  7. Page for climate data by geographical positions:
  8. Page for monthly and annual climate data maps:

Article full text

Download PDF