Dielectric properties of a target describe its relative permittivity. The related dielectric constant is a measure of the electric characteristics of any surface material. It consists of two parts (permittivity and conductivity) that are both highly dependent on the moisture content of the material considered. Consequently, the amount of energy that is scattered at surface level depends on how much energy penetrates into the medium and how much is absorbed by the radar target.
With changes in the amount of liquid water that is stored in soils, vegetation or other radar targets, we can also observe variations in radar reflectivity. Thus, there is a strong relationship between radar backscatter, precipitation, soil moisture as well as plant health.
The term dielectric constant describes the electric permittivity of a given material as a ratio, relative to the permittivity of a vacuum. It is directly related to the water content of an object or surface, as well as the material.
The term dielectric constant can be misleading, since it is not constant over various materials, only for a specific material or object. Keep that in mind when you think about this parameter.
Dielectric constant of common land cover surfaces
Below you can find a table displaying common surfaces and land cover types as well as their respective dielectric constants.
Dependence of dielectric properties and wavelength
Between the two SAR parameters dielectric constant and wavelength exists a substantial relation that defines the strength of the radar echo we are measuring.
How strong the moisture content of a radar target influences the backscatter is therefore also relying on the wavelength that is utilised. For a fixed soil moisture content different wavelengths will also penetrate the soil with varying depths.
The influence of soil moisture on the radar backscatter
Soil moisture measurement using radar
While soil moisture only accounts for a very small portion of the global freshwater resources, it plays an important and complex role in large-scale energy cycles. The amount of water that is stored within soils does also have a significant impact on the radar backscatter we interpret in our radar images.
To measure soil moisture using radar sensors, models in which the influence of vegetation and surface roughness must be known, are necessary. In a first step, effects related to vegetation and surface roughness must be corrected. This step is necessary to correlate the backscatter signal to the dielectric constant. Then, an empirical relationship between the relative dielectric constant and soil moisture is used to calculate the volumetric soil moisture.
Especially agricultural fields can be subject to rapid changes in soil moisture content, which is also strongly dependent on the type of cultivation in the respective site. The figure below shows an example of how strong the water content, that is stored in the soil, can vary over rather short periods.
Water content in soils
Compared to sealed surfaces or bare rock, the composition of natural soils are quite different. The mixture of loose grains of soil, water and air produces an endless number of possible variations in its appearance. In remote sensing applications, focus lies on the bulk properties rather than the microstructures as these are mostly significantly smaller than the wavelength. The most important of these three variables, at least when it comes to a radar perspective, is the water content of the soil. When comparing different soils to each other, a substantial bandwidth in water content can be found.
Scattering properties of (wet) soils
If water gets in contact with soil, the water molecules adhere to the soil particles. When the soil moisture increases, the scattering increases likewise while emission decreases. This effect is shown in the figure on the right and true for the real as well as the imaginary part of the dielectric constant. Another process we can observe with increasing wetness in soils is that the penetration depth decreases. Since this effect is added to the variance of surface roughness and the wetness can vary significantly within different layers of the soil, radar measurements of soils (especially in wet conditions) can be quite complex to interpret.
In this interactive tool you can simulate different dielectric constants and how these influence the scattering behaviour.
The influence of the dielectric properties of vegetation on the radar backscatter signal
Influence of water content on vegetation backscatter
The radar echo is also influenced by the amount of moisture that is stored in vegetation, as it affects the absorption and propagation of electromagnetic energy. When the moisture content within the plant rises, the penetration of the radar signal through a vegetation canopy gets decreased. Consequently, drier plants will exhibit less backscatter in the radar data. Different types of vegetation exhibit varying water storages and therefore also vary in the backscatter behaviour. Studies in the early 1990’s already demonstrated the sensitivity of the ERS-1 satellite towards vegetation wetness.
However, not only the water content within the plant has a crucial impact on the backscatter signal but also the amount of water that is being found on top of the vegetation surface and on the Earth’s surface surrounding the vegetation. Studies showed that the backscatter in temperate forests can vary around 2-3 dB before and after a rainfall event, which shows the significant influence of the moisture content on the backscatter. For smaller vegetation this effect can be even stronger depending on the wavelength that is applied.
When a forest is flooded, this can change the backscatter signature completely for the respective area. Sensors operating at low frequency can penetrate forest canopy and thus detect standing water at ground level. L-band instruments exhibit characteristic bright radar signatures from forests that are currently flooded. This is due to the dihedral effects that were described earlier in this course, which lead to double-bounce scattering with increased radar return.
In contrast to a flooded forest, a non-inundated forestabsorbs more of the transmitted radiation through the ground as well as the canopy. When interpreting the radar image pixels, they will appear not as bright as in a flooded scenario. Sensors working with C-Band are less suitable (in most cases) to detect standing waters under canopies as they are not capable of ‘seeing’ through the canopy. Exceptions would be scenarios with open canopies or during leaf-off season.