An uncertainty and sensitivity analysis approach for GIS-based multicriteria landslide susceptibility mapping

Figures & data

Figure 1. Location of the case study area in Iran (left) and within northern Iran (right).

Table 1. Classification of GIS-MCDA based on their combination rules and applied methods for LSM.

	MCDA-Combination rules	Type of MCDA	Examples in the literature for LSM
MADA	Weighted summation/Boolean overlay	WLC	Feizizadeh and Blaschke (2013a), Kritikos and Davies (2011)
	Ideal/reference point	TOPSIS, MOLA
	Analytical hierarchy process	AHP, Fuzzy-AHP	Feizizadeh et al. (2012a, 2014), Pourghasemi et al. (2012b), Gorsevski and Jankowski (2010), Oh and Pradhan (2011), Kritikos and Davies (2011)
	Outranking methods	ELECTRE, PROMETHEE
MODA	Multi-objectives programming algorithms (linear-integer programming)
	Heuristic/search/evolutionary/genetic algorithms
	Goal programming/reference point algorithms
Ordered Weighted Average			Feizizadeh et al. (2012b, 2014), Feizizadeh and Blaschke (2013a), Kritikos and Davies (2011)
Dempster–Shafer Theory, weights-of-evidence			Gorsevski and Jankowski (2005), Park (2011), Feizizadeh et al. (2012b), 2014, Althuwaynee et al. (2012), Pourghasemi et al. (2012b)
Artificial neural networks, Support vector machines			Yilmaz (2010)
Fuzzy set- MCDA and Kalman filter			Gorsevski and Jankowski (2010), Oh and Pradhan (2011)
Logistic regression			Ayalew and Yamagishi (2005)

Notes: MADA, Multi-attribute decision analysis; MODA, multi-objective decision analysis.

This table is adapted based on MCDA classification in the references of Malczewski (2006).

Figure 2. Methodology flowchart of the research.

Table 2. The mean consistency index of randomly generated matrices (Saaty 1977).

n	1	2	3	4	5	6	7
RI	0	0	0.52	0.89	1.12	1.26	1.36
n	8	9	10	11	12	13	14	15
RI	1.41	1.46	1.49	1.52	1.54	1.56	1.58	1.59

Figure 3. AHP decision tree for LSM.

Table 3. Pairwise comparison matrix for data set layers of landslide analysis.

Criterion	1	2	3	4	5	6	7	8	9	Weight
(1) Aspect	1									0.025
(2) Distance to road	1/5	1								0.036
(3) Elevation	1/2	1/3	1							0.020
(4) Distance to stream	1/3	1/3	1/3	1						0.112
(5) Distance to fault	1/3	1/5	1/5	1/3	1					0.124
(6) Slope	7	1/5	9	1/3	1/4	1				0.141
(7) Land use	8	6	1/5	1/5	1/3	1/3	1			0.160
(8) Precipitation	8	6	7	7	4	3	1/5	1		0.172
(9) lithology	9	7	1/3	8	7	4	1/5	8	1	0.210

Note: CR: 0.053.

Table 4. Minimum and maximum withed for PDF.

Factor	Reference weights	Maximum weights
Aspect	0.025	0.2
Distance to road	0.036	0.5
Elevation	0.020	0.55
Distance to stream	0.112	0.6
Distance to fault	0.124	0.7
Slope	0.141	0.75
Land use	0.160	0. 65
Precipitation	0.172	0.6
lithology	0.210	0.95

Figure 4. Scheme of MCSs to GIS-MCDA-based LSM.

Figure 5. Schematic relationships of evidential belief functions (Althuwaynee et al. 2012).

Figure 6. Landslide susceptibility maps (LSM) of ‘conventional’ and ‘novel’ approaches: (a) LSM derived from WLC, (b) LSM based on the AHP method, (c) LSM from the OWA method, (d) LSM derived from MCS-WLC, (e) LSM based on the MCS-AHP method, and (f) LSM from the MCS-OWA method.

Figure 7. Uncertainty assessment (UA) for results of landslide susceptibility maps (LSM) of ‘conventional’ and ‘novel’ approaches: (a) UA results of the belief function for LSM derived from WLC, (b) UA results of the belief function for LSM based on the AHP method, (c) UA results of the belief function for LSM of the OWA method, (d) UA results of the belief function for LSM based on the MCS-WLC method, (e) UA results of the belief function for LSM of the MCS-AHP method, and (f) UA results of the belief function for LSM of the MCS-OWA method.

Table 5. Validation and certainty assessment (CA) of GIS-MCDA and GIS-MCS-MCDA results.

MCDA	Susceptibility category	Validation of MCDA results (%)	Validation of MCS-MCD results 2 (%)	CA of MCDA	CA of MCS-MCDA
WLC	High susceptibility	10.57	17.64	0.64–0.67	0.75–0.77
	Moderate susceptibility	86.33	79.26	0.56–0.63	0.71–0.75
	Low susceptibility	3.1	3.1	0.46–0.55	0.61–0.70
	No susceptibility	0	0	0.15–0.45	0.50–0. 60
	Sum	100	100	–	–
AHP	High susceptibility	21.2	29.41	0.85–0.87	0.86–0.89
	Moderate susceptibility	75.7	67.49	0.65–0.84	0.81–0.85
	Low susceptibility	3.1	3.1	0.46–0.64	0.76–0.80
	No susceptibility	0	0	0.22–0.45	0.68–0.75
	Sum	100	100	–	–
OWA	High susceptibility	20.16	29.26	0.85–0.89	0.86–0.93
	Moderate susceptibility	76.74	67.64	0.65–0.84	0.81–0.85
	Low susceptibility	3.1	3.1	0.46–0.64	0.76–0.80
	No susceptibility	0	0	0.24–0.45	0.71–0.75
	Sum	100	100	–	–

Table 6. Representation overall certainty and validation of results by DST and ROC curve.

Approach MCDA		Plausibility	Belief interval	Belief	AUC
Convectional	WLC	0.09–0.63	0.07–0.36	0.15–0.67	0.42451
	AHP	0.48–0.89	0.14–0.41	0.22–0.87	0.76912
	OWA	0.16–0.46	0.15–0.51	0.24–0.89	0.75645
Novel	MCS-WLC	0.21–0.71	0.23–0.57	0.50–0.77	0.71457
	MCS-AHP	0.52–0.91	0.44–0.63	0.68–0.89	0.87945
	MCS-OWA	0.63–0.92	0.49–0.71	0.71–0.93	0.83457

Figure 8. Results of UA for a selected area for ‘conventional’ and ‘novel’ approaches of landslide susceptibility maps (LSM) of (a) UA results of the belief function for LSM derived based on WLC, (b) UA results of the belief function for LSM based on AHP, (c) UA results of the belief function for LSM based on OWA, (d) UA results of the belief function for LSM based on MCS-WLC, (e) UA results of the belief function for LSM based on MCS-AHP, and (f) UA results of the belief function for LSM based on MCS-OWA.

Figure 9. Landslide areas delineated by OBIA from IRS-ID satellite images (Feizizadeh and Blaschke 2013c).

Figure 10. Comparative results of GIS-MCDA methods: (a) results of DST method for measuring the certainty and improved accuracy of GIS-MCDA by novel approach and (b) results of validation of methods based on ROC curve.

Feizizadeh, B. and Blaschke, T., 2013b. Land suitability analysis for Tabriz County, Iran: a multi-criteria evaluation approach using GIS. Journal of Environmental Planning and Management, 56, 1–23.

 Web of Science ®Google Scholar

Kritikos, T. and Davies, T.R.H., 2011. GIS-based multi-criteria decision analysis for landslide susceptibility mapping at northern Evia, Greece. Zeitschrift der Deutschen Gesellschaft für Geowissenschaften, 162, 421–434.

 Web of Science ®Google Scholar

Feizizadeh, B., Jankowski, P., and Blaschke, T., 2014. A GIS based spatially-explicit sensitivity and uncertainty analysis approach for multi-criteria decision analysis. Computers and Geosciences. doi:10.1016/j.cageo.2013.11.009

 Google Scholar

Pourghasemi, H.R., Pradhan, B., and Gokceoglu, C., 2012a. Application of fuzzy logic and analytical hierarchy process (AHP) to landslide susceptibility mapping at Haraz watershed, Iran. Natural Hazards, 63 (2), 965–996.

 Web of Science ®Google Scholar

Gorsevski, P.V., Jankowski, P., and Gessler, P.E., 2006. An heuristic approach for mapping landslide hazard by integrating fuzzy logic with analytic hierarchy process. Control and Cybernetics, 35, 121–146.

 Web of Science ®Google Scholar

Oh, H.J. and Pradhan, B., 2011. Application of a neuro-fuzzy model to landslide susceptibility mapping for shallow landslides in a tropical hilly area. Computers & Geosciences, 37, 1264–1276.

 Web of Science ®Google Scholar

Gorsevski, P.V. and Jankowski, P., 2010. An optimized solution of multi-criteria evaluation analysis of landslide susceptibility using fuzzy sets and Kalman filter. Computers & Geosciences, 36, 1005–1020.

 Web of Science ®Google Scholar

Park, N.W., 2011. Application of Dempster–Shafer theory of evidence to GIS-based landslide susceptibility analysis. Environmental Earth Science, 62, 367–376.

 Web of Science ®Google Scholar

Althuwaynee, O.F., Pradhan, B., and Lee, S., 2012. Application of an evidential belief function model in landslide susceptibility mapping. Computers & Geosciences, 44, 120–135.

 Web of Science ®Google Scholar

Yilmaz, I., 2010. Comparison of landslide susceptibility mapping methodologies for Koyulhisar, Turkey: conditional probability, logistic regression, artificial neural networks, and support vector machine. Environmental Earth Science, 61, 821–836.

 Web of Science ®Google Scholar

Ayalew, L. and Yamagishi, H., 2005. The application of GIS-based logistic regression for landslide susceptibility mapping in the Kakuda-Yahiko Mountains, Central Japan. Geomorphology, 65, 15–31.

 Web of Science ®Google Scholar

MNR, Ministry of Natural Resources, East Azerbaijan Province, 2010. Landslide event report. Iran: Tabriz.

 Google Scholar

Saaty, T.L., 1977. A scaling method for priorities in hierarchical structures. Journal of Mathematical Psychology, 15, 234–281.

 Web of Science ®Google Scholar

Feizizadeh, B. and Blaschke, T., 2013c. A semi-automated object based image analysis approach for landslide delineation. In: The 2013 European space agency living planet symposium, 9–13 September, Edinburgh.

 Google Scholar