Satellites are one of the most used means in agriculture to perform remote sensing. The satellite imagery in fact allows to monitor crops remotely in a precise and efficient way.
Spatial and temporal resolution of satellites
There are many satellites that acquire multispectral images from space: the most common are Sentinel-2and Landsat 8 (both used in Agricolus platform), Planetscope and Sky Sat. The images obtained have a spatial resolution of a few meters: Landsat 8 provides data with a spatial resolution of 30 m, while Sentinel-2 of 10, 20 or 60 m (depending on the band), Planetscope of 3 m and SkySat of 1 m.
The temporal resolution is in most cases regular. For example, Landsat 8 is available every 16 days, while Sentinel-2 is available every 3/5 days (depending on the zone). Planetscope and Skysat have a daily resolution.
The regular passage of the satellites determines the availability of the data in several phases of the growing season, but it is also important to underline that during the satellite transit, where the area under examination is covered by clouds, the data is not usable.
Vegetation indices obtained by satellite
When talking about satellite images, and remote sensing in general, it is necessary to introduce the concept of vegetation index to understand how they allow the monitoring of the health of crops without the need to go to the field.
Vegetation indexes are a key tool of Smart Farming: the use of satellite data and their correct interpretation reduce the interventions in the field and make sustainable, from the economic point of view, a structured crop scouting activity.
The indices can describe the vigor of the plant, providing a measure of its general health, or specific problems such as water stress or the amount of chlorophyll. There are therefore several types that describe the conditions of the vegetation, analyzing different bands of the electromagnetic spectrum.
The electromagnetic spectrum
The entire spectrum is divided into the visible part spectrum, which originates light, and the parts of the non-visible spectrum at longer and shorter wavelengths of the visible spectrum. In agriculture, the wavelengths of interest are those of the visible and infrared, which has wavelengths longer than those of visible light (0.8 to 13 μm).
Infrared can be further classified into near infrared (NIR), medium infrared and far infrared.
The most used are undoubtedly near infrared, or NIR, in addition to the wavelengths of green and red. These wavelengths are sensitive to changes in plant vigor. The wavelengths in SWIR are sensitive to changes in plant water stress.
Types of vegetation indexes
NDVI: it allows to evaluate the health of the vegetation, analyzing the reflectance of the vegetation in the Red and NIR bands.
GNDVI (Green-NDVI): it provides an indication of the health of the vegetation and reduces the saturation effect when the vegetation is particularly developed.
WDRVI: it analyzes the health of the vegetation and is particularly useful when the vegetation is well developed and lush and the other vegetation indices tend to saturate.
SAVI: allows to evaluate the conditions of vegetation development in the emergency and early stages of development, as it applies a correction to bare soil.
LAI: leaf area index that estimates the leaf area of the plant expressed in m2 on m2.
TCARI/OSAVI: specific index that allows to identify chlorotic areas within the field.
NDMI: specific index that evaluates the water content of the vegetation, therefore usable only with developed vegetation.
NMDI: can be used to assess the water content of the soil; in case of bare soil, a high index value indicates dry soil. In the presence of vegetation, a high index value indicates that the plant is not under water stress.