Science Service System

Summary of Proposal LAN0198

TitleDetecting landsliding after major storm events using TerraSAR-X
Investigator Belliss, Stella - Landcare Research New Zealand Limited, Rural Land Use
Team Members
Samonov, Sergey - GNS Science,
Dr Pairman, David - Landcare Research New Zealand Ltd, Rural Land Use
Dr McNeill, Stephen - Landcare Research New Zealand Ltd, Rural Land Use
Joyce, Karen - GNS Science,
Mr Dymond, John - Landcare Research New Zealand Ltd, Rural Land Use
Mr Todd, Malcolm - Manawatu-Wanganui Regional Council, Regional Planning
Mt Steffert, Andrew - Manawatu-Wanganui Regional Council, Regional Planning
Summary Objective

Although optical space-based data have proved effective for identifying and quantifying medium-sized landslides after major storms, imagery is available only once the storm has passed and skies have cleared. In our experience, this can take from days to weeks, and so our objective is to demonstrate an on-demand capability to map damage using high resolution space-based radar. This would enable us to provide information and services faster, and to detect much smaller landslides that are currently undetected by other medium-resolution space-based sensors.

Landslides and flooding caused by heavy rainfall remove a great deal of vegetation and soil, and have a considerable effect on waterways, agricultural activity, road routes, and farming infrastructure. Collectively, the large number of small landslides have a profound effect on the environment, and they are the most difficult to map. Landcare Research has investigated the use of Radarsat-1 imagery but this is constrained by the spatial resolution of these data. We believe that TerraSAR-X data would overcome this spatial resolution constraint.

Method

We plan to acquire and process two radar stereo pairs to detect landslide scars, captured from both the single polarisation stripmap and high single polarisation spotlight modes of our test site, taken as close together in time as practicable. The imagery will be fused to produce a post-storm DEM, using information from an existing DEM to constrain the processing. Landslides will be detected by calculating the difference between the reference and stereo-fused DEMs, with additional land cover information providing detection assistance.

We will also calculate the likely distributions of acquisition dates for TerraSAR-X radar data and the delivery time for a landslide detection product so our capability will include realistic delivery times to civil emergency response efforts from central and regional government.

Data requirements

A single polarisation stripmap stereo pair, as well as a high resolution single polarisation spotlight product of part of the test site. The images within each stereo pair should be taken as close together in time as practicable to compare with existing results from data analysis after the 2004 storm. If the test site was to be affected by an extrememe weather event after these data acquisitions, then we would request a second pair of images to be taken.

Deliverables
  1. A methodology to rapidly detect deep-seated landslides in storm-affected areas
  2. A series of recommendations to emergency response decision-makers as to the most appropriate data sources and modes for use in a given situation
  3. A presentation at the TerraSAR-X results reporting meeting
  4. Publication of the research findings in an international journal
  5. With DLR’s permission, publication of the project and results on the company websites of the Landcare Research and Manawatu-Wanganui Regional Council.
Funding

Landcare Research plans to obtain funding from the Foundation for Research Science and Technology (FRST). This is the New Zealand Government Department that allocates research funding from a contestable pool. Our research area is due to be re-tendered in early 2006 for research funding which may continue through to 2012, so the timing of this mission fits well with our funding cycles. Manawatu-Wanganui Regional Council will be included as sub-contractors on our bid to FRST. In addition, they are able to undertake fieldwork in support of this research proposal as part of their on-going environment monitoring schedules. We are aware that our participation will depend upon gaining this funding.

Back to list of proposals

© DLR 2004-2016