Introduction - Semi-quantitative Approaches for Landslide Assessment and Prediction - page 46

physical models studied after Carrara et al. ( 1995 ), Soeters and Van Westen
( 1996 ), Guzzetti et al. ( 1999a , b , c , 2003 ).

Essentially, the following two broad options are available at the start of the
mapping program.

1. Direct mapping of landslide hazards, based on geological, geomrophological,
and geotechnical investigation, and mapping by a single multidisciplinary team.
2. State-of-nature (factor) mapping and integration of factor maps into landslide
hazards map.

British geomorphologists have cited some of the best examples of landslide hazard
maps by using the direct hazard mapping approach. The second option is, however,
normally preferred because Factor Maps are, in themselves, a great value to the users
of the project output and provide the best assurance of independent evaluation, easy
revision, and constant upgradation and revalidation (Bhandari 1994 ). Guzzetti et al.
( 2003 ) evaluated the requirements of indirect/state-of-nature method as: (i) the rec-
ognition and mapping of landslides over a target region which is obtained by pre-
paring a landslide inventory map; (ii) the identi cation and mapping of the physical

factors which are directly or indirectly correlated with slope instability; (iii) an
estimate of the relative contribution of the instability factors in initiating slope
instability; (iv) the classi cation of land surface into domains of different levels of

susceptibility; and (v) the assessment of the model performance.
Crrara et al. (1992, 1995 ), Van Westen (1993), Hutchinson (1995), Soeters and
Van Westen ( 1996 ), Van Westen et al. (1997a), Guzzetti et al. ( 1999a , b , c ) and
Committee on the National Landslide Hazards Mitigation Strategy (2004), India
suggested the most common approaches in the literature review of landslide
susceptibility mapping procedure and which can be grouped into ve major cate-

gories, namely: (i) direct geomorphological mapping; (ii) analysis of landslide
inventories; (iii) heuristic or index based methods; (iv) statistical methods,
including neural networks, fuzzy logic and expert systems; and (v) process based
conceptual models.
Bureau of Indian Standards (BIS-1998), proposed a guidelines for landslide
hazard zonation map on 1:25,000 or 50,000 scale. Bhandari ( 1987 ) proposed
landslide hazard zonation on the basis of the hazard rating values relating to the
responsible factors for landslide. The landslide zonation map at regional scale has
been attempted in different parts of the world for past two/three decades (Nilsen and
Brabb 1973; De Graf 1978; Varnes 1981, 1985). A landslide hazard map (macro)
on 1:25,000 scales has been prepared following the landslide hazard evaluation
factor (LHEF) rating scheme proposed by the Indian Bureau of Standards (1998)
for the area between Singtam and Ranipool (after S.R. Basu and B. Bera). Data for
LHEF have been computed from interpretation of 1:50,000 Survey of India topo-
graphical maps, 1:25,000 prints of panchromatic aerial photographs (stereoscopic
coverage), 1:50,000 geo-coded LISS III Satellite data, existing geological maps and
extensive field work. Bhandari ( 1987 ) proposed landslide hazard zonation on the
basis of the hazard rating values relating to the responsible factors for landslide.