Keywords Landslide Radar image Subordinate shear plans Micro-geography
Abstract
Hong Kong Radar Image provides us lots of detail information of subordinate
shear planes. These low-grade shear planes are very important factors of landslide. By image
procession and field investigation, we find that subordinate shear planes often make rock lose
bulking property, displays unbalance feature, and expedite the development of landslide. In
processed radar image, subordinate shear planes which can be traced are R planes, P planes and
R’ shear planes. These subordinate shear planes are close related to the evolution of micro-geography,
and also to the development of landslide. Affected by the regional N45E strike-slip fault, the
subordinate N60E dextral shear planes formed at the direction of P shear plan, the N30E dextral
shear plans formed at the direction of R shear plan, and the NW faults formed at the direction of
R’ direction. The boulder falls landslide can only be formed at the extension fault direction of R
shear plan, and the alluvium gravity landslide formed at N60E shear direction.
1. Geology setting
1.1 The rock exposed
The working area is around Jubilee Reservoir. The main exposed rock around Jubilee Reservoir
is fine-grain granite with weak pink color, granodiorite, and pyroclastic rock with some lava.
Rocks are strongly weathered along fault zone, especially at the area of tensely distributed joints
or faults.
1.2 The structure frame
The faults are well developed nearby the Jubilee Reservoir(JR, Fig.1). Three fault sets can be
traced: NE, NW, and EW. The NE trend faults are usually parallel to the long direction of Tolo
Harbor, which was considered as the offset of Lianhua Shan regional fault zone(Bennet,1984,
C.F.Lee,1998, Fletcher,1997,Y.Z.Ding,1997). The NW trend faults can be divided into two
types: regional faults and subordinate faults. The regional NW faults have large scale and can be
easily recognized on radar image. It usually cut through the whole mountain area and shows a
regular lineation. The subordinate faults are in small scale, usually performed as structural
lamellae and restricted in-between the regional faults. In radar image, both NE and NW fault can
be traced at three scales. The first scale includes the elongation direction of mountain, the
shoreline and the shape of harbors. The second scale includes the local topography, the
distribution of valley and streamlet, and the middle scale lineations. The third scale includes the
joints and the micro-topography, they display as image lamellae and the regular digital zones in
processed image.
1.3 The subordinate shear plans
The subordinate shear plans which can be traced are R plans, P plans and R’ plans. R shear plan
has close relation to micro-topography and to image lamellae. It located at the tense and shear
junction has strongly damaged the bulking property of rocks during it development. R shear plan
usually formed at the side of NE fault with 15 degrees, and has the same sinistral
movement as NE fault(Batlett,1981, Lu Huafu,1998,Shi Huosheng,1996). The shape of Jubilee
Reservoir itself can be represented as a combination of transtensional structures . The recent
landslide was occurred at the pure tense boundary. The deep valley, at the south end of Jubilee
Reservoir, has impressive of structural significance. The steep slopes together with R shear plan
and the tense boundary are the most important factors of boulder falls land slide in this area. R’
shear plan is against the move direction of main shear plan, called anti-movement shear plan,
and exhibits itself as a fault or shear joint. R’ plan exhibits the distribution of sinistral shear set
in main fault zone of dextral movement, and exhibits the distribution of dextral shear set in main
fault zone of sinistral movement. P shear plan located at the direction of N60E, where is easy to
form the alluvium gravity land slide.
2. Subordinate shear plans and Landslide
2.1 Image procession
The Radar image used here was made by Chinese Academy of Science, which parameters are in
table 1. Synthesis Aperture Radar image can give a strong stereoscopic view and very rich
information on texture, which are helpful to show the features of linear structure(Armin
Gruen,1997,Chen Shupeng,1998,Guo Huadong.1991). For such merits, it is widely used in the
study of geological structure. However, SAR image has noticeable speckle noise which need a
filter processing. Based on the analysis and comparing of several noise removing methods, we
choose K-self-adaptive filter which has a better result for such a process. In addition, because
SAR and TM use different way to receive different band range of spectrum of ground features,
they have obvious difference on showing the characteristics of spectrum of objects on the
ground(Pang Xizhe,1996). If these two image sources combined together, they can reveal the
characteristics of objects on the ground and information of subordinate structure to a great
extent. In order to compose the monochromatic SAR image and multi-spectrum color image, we
make ISH transformation on the color composite image of TM3, TM4 and TM5 bands at first.
After that, we use filtered SAR image to substitute for I component. Then make invert-IHS
transformation on that. Continually, make KL transformation on TM 3, TM 4, TM 5 and SAR.
At last, we do color composite on TM4 image, G component that comes from invert-IHS
transformation and the first component from KL transformation to get the final result. Studying
on the final image, we found that the subordinate information has been considerably enhanced
and the characteristics of both radar and multi-spectrum images have been reflected, and that
will provide great convenience for geology interpretation(Fig.1).
Fig.1 Radar Image with processed results (scale bar is 2000 m)
Table 1. The Parameters of Hong Kong Radar Image
| Date of flight |
Sep. 18, 1998 |
Sep. 19, 1998 |
| Site Name |
Hong Kong, China |
Hong kong, China |
| SAR Band |
L |
L |
| Transmit Polarization |
H |
H |
| Receive Polarization |
H |
H |
| Line Spacing (m) |
6.25000000 |
6.25000000 |
| Pixel Spacing (m) |
6.25000000 |
6.25000000 |
| Height of Flight(m) |
8400 |
8400 |
