Classification
Terrestrial phenomena disasters/Earthquakes/Urban structure damage, slope disasters, Geotechnical disasters.

Urban disasters/Urban structure damage. Geotechnical disasters/Crustal movement.
Disaster name
Monitoring of Changes to the Northern Part of Awaji Island Caused by the 1995 Southern Hyogo Prefecture Earthquake
Author of WEB conversion
Iwashita Atsushi

Case Study

No. 26

1. Analysis objective

Sample geographic movement and understand its characteristics from lineament analysis in areas in the vicinity of earthquake faults.

Understand characteristics based on examination of lineament automatic sampling results.


2. Analysis procedure Analysis flow chart

This is a technique to estimate the ground level and immediate underlying features, using lineament based on visual interpretation and automatic sampling obtained from high-resolution satellite images. Compared to traditional methods using lineament, this approach is characterized by the use a many lineament groups, to determine the distribution density or the "ratio of decline in fracture frequency" spatial characteristics, and clarify the fracture characteristics of the target region.

The ground conditions in the vicinity of an earthquake fault can be qualitatively understood.

We performed lineament using automatic sampling and visual interpretation. Lineament sampling is a technique that recognizes the information contained in the images used based upon the knowledge and judgment of analysts who are familiar with the geologic structure of the target region and is performed visually. As a technique to understand the characteristics of geological structures, this method has been used for many years and is frequently used for the analysis of geologic structures in images.

It is well known, however, that methods employing visual observation will change depending upon the analyst and target region, and overcoming this aspect of such methods was a problem. In contrast to traditional lineament sampling, the method used for our research analyzed the information contained in the images used through the use of objective results sampled automatically based on a specific algorithm. The images used were SPOT/HRV obtained after the earthquake. From these we removed the northern part of Awaji island that includes the Nojima fault, applied a high-emphasis filter using a 3x3 local operator, and created an image suitable for lineament sampling. The parameters of the image used were a SPOT/HRV multi-spectral image from bands 1-2-3 obtained on January 1, 1995.

We were able recognize the features seen from this image as locations where the Nojima fault is assumed to lie under the northwest edge of Awaji Island as a clear topographic change.
Moreover, the development of several annular structures can be noted on the east side of the fault, which we perceived to be the result of a comparatively intense topographic change. While there are various opinions about the primary origin of this annular structure, the results of the lineament sampling that captured the topographic changes reflect the unique features of the annular structure.


3.Analysis results

From the distribution of the lineaments we were able to clarify that the Nojima fault running nearly north-south across the coast of northern Awaji Island was active, and that there were topographic changes along the earthquake fault that appeared and its vicinity.

Moreover, we clarified that the correlation between the lineament pattern and the fringe that appeared on the interferometry image that used JERS-1/SAR showed qualitatively excellent results.


4. Results from using the analysis results

The utility of the applied aspects of lineament geological analysis as fracture information obtained from satellite images has long been viewed positively.

Lineament automatic sampling algorithms have also been proposed before; among these the DSDA method is especially handy and comparatively easy to use. For lineament analysis in the past, most were methods for dealing with the characteristics of long, large lineaments through visual interpretation.

As one direction for the future, techniques for statistical handling of numerous lineament grouping that are obtained from the results of automatic lineament sampling are beginning to be tested.

As these methods found greater application, there is a possibility we will be able to discover the parameters that show ground strength by applying lineament analysis to disaster regions.

We look forward to greater quantitative handling of fracture information, which is part of the phenomena in the natural world.


5. Sources

Daiichi M. and others: Automatic Sampling of Lineaments from Satellite MSS Data and its Geologic Evaluation, Mining Geology, Vol. 37, No. 2, pp. 99-108, 1987

Hase H. and Yamaguchi Y.: Analysis of Lineament Frequency Using Various Images - Example of an Application of Radar Images to the Yakushima Region -, Journal of the Japan Society of Photogrammetry and Remote Sensing, Vol. 22, No. 3, pp. 4-13, 1983

Iwashita A . Others: A Quantitative Evaluation of Image Emphasis Results for the Purpose of Lineament Sampling by Comparisons to Active Faults, Journal of the Japan Soceity of Photogrammetry and Remote Sensing, Vol., 33, No. 6, pp. 6-15, 1994