| Space

TerraSAR-X/TanDEM-X Science Meeting 2019: Monitoring very slow landslides from space

Credit:
Credits:

left image: ALOS-2 data (JAXA); centre image: modified Copernicus Sentinel data; right image: Terra-SAR-X data (DLR); prozessing each: by Gamma Remote Sensing and SUPSI

Global change is no more a secret – neither is the annual TerraSAR-X and TanDEM-X Science Meeting at DLR Oberpfaffenhofen! High-level experts from all over the world presented the latest state of research from 21 to 24 October 2019 and point the way for future technologies in satellite-based remote sensing. Data source and mutual starting point are the two radar satellite missions TerraSAR-X and TanDEM-X. They provide indispensable data for environmental and climate research, for a better understanding of our earth. The applications are correspondingly diverse. The main research areas include forest, ocean, urbanization, ice cover, glacier, geohazards, agriculture, archaeology and technology development.

In the DLR-Blog we introduce some of the work presented at this year's science meeting. The short examples provided outline how the data of the German radar satellites support researchers worldwide

Over Switzerland, hazards due to slope instabilities affect about six percent of the territory. Particularly in the Alpine areas, landslides repeatedly lead to disasters. It is therefore of outstanding importance to continuously monitor the rate of movement of landslides for risk assessment and to survey their activity over time. Satellite SAR interferometry (InSAR) is one option for surface deformation monitoring over large areas. Recently, various nationwide maps of land deformation have thus been released in Norway, Germany and the United Kingdom. In Switzerland there is also high interest in creating deformation maps and time series of surface movement using InSAR.

Invisible danger

The very high spatial resolution of TerraSAR-X data allows for observing "invisible" or very slow landslides as shown in the latest research by Tazio Strozzi, Rafael Caduff and Andrea Manconi from Gamma Remote Sensing AG and Christian Ambrosi from the University of Applied Sciences and Arts of Southern Switzerland (SUPSI).

They report: Our work was based on large datastacks on the Loderio landslide of in the Canton of Ticino. The Loderio landslide is an example of a very slow but large and active landslide – with movements rates of a few centimeters per year. We used multi-temporal interferometric approaches to process the radar data. Thereby we obtained meaningful results for satellite-based monitoring of landslides in the Alps.

Sparse urbanization, large vegetated areas, snow cover, shadows and layover, atmospheric stratification and summer turbulences pose major challenges for InSAR processing. In particular, we investigated the potential and limitations of current satellite SAR data with different carrier frequencies (L-, C- and X-band), ground resolutions (around 10, 20 and 2 meters), time intervals (46, 6 and 11 days) and acquisition strategies (global versus on-demand, free versus commercial data) for operational monitoring of the Alps.

Our results show that the L-band has an advantage when it comes to temporal decorrelation over vegetated areas and relatively fast movements. The long-wave frequency range is less affected here than are the C- and X-bands. For this reason, the fastest moving part of the Loderio landslide could only be detected with PALSAR-2 data from the Japanese earth observation satellite ALOS-2. On the other hand, the C- and X-band are more sensitive to the very slow moving sectors of the landslide. Then TerraSAR-X data are preferred: they offer a high-precision spatial resolution, so we obtain a higher density of measurement points and are thus able to study local phenomena.