Spectral saturation in the remote sensing of high-density vegetation traits: A systematic review of progress, challenges, and prospects

The saturation of spectral reflectance within densely vegetated regions is a renowned challenge that has precluded the optimal use of broad-band remotely sensed data and its derivatives for vegetation monitoring. While several reviews on the remote sensing of biomass have been published to date, an attempt to document and better understand the spectral reflectance saturation of vegetation has largely remained elusive. This article provides a comprehensive bibliometric assessment of the spectral reflectance saturation problem. The review profiles historical developments and maps the current remote sensing landscape of high-density Aboveground biomass and Leaf Area Index, with emphasis on the physical principles, proxies and methodologies as well as exploring the challenges and opportunities thereof. The review showed a skewed distribution of research between the Global North and South as well as variability in signal saturation levels of sensors according to the type, structure, and species composition of vegetation. Signal saturation is also dependent on the type of sensor used with the wavelength position and polarisation playing a pivotal role. The review also showed frequent usage of SAR backscatter and Lidar-based sensors for large-scale mapping, particularly in forests as compared to optical sensors. The fusion of waveform lidar indices with other sensors provides unprecedented opportunities for solving signal saturation problems. While used at localised and laboratory scales, narrow-band vegetation indices from hyperspectral sensors also significantly improved high-density biomass estimation. It is concluded that despite improvements generally in sensor capabilities and algorithm development, there is no uniform method for improving biomass estimation in dense vegetation. This calls for further research on understanding the fundamental relationship between spectral reflectance measurements and vegetation type, leaf orientation, vertical and horizontal structural parameters in dense vegetated regions. The development of appropriate sensors, fusion of optical and microwave data and improvement of retrieval approaches that transcend vegetation types and their associated traits is critical.