Advantages of using radar to estimate vegetation variables: Part 1 of 2
The ability to obtain quantitative information about the vegetation via remote sensing is an increasing demand of professionals interested in a range of applications: from estimating forest carbon stocks, production of commercial forest inventories (see also Forest Inventory: Can we do without it? by Ian Moss), land management, to monitoring the impacts of human activities in a variety of ecosystems. A plethora of remote sensing technologies are becoming available with lower costs. Whether airborne or satellite, radar — in particular, SAR or Synthetic Aperture Radar — stands out as an active sensor able to provide data even with cloud cover.
Radar — stands out as an active sensor able to provide data even with cloud cover.
However, radar technology still triggers certain resistances due to its less intuitive processing and interpretation. As new software and libraries are developed, and both data and the ability to process it become accessible to a non-specialized public, the development of commercial applications for radar data remains the bottleneck before people can fully take advantage of radar technology. The quantitative assessment of the vegetation is one of these sub-utilized applications.
The longer the wavelength, the deeper Radar bands penetrate dense vegetation.
Typical available radar bands are X, C, L, and P bands. The longer the wavelength, the deeper Radar bands penetrate dense vegetation. While the X-band, for example, having roughly the size of a leaf, will reflect back straight from the canopy, longer wavelength bands — such as L and P-bands — will pass through the canopy and can even reach the ground underneath the canopy.
In general, people resort to the analysis of radar backscatter to assess vegetation attributes. Such analyses have a somewhat simple underlying principle: if one illuminates the landscape with radar waves, the vegetation will act as a barrier and will scatter these waves back to the radar where its intensity (or amplitude) is recorded. Thus, the amount of backscatter from the landscape is roughly proportional to how much vegetation — measured in biomass, volume, and other correlated variables — of the landscape. While this seemingly solves the problem, a few complexities add difficulties to it.
First, the relationship between vegetation biomass and radar back scatter intensity is nonlinear. As the vegetation biomass goes up, this relationship levels off, i.e., it saturates. As a result, the quantification of forest attributes loses efficacy in very dense vegetation. In addition, available radar bands interact differently with the vegetation. Due to their physical properties, different radar bands saturate at different biomass levels. As an example, the L-band tends to saturate at 100 ton/ha of forest biomass.
As a result, the quantification of forest attributes loses efficacy in very dense vegetation.
Alternatives to overcome these limitations exist and involve a wise choice of radar bands, and/or a combination thereof, in combination with processing and analytical techniques.
Learn more about advantages and limitations to using radar to estimate vegetation attributes in Part 2.
Read more about Tesera’s approach to high resolution forest inventory.
Dr. Julianno Sambatti is a Data Scientist working on high resolution inventory solutions at Tesera.com
