In January 2010, TRE will be launching SqueeSAR™: a huge advancement for satellite data analysis and a breakthrough in Earth observation capabilities. It is the result of ten years of intense research and development, here at TRE.

This issue introduces the new features of SqueeSAR™ and its unprecedented ability to analyse ground displacements over areas with poor PS density (e.g. areas with low reflectivity). This new technology has strong advantages for monitoring ground movements over areas where PSInSAR™ analyses couldn’t provide satisfactory results.

In this issue:

Beyond PSInSAR™: ten years of innovation and research
Permanent Scatterers (PS) and Distributed Scatterers (DS):
more information and better quality
Examples of SqueeSAR™ applications
Frequently Asked Questions

Beyond PSInSAR™: ten years of innovation and research

Our history originates from the Department of Electronics and Information at the Politecnico di Milano University in the nineties. There, a pioneering group headed by Professor Rocca commenced their research into Synthetic Aperture Radar (SAR) Interferometry to generate surface deformation maps.

In 1999, the group produced and patented the PSInSAR™ algorithm, which is a significant evolution of conventional SAR Interferometry. The innovation consisted of the use of multi-image datasets, which enable the identification of stable reflectors, referred to as permanent scatterers, or PS. They are points on the ground that consistently return stable signals to the satellite sensor, allowing ground displacement velocities to be measured with millimetre accuracy.
After PSInSAR™, similar algorithms have been developed in the following years with the same aim of singling out point-wise radar targets on the ground, that can be measured over time.

In 2000 POLIMI founded TRE as its first spin-off company to receive exclusive rights for the worldwide use of the patent.  Since its outset, TRE’s main thrust has been maintaining its lead position both in the scientific community and the market by fostering research into developing advanced algorithms, proprietary software and hardware architecture to enhance SAR data analysis.

Now, a  decade after PSInSAR™, TRE is launching SqueeSAR™. It will provide a unique innovation in InSAR analysis that our competitors will find difficult to replicate. This will also allow us to expand our competitive advantage, confirming us as the ’icebreaker’ of SAR interferometry.

Permanent Scatterers (PS) and Distributed Scatterers (DS):
more information and better quality

Distributed Scatterers (DS) represent the core of the new patent (SqueeSAR™) enabling a real advancement in the monitoring of extra-urban areas, where our point of measurement – PS – density could be low.

PS are point-wise targets, which reflect back to the sensor a strong, stable signal. PS typically correspond to man-made structures such as buildings, bridges, dams, water-pipelines, antennae, as well as stable natural reflectors (e.g. exposed rocks). However, we noticed that distributed scatterers also exist and that they too can be used for monitoring ground displacement. Distributed Scatterers or DS consist of an extensive area where the back-scattered energy is less strong in some way, but statistically homogeneous within the area. Hence using our SqueeSAR™ algorithm, it is also possible to ’process’ this energy and detect the movement of areas namely dominated by DS, with the same accuracy as analysis with PS. DS displacement time series are indeed less noisy too. DS typically correspond to debris flows, non-cultivated lands and desert areas.

It is also important to highlight that the PSInSAR™ processing chain is maintained and used within the SqueeSAR™ algorithm: the result is an enhancement of our information output capacity, meaning PS plus DS, to gain an enhanced insight into ground deformation and associated surface movements.

In summary:

PSInSAR™

SqueeSAR™

Ground points identified: PS High density of ground measurement points identified in urban areas (PS) Satisfactory density of ground measurement points identified in non-urban areas (PS) Time series provided for each ground point (PS) Millimetre accuracy on ground displacement values

Ground points identified: PS and DS High density of ground measurement points identified in urban areas (PS) High density of ground measurement points identified in non-urban areas (PS and DS) Time series provided for each ground point (PS and DS) Millimetre accuracy on ground displacement values Time series standard deviation reduces – i.e. coherence increases and noise decreases Increased confidence on ground behaviour due to increased coverage of points – especially significant for landslides, outcrops and generic areas with low reflectivity

Examples of SqueeSAR™ application

The images, below, illustrate the results of applying PSInSAR™ and SqueeSAR™ to the same area of interest and using the same radar imagery. 

Figure 1: The area in question is located in Southern Saskatchewan, Canada. The small concentration of PS in the left image is the Town of Weyburn, which also appears in the right hand image. All the additional mesurements, in the right hand image, relate to areas of pasture or natural grassland where signal to noise ratios would typically be low.

Figure 2: Another comparison between displacement velocity maps over a mountain area: PSInSAR™  results (left) and SqueeSAR™ results (right), using the same RADARSAT dataset.

Figure 3: Comparison between time series, PSInSAR™ processing (left) and SqueeSAR™ processing (right).

top

Frequently Asked Questions

What's the main difference between PSInSAR™ and SqueeSAR™?

With SqueeSAR™ we are able to monitor the behaviour of two different sets of targets on the ground (PS and DS) making it possible to improve the size of areas covered by our analysis, while PSInSAR™ provide only PS. However, there are some situations that remain undetectable, i.e. forest or densely vegetated areas.

Am I at risk of being swamped with data from SqueeSAR™ analysis and, if so, can I request to have the data processed using PSInSAR™?

There is no doubt that SqueeSAR™ will generate more information than PSInSAR™. However, that increase is directed at generating more information in areas where, typically, PS density is low. Within urban areas, we do expect an increase in radar target density but there are likely to be small because the natural reflectors in such areas are already generating higher PS densities.

TRE’s policy is to phase out PSInSAR™, per se, because it has become redundant with the introduction of SqueeSAR™.

Why do some time series look significantly less noisy?

One of the features of the SqueeSAR™ algorithm is that is able to smooth the data scatter, usually caused by noise in the pixel, that is often visible in time series generated by PSInSAR™.
A DS time series is obtained after averaging all signal contributions belonging to the DS extension itself. Since a DS is spread over many contiguous resolution cells - which are statistically homogeneous - it is  possible to significantly reduce the associated noise by simply averaging those cells belonging to the same DS. As a result, a significantly less noisy time series is expected.

What are the future perspectives of SqueeSAR™ use within our markets?

We asked our sales manager which opportunities they are seeing for S. in their reference markets.

Starting with Stefano Cespa, TRE sales manager, he quoted:
“we have seen a clear increase in the number of ways in which our clients are benefitting from  SqueeSAR™. For example: Government and Civil protection authorities in Italy, that are only considering interferometry as an effective tool for landslide analysis and monitoring, will appreciate SqueeSAR™’s  increased ground point coverage in Alpine and mountainous areas. SqueeSAR™ is undoubtedly best in areas where vegetation begins to thin out and where open outcrops emerge. Here the new algorithm allows the identification of sliding areas (previously undetected) and provides a clearer identification of landslide boundaries to update inventory maps, leading to better risk management and hazard mitigation.

The Oil & Gas industry will take advantage of the increased capability in monitoring the extra-urban areas in which they typically operate. A higher density of ground measurement points means that reservoir  deformation patterns, seismic faults positions, and their possible reactivation, will be able to be detected more precisely whatever type of operation is occurring over the field.

Transportation, utilities and construction companies now can count on an enhanced tool for checking infrastructure stability. If a deeper comprehension of regional trends can help the designing phase ante-opera, more precise estimations of time series, as well as the capability to follow complex displacements, will positively impact on monitoring activities during construction and operations  over years”.

Then  Brian Young, TRE Canada CEO, added with regard to the American market:
“First of all, North America was slow to recognize the potential of InSAR technology and there wasn’t the kind of driver like the European Space Agency to push the technology forward, as was the case in Europe.  By default, if you like, the Europeans had a head start on North America. So, North America is still in ‘discovery’ mode, more so than most European countries.

Another consideration is that the population densities, in terms of persons per square kilometer, of Canada and the USA are generally a lot lower than in many European countries. This typically means that there are wide open spaces between population centres in North America, in which land use is either wilderness or agriculture. PSInSAR™ was generally unable to produce meaningful data in such areas and so the opportunities for its use were limited to areas where there was some evidence of habitation. SqueeSAR™, on the other hand, has demonstrated how it can extract usable information on ground motion in areas that have low-lying vegetation cover such as natural grasslands and pasture.  This land use covers large tracts of terrain in North America. 
Now, as you know, North America is rich with sub-surface natural resources. Metal ores and minerals abound and, in many cases, oil & gas and mining operations are located in rural areas. Further, the push to find CCS sites focuses on rural environments. The activities of these facilities invariably impact the environment and both Canada and the USA have reporting requirements that mandate development companies to report, to government regulatory bodies, the physical and polluting impacts on the environment arising from their operations.
Thus, in all areas where natural vegetation is low-lying and generally undisturbed, SqueeSAR™ is able to shed new light on ground displacement. So, as awareness of SqueeSAR™ grows, I believe that new windows of opportunity for InSAR will arise and that the technology, as a whole, will gain increased credibility”.

top