Solar photovoltaic module detection using laboratory and airborne imaging spectroscopy data.

Area of Science:

• Remote Sensing
• Photovoltaics
• Spectroscopy

Background:

• Solar panels (PV modules) are crucial for renewable energy, necessitating effective remote monitoring.
• Traditional PV detection methods using aerial photography have limitations.
• Imaging spectroscopy offers detailed spectral data for automated PV detection but faces challenges with spectral diversity.

Purpose of the Study:

• To develop and validate a physics-based approach for detecting solar PV modules using airborne imaging spectroscopy data.
• To address spectral intra-class variability and inter-class similarity in PV detection.
• To create a robust and transferable method for mapping PV installations.

Main Methods:

• Developed a physics-based approach analyzing physical absorption and reflection characteristics of PV modules.
• Normalized the Hydrocarbon Index (HI) to mitigate detection angle-induced variability.
• Utilized a large database including spectra-goniometric measurements, a HyMap spectral library, and HySpex imaging data.
• Applied six spectral indices to HySpex data for PV mapping and validated results.

Main Results:

• Successfully mapped PV modules in Oldenburg, Germany, using the developed approach.
• Achieved high overall, producer's, and user's accuracies in validation subsets.
• Demonstrated the approach's robustness and transferability through validation against diverse datasets.

Conclusions:

• The developed physics-based approach provides a robust, transferable, and applicable method for detecting PV modules using imaging spectroscopy.
• Airborne and spaceborne imaging spectroscopy hold significant potential for PV module identification and monitoring.
• This method offers a promising solution for automated and operational PV detection.