This issue:

DInSAR, PSI, PSInSAR, …: make it clearer
Future trends and perspectives with X-band sensors
A new stream of research ...
Improving our CED (data center) hardware capacity

DInSAR, PSI, PSInSAR, …: make it clearer

During the last few years an increasing number of acronyms related to radar interferometry has appeared in the Earth Observation sector. It’s time to clarify what each of them means, within their historical context.

In the nineties, after the launch of the ESA-ERS satellites, many research groups started working on SAR (Synthetic Aperture Radar) interferometry to derive surface displacement maps.
InSAR (Interferometric SAR), DInSAR (Differential SAR Interferometry) and also DiffSAR were, and are, used synonymously in reference to this technology.

Traditional interferometry measures ground displacement occurring during the time interval between a pair of radar images acquired over the same area. Unfortunately, atmospheric artefacts and decorrelation overlap the signal, sometimes compromising the analysis.

In 1999 Politecnico di Milano (POLIMI) developed a new multi-image approach (not just two but more then twenty images), that took traditional interferometry a step further. This approach works on specific point targets (Permanent Scatterer or PS) on the ground.
POLIMI registered the patent (1999) and two international trademarks: PSInSAR™ and POLIMI PS Technique™; while our company (TRE) was established by POLIMI in 2000. TRE is the exclusive licensee of the PSInSAR™ technique. Hence PSInSAR™ refers to a specific algorithm developed and patented by POLIMI.

As the result of our success some research centres and other companies were inspired to develop their own algorithms which, like PSInSAR™, would overcome DInSAR’s limitations. Today these techniques are collectively referred to as Persistent Scatterer Interferometry or PSI techniques, within scientific literature. The term Persistent Scatterer Interferometry (PSI) was created by ESA to define the second generation of radar interferometry techniques. PSInSAR™ is a PSI technology but PSI and PSInSAR are not synonymous.

Future trends and perspectives with X-band sensors

Today the two data sources available for commercial applications of PSInSAR™ are the ENVISAT and RADARSAT satellites operated by the European Space Agency (ESA) and the Canadian Space Agency, respectively. Both satellite platforms have two C-band sensors allowing the generation of InSAR data-stacks.
Before the end of this year three new sensors will be available, namely: RADARSAT-2 (C-band), TerraSAR-X and the first Cosmo-SkyMed (both of them with X-band sensors, wavelength ~3 cm), while the Japanese sensor ALOS-PALSAR is already able to acquire L-band (wavelength ~24 cm) radar images. TerraSAR-X and the Cosmo-SkyMed satellites have been successfully launched last June and are now undergoing commissioning procedures.
It is then clear that, in the near future, SAR data should become easier to get and more widely used by both the scientific and commercial communities. Based on the technical features of these new sensors, it is possible to foresee a significant impact on applications, for example:

• The improved spatial accuracy (up to 1 m) will allow a higher PS density for surface deformation monitoring and a significant reduction in the dimensions of corner reflectors, which are used to create good radar targets when “natural” PS are unavailable. This improved resolution will greatly enhance the monitoring of dams, bridges and pipelines.
• The lower repeat cycle of the satellites, particularly in constellations, will make it possible to achieve almost “real-time monitoring” - a particular benefit to civil protection applications.
• The shorter wavelengths will allow better accuracies on both average displacement rates and displacement time series for PSInSAR™ applications.
• The polarimetric capabilities will allow a better characterization of radar targets in terms of the scattering mechanism (single bounce, diehedral, trihedral), while the improved spatial resolution will strongly impact on the positioning (geocoding) accuracy of the measurement points (<1m in all directions).

While the properties of these “second generation” SAR sensors should open new scenarios in SAR applications, success will strongly depend on data acquisition policies, specifically, the creation of historical archives of radar data. It is extremely important to maintain regular acquisitions of SAR.

A new stream of research within the SAR.net project by the National Department of Civil Protection (DPCN) in Italy

Since 2005 TRE has been providing interferometric data to the Italian DPCN within the framework of the “SAR.net” project, which was conceived to transfer the results of scientific research to effective monitoring of the hydrogeological risk in Italy. The analysis of different sites, using the PSInSAR™ technique, has also stimulated the improvement of our processing chain.

A new opportunity
As it is well known, the precise measurement of PS displacement velocity is affected by the presence of the Earth’s atmosphere. In a vacuum, the speed of the electromagnetic signal remains constant; however, the Earth’s atmosphere alters the speed, introducing errors into the phase values. So the PS processing chain is partly concerned with removing this error.

However, might it be possible to assess such a contribution before processing starts? If the answer is ‘yes’, we will be able to determine, more precisely, the contribution to signal phase from the Earth’s atmosphere.
Some “a priori” information exists and consists of a large amount of data collected by weather stations in Italy and already processed by the DPCN.
The first test, using weather and satellite (RADARSAT) data over Rome for about 1,500 km2, has been completed. Results are very promising and TRE will continue its research in the coming months. The integration of weather data with PSInSAR™ and then with GPS is expected to improve the analysis of ground displacements.

Improving our CED (data center) hardware capacity

TRE is pleased to announce that we have expanded our computing (by increasing the number of nodes belonging to the PS processing cluster) and storage capacity, giving us the benefit of a much improved back up facility.
We are now able to process thousands of square kilometres in a short period of time which will benefit many of our cutomer sectors.