Development of a Cut-Slope Stability Assessment System for Peninsular Malaysia
The purpose of this research is to evaluate the accuracy of four existing slope assessment systems (SAS) in Malaysia in predicting landslides on granitic and sediment/metasediment formation slopes. The four existing SAS in Malaysia are namely Slope Management System (SMS), Slope Priority Ranking...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2006
|
| Subjects: | |
| Online Access: | http://psasir.upm.edu.my/id/eprint/219/1/549049_FK_2006_5.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The purpose of this research is to evaluate the accuracy of four existing slope assessment
systems (SAS) in Malaysia in predicting landslides on granitic and
sediment/metasediment formation slopes. The four existing SAS in Malaysia are namely
Slope Management System (SMS), Slope Priority Ranking System (SPRS), Slope
Information and Management System (SIMS), and Slope Management and Risk Tracking
System (SMART)
Assessment on 139 slopes underlain by granitic formation from the Gunung Raya Road,
the East-West Highway and the Kuala Kubu Baru – Gap Road showed that none of the
existing SAS is satisfactory for predicting landslide. The most accurate prediction was
made by SMART System with only 61% accuracy. For the assessment of 47 slopes
underlain by sediment/metasediment formation from the Gunung Raya Road and the
East-West Highway, the results showed that the accuracy produced by the SMART System was 90%, which was considered as very good prediction. None of the other three
SAS gave satisfactory prediction.
Based on the accuracy evaluation above, two new SAS models were developed for the
slopes in granitic formation. Using the slope database (139 cut-slopes) from the Gunung
Raya Road, the East-West Highway and the Kuala Kubu Baru – Gap Road, twenty five
slope parameters was analysed for development of the new SAS. Development of Model
1 using stepwise discriminate analysis found that ten slope parameters, namely; slope
angle, feature area, distance to ridge, slope shape, percentage of feature uncovered,
presence of rock exposure, rock condition profile, presence of bench drain, horizontal
drain and sign of erosion were significant in predicting landslides occurrences. However,
development of Model 2 using stepwise linear regression analysis found that only nine of
the parameters (same parameters as Model 1 except without rock condition profile) were
significant. The overall correct classification for Model 1 and Model 2 were 77% and
73% respectively.
In order to validate the accuracy of these two newly developed SAS, slope assessment
was carried out on two sites which were different from the ones used in the development
of the new SAS models. The assessment on 36 slopes underlain by granitic formation
from the Kuala Lumpur – Bentung Old Road and the Tapah – Cameron Highland Road,
found that the accuracy in predicting landslides by Model 1 and Model 2 is 88% and 84%
respectively. Hence the degree of accuracy by the 2 newly developed models is within
the accuracy produced by other previous researchers. |
|---|