Two years ago, I came across an interesting article by Avi Steinberg in The New Yorker written about a book titled ‘Is that Kafka?’. This book is a collection of 99 items from Kafka’s life to overturn his negative stereotype, which describes him as a tortured and neurotic person. Item 13 presents the statement of 15 witnesses describing Kafka’s eye color. Surprisingly, the witnesses could not agree on a single color. Instead, they described his eyes as gray, blue, black and brown. How can that be possible? Can hues be so subjective from one person to the next? Color identification becomes even trickier when discussing it with people of different nationalities. For example, don’t get surprised if a person from Vietnam stubbornly assures that tree leaves and the sky are the same color: xanh.
At that time, even the most ordinary activities such as reading non-scientific posts made me think about the main topic of my thesis where I explore the suitability of remote sensing data to monitor vegetation health and well-being. This idea relies on the fact that it is possible to link the health of a certain ecological system of interest to their reflectance behavior (which defines their color). This reflectance can be captured by remote sensing instruments, which are usually mounted on aircrafts and satellites. After reading the article about Kafka, some questions immediately came to my mind: how ‘green’ is a healthy plant? How can we quantify its ‘greenness’?
Let me start with the basics. The solar radiation spectrum striking the Earth’s atmosphere spans a wavelength range of approximately 100 nm to 4000 nm. Once reflected by the Earth’s surface, the radiation spectrum decreases to a range of 400 nm to 2500 nm. This reflected optical spectrum can be divided into three different wavelength categories: visible (VIS – the one visible to the human eye), near infra-red (NIR) and shortwave infra-red (SWIR).
Healthy canopies of green vegetation have a very distinct interaction with certain portions of light. In the VIS region, leaf chlorophyll causes strong absorption of energy, primarily for use in photosynthesis. Pigments in chloroplasts are responsible for the absorption of solar radiation in the blue and red wavelengths of the visible spectrum. On the contrary, radiation in the green region (~ 530nm) is reflected, commonly giving to leaves their characteristic green color. At canopy level, some structural characteristics of plants influence the amount of reflectance. For example, the majority of ecosystems, in addition to green vegetation, contain aging or dead vegetation. These latter components can increase NIR reflectance (Figure 1).
Expanding this knowledge, a sort of ‘vegetation indices’ based on vegetation spectral responses has been used in the last decades. Among many of these indices, the so-called NDVI (Normalized Difference Vegetation Index) is one of the most successful of many attempts to quickly identify vegetated areas and their condition. This index can be interpreted as a proxy of vegetation well-being and can be recorded by remote sensors. Remote sensing becomes an alternative to monitoring vegetation changes with many advantages. For example, it permits to gather information about inaccessible sites and to replace costly and time-consuming data collection on the ground, while leaving the area of investigation undisturbed. In several remote sensing products, many years of worldwide data are available, making possible long term and past analyses. It also covers large areas at fast speed, which is a revolutionary change compared with the traditional field measurements.
There are countless remotely sensed products that can be used for environmental monitoring. In my personal experience, I used remote sensing data from various environments, such as the warm Kenyan savanna ( Ruiz-Pérez et al., 2017), the mild Mediterranean forest (Ruiz-Pérez et al., 2016) and the cold Scandinavian boreal forest. Three places where, due to economic reasons and/or inaccessibility, there are limited amount of field measurements and/or data with poor quality.
Lately, great strides have been made in remote sensing with a plethora of platforms – satellites, unmanned aerial systems (UAS), airplanes, balloons and helicopters. The growth of remote sensing technologies has lead to innovative applications and data assimilation techniques. Made up of 28 European countries, the European Cost Action Harmonious will co-ordinate efforts to establish common remote sensing monitoring practices, facilitate data transfer, and upgrade knowledge through networking, exchange and training (Figure 2). If you feel drawn by the curiosity of how much one can learn from the sky, don’t lose track of all these activities. I definitely will not!
Ruiz-Pérez, G., González-Sanchis, M., Del Campo, A. D., & Francés, F. (2016). Can a parsimonious model implemented with satellite data be used for modelling the vegetation dynamics and water cycle in water-controlled environments?. Ecological modelling, 324, 45-53.
Ruiz-Pérez, G., Koch, J., Manfreda, S., Caylor, K., & Francés, F. (2017). Calibration of a parsimonious distributed ecohydrological daily model in a data-scarce basin by exclusively using the spatio-temporal variation of NDVI. Hydrology and Earth System Sciences, 21(12), 6235.