Vujičić et al. (2026) Regional Ground-Based IoT Solar Irradiance Monitoring: A Multi-Site Study Across Mountain, Rural, and Urban Environments

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Sensors
• Year: 2026
• Date: 2026-09-11
• Authors: Dejan Vujičić, Dušan Marković, Pranay Anandbabu Obla, Shrihari Rajeev Kulkarni, Zoran Stamenković, Siniša Randjić
• DOI: 10.3390/s26185781

Research Groups

• Institute of Physics Belgrade, Serbia
• Faculty of Electrical Engineering, University of Belgrade, Serbia

Short Summary

This study investigates solar irradiance in a part of central Serbia using a low-cost IoT sensor network and compares the results to satellite reference data. The research reveals significant differences in measured irradiances at various locations due to terrain types.

Objective

• Investigate the accuracy of ground-based solar irradiance measurements in different terrain types

Study Configuration

• Spatial Scale: Central Serbia, with three measurement locations: mountain slope, open field near a village, and city center
• Temporal Scale: 25-year monthly NASA POWER record for the same cell used as training data

Methodology and Data

• Models used: Gradient boosting (XGBoost), recurrent neural networks, convolutional neural networks, fully connected networks, graph-based models, and seasonal models with calendar features
• Data sources: Satellite reference data from NASA POWER, ground-based IoT sensor network measurements

Main Results

• Measured mean daytime irradiances at three locations: 267.7 W/m2 (rural), 351.5 W/m2 (open field), and 19.6 W/m2 (urban)
• Best agreement with satellite reference found at rural station (R2 = 0.567)
• Significant shading bias at mountain station (MBE = −138.9 W/m2) and attenuation of signal at urban station
• Trained models achieved high accuracy: XGBoost R2 = 0.998, hybrid TCN-GNN R2 = 0.968, plain ReLU network R2 = 0.912

Contributions

• First multi-site ground-based IoT irradiance record for central Serbia with different terrain types
• Development of edge–cloud partitioning for local inference and bias correction

Funding

• This research was funded by the Ministry of Education, Science and Technological Development of the Republic of Serbia (project number 451-03-68/2020-14/200002)

Citation

@article{Vujičić2026Regional,
  author = {Vujičić, Dejan and Marković, Dušan and Obla, Pranay Anandbabu and Kulkarni, Shrihari Rajeev and Stamenković, Zoran and Randjić, Siniša},
  title = {Regional Ground-Based IoT Solar Irradiance Monitoring: A Multi-Site Study Across Mountain, Rural, and Urban Environments},
  journal = {Sensors},
  year = {2026},
  doi = {10.3390/s26185781},
  url = {https://doi.org/10.3390/s26185781}
}