Imaging lichen water content with visible to mid-wave infrared (400–5500 nm) spectroscopy

• Published in: Remote sensing of environment Vol. 216; pp. 301 - 310
• Main Authors: Granlund, Lars, Keski-Saari, Sarita, Kumpula, Timo, Oksanen, Elina, Keinänen, Markku
• Format: Journal Article
• Language: English
• Published: New York Elsevier Inc 01.10.2018; Elsevier BV
• Subjects: Drying; heat; image analysis; Indices; Infrared imaging systems; Inspection; Lichen; Lichens; Moisture content; MWIR; Nondestructive testing; Rehydration; Remote sensing; Remote sensing indices; Short wave radiation; Species; Spectral imaging; Spectroscopy; Spectrum analysis; SWIR; VNIR; Water content; wavelengths; Remote sensing indices; Spectral imaging; Water content; MWIR; Indices; VNIR; SWIR; Lichen
• ISSN: 0034-4257; 1879-0704
• DOI: 10.1016/j.rse.2018.06.041

The unique adaptation of lichens to repeated cycles of drying and rehydration makes them ideal subjects for developing remote sensing methodologies for water content estimation. This laboratory-based study evaluates the suitability of simple ratios (SR) and normalized difference indices (NDI), from several spectral regions; visible to near infrared (VNIR, 400–1000 nm), short-wave infrared (SWIR, 1000–2500 nm) and mid-wave infrared (MWIR, 2500–5500 nm) in the water content estimation of seven lichen species. The utilization of a wide wavelength range and several different lichen species allowed the evaluation of the robustness of the water content indices. Our results show that although there was high variability between different lichen species in their spectral responses to increasing water content, the best indices predicted water content accurately for the combination of species (RMSE 16.3%, 10.3% and 13.9% for the test set, in the VNIR, SWIR and MWIR regions, respectively). Generally the NDI indices were found to be slightly better than the SR indices. We also demonstrate the capability of imaging spectroscopy in creating detailed heat maps of the water content with these indices. •Spectral changes caused by water are dependent on the lichen species.•SR and NDI indices can predict lichens water content accurately.•In the SWIR and MWIR regions, the best water content indices are species independent.•Spectral responses of lichens differ from plants in the VNIR region.•Hyperspectral imaging data can be used to create pixel-wise maps of water content.