A review of different multispectral indices for monitoring plant health in mid-mountain sites

The effects of climate change, such as drought and pest infestation, will pose new challenges for forest management in the coming years to ensure the preservation of biodiversity and vegetation balance. A combination of various sensor technologies enables early detection of changes and initiation of necessary mitigation steps. Here, hyperspectral cameras provide direct measurement of the health status on the plants themselves. The achievable spatial and spectral resolutions have been steadily increasing due to the use of drones instead of airplanes and satellites. Nevertheless, only canopy measurement is possible in this case. The measurement below the tree canopy can grant new insights and increase the resolution up to the level of the leaf. The aim of this work is to define the basic requirements for a spectral system suitable for this purpose. For these high-resolution spectral images of typical plants of the mid-mountain range during desiccation were acquired. On the basis of these, various vegetation indices were calculated and the influence of filter properties such as the half-width were simulated. During this investigation, a clear reaction to desiccation was observed in all samples after a brief period of time. Different vegetation indices show a comparable behavior despite the application of different wavelengths.