Comparative Study of Geological Hazards Susceptibility Assessment: Constraints from the Information Value+Logistic Regression Model and the CF+Logistic Regression Model

doi:10.19657/j.geoscience.1000-8527.2018.03.18

Abstract

Abstract:

Susceptibility assessment constitutes an important part of geological hazard research. This paper discussed two integrated approaches, namely the Information Value+Logistic Regression model (I+LR) and CF+Logistic Regression model (CF+LR), to evaluate the geological hazard susceptibility of Yanchi County, Ningxia Hui Autonomous Region. 462 samples (231 hazardous samples and 231 non-hazardous samples) were divided into two groups: 75% and 25% of the samples were used for model training and validation, respectively. Nine influencing factors, including slope dip angle and direction, slope height, elevation, strata, distance to rivers and roads, and the normalized difference vegetation index (NDVI) were considered in this evaluation. Based on such information, the information value and C_F of the training samples were calculated, which were then fed into the SPSS for analysis via logistic regression. After collinearity diagnostics and correlation analysis, six factors were incorporated into the model eventually. Regression equation was determined using the obtained constants and coefficients, and the Receiver Operating Characteristic (ROC) curve was used to assess the accuracy of the two models. The results show that: (1)Susceptibility maps reveal four susceptibility classes, i.e., very low, low, moderate and high. The area percentage of the (I+LR) model for the four class accounts for 60.79%, 23.44%, 11.34% and 4.43%, respectively, with that of the (CF+LR) model being 54.49%, 22.89%, 10.30% and 12.32%. This indicates that the areas of the low and moderate classes are basically the same. The high-class area of the (CF+LR) model has increased for 533.6 km² more than the (I+LR) model, but the very low-class area has dropped by as much as 6%. (2) The area under the curve for the successful rates are 0.868 and 0.829, respectively, and asymptotic Sig.^b is lower than 0.05 for the two models. Both integrated approaches can produce reasonable accuracy. (3) The ROC accuracy and the geological hazard development conditions at Yanchi both indicate that the (I+LR) model has higher accuracy over the (CF+LR) model.

Key words: geological hazard, susceptibility, CF model, logistic regression model, Yanchi of Ningxia

CLC Number:

P627
P642.2

ZHANG Xiaodong, LIU Xiangnan, ZHAO Zhipeng, WU Wenzhong, LIU Haiyan, ZHANG Yong, GAO Yuliang. Comparative Study of Geological Hazards Susceptibility Assessment: Constraints from the Information Value+Logistic Regression Model and the CF+Logistic Regression Model[J]. Geoscience, 2018, 32(03): 602-610.

Figures/Tables 9

References 15

[1]	GUZZETTI F, REICHENBACH P, CARDINALI M, et al. Probabilistic landslide hazard assessment at the basin scale[J]. Geomorphology, 2005, 72(1/4) : 272-299. DOI URL
[2]	倪化勇, 王德伟, 陈绪钰, 等. 四川雅江县城地质灾害发育特征与稳定性评价[J]. 现代地质, 2015, 29(2):474-480.
[3]	孟庆华, 孙炜锋, 王涛. 陕西凤县地质灾害易发性评价研究[J]. 工程地质学报, 2011, 19(3):388-396.
[4]	张超, 陈艳, 张宇飞, 等. 基于多元线性回归模型的云南昭通地质灾害易发性评价[J]. 水文地质工程地质, 2016, 43(3):159-163.
[5]	CHAU K T, SZE Y L, FUNG M K, et al. Landslide hazard analysis for Hong Kong using landslide inventory and GIS[J]. Computers & Geosciences, 2004, 30(4):429-443. DOI URL
[6]	田春山, 刘希林, 汪佳. 基于CF和Logistic回归模型的广东省地质灾害易发性评价[J]. 水文地质工程地质, 2016, 43(6):154-161.
[7]	岳溪柳, 黄玫, 徐庆勇, 等. 贵州省喀斯特地区泥石流灾害易发性评价[J]. 地球信息科学学报, 2015, 17(11):1396-1403.
[8]	许冲, 戴福初, 姚鑫, 等. 基于GIS与确定性系数分析方法的汶川地震滑坡易发性评价[J]. 工程地质学报, 2010, 18(1):15-26.
[9]	方苗, 张金龙, 徐瑱. 基于GIS和Logistic回归模型的兰州市滑坡灾害敏感性区划研究[J]. 遥感技术与应用, 2011, 26(6):845-853.
[10]	杜谦, 范文, 李凯, 等. 二元 Logistic 回归和信息量模型在地质灾害分区中的应用[J]. 灾害学, 2017, 32(2):220-226.
[11]	张晓东, 刘湘南, 赵志鹏, 等. 盐池县地质灾害遥感调查及空间分布特征分析[J]. 水文地质工程地质, 2017, 44(1):164-170.
[12]	SHORTLIFFE E H, BUCHANAN B G. A model of inexact reasoning in medicine[J]. Mathematical Biosciences, 1975, 23(3):351-379. DOI URL
[13]	HECKERMAN D. Probabilistic interpretation for mycin’s certainty factors[J]. Machine Intelligence & Pattern Recognition, 2013, 4:9-20.
[14]	LIANG J W, KAZUHIDE S, SHUJI M. Landslide susceptibility analysis with logistic regression model based on FCM sampling strategy[J]. Computers & Geosciences, 2013, 57:81-92. DOI URL
[15]	HAMID R P, MAJID M, BISWAJEET P. Landslide susceptibility mapping using index of entropy and conditional probability models in GIS:Safarood Basin,Iran[J]. Catena, 2012, 97:71-84. DOI URL

评价因子	分级数	指标分级
坡度/(°)	8	0~5,5~10,10~15,15~20,20~25,25~30,30~35,>35
坡向	9	平地,北向,东北,东向,东南,南向,西南,西向,西北
坡高/m	5	0~10,10~20,20~30,30~40,>40
高程/m	6	1 294~1 400,1 400~1 500,1 500~1 600,1 600~1 700,1 700~1 800,1 800~1 950
地层	10	全新统,更新统,新近系,古近系,白垩系,侏罗系,三叠系,二叠系,奥陶系,中元古界
距河流距离/m	6	0~200,200~400,400~600,600~800,800~1 000,>1 000
距道路距离/m	5	0~100,100~200,200~300,300~400,>400
NDVI	5	0~0.1,0.1~0.2,0.2~0.3,0.3~0.4,>0.4

评价因子	分级数	指标分级
坡度/(°)	8	0~5,5~10,10~15,15~20,20~25,25~30,30~35,>35
坡向	9	平地,北向,东北,东向,东南,南向,西南,西向,西北
坡高/m	5	0~10,10~20,20~30,30~40,>40
高程/m	6	1 294~1 400,1 400~1 500,1 500~1 600,1 600~1 700,1 700~1 800,1 800~1 950
地层	10	全新统,更新统,新近系,古近系,白垩系,侏罗系,三叠系,二叠系,奥陶系,中元古界
距河流距离/m	6	0~200,200~400,400~600,600~800,800~1 000,>1 000
距道路距离/m	5	0~100,100~200,200~300,300~400,>400
NDVI	5	0~0.1,0.1~0.2,0.2~0.3,0.3~0.4,>0.4

评价因子	评价因子分级	灾害点/个	类别面积/km²	I	C_F	评价因子	评价因子分级	灾害点/个	类别面积/km²	I	C_F
坡度/(°)	0~5	12	5 011.41	-2.658 6	-0.932 18	地层	全新统	9	2 271.03	-2.154 76	-0.887 59
	5~10	27	596.50	0.280 8	0.253 46		更新统	189	3 706.01	0.400 04	0.341 38
	10~15	13	264.36	0.363 6	0.315 65		新近系	3	57.03	0.431 04	0.362 56
	15~20	31	133.18	1.918 3	0.883 34		古近系	2	348.50	-1.784 48	-0.836 92
	20~25	43	262.37	1.567 5	0.819 44		白垩系	28	331.56	0.904 40	0.616 28
	25~30	63	124.53	2.694 6	0.965 43		侏罗系	0	0.85	0	-1.000 00
	30~35	17	181.26	1.009 3	0.658 02		三叠系	0	19.68	0	-1.000 00
	>35	25	184.01	1.379 9	0.774 88		二叠系	0	11.53	0	-1.000 00
坡向	平地	17	528.11	-0.060 08	-0.060 25		奥陶系	0	4.73	0	-1.000 00
	北向	69	905.96	0.801 12	0.570 68		中元古界	0	6.70	0	-1.000 00
	东北	25	648.00	0.120 99	0.117 99	距河流距离/m	0~200	129	165.12	3.129 14	0.990 09
	东向	12	651.21	-0.617 92	-0.469 59		200~400	36	155.36	1.913 78	0.882 65
	东南	27	836.37	-0.057 23	-0.057 47		400~600	11	156.31	0.722 06	0.532 45
	南向	40	983.37	0.173 90	0.165 27		600~800	12	148.45	0.860 67	0.597 55
	西南	15	818.46	-0.623 37	-0.472 52		800~1 000	5	151.03	-0.032 03	-0.032 60
	西向	10	889.92	-1.112 54	-0.678 91		>1 000	38	5 981.34	-1.682 81	-0.819 35
	西北	16	496.22	-0.058 42	-0.058 64	距道路距离/m	0~100	8	147.50	0.461 62	0.382 83
坡高/m	0~10	27	5430.71	-1.927 98	-0.858 83		100~200	5	117.78	0.216 63	0.201 66
	10~20	31	655.24	0.324 99	0.287 29		200~300	10	129.05	0.818 39	0.578 64
	20~30	37	306.56	1.261 51	0.742 14		300~400	9	109.17	0.880 33	0.606 07
	30~40	41	181.10	1.890 53	0.879 06		>400	199	6 254.12	-0.071 68	-0.071 45
	>40	95	184.01	2.714 90	0.966 84	NDVI	<0.1	9	813.75	-1.128 42	-0.684 02
高程/m	1 294~1 400	14	1418.34	-1.242 18	-0.718 34		0.1~0.2	91	2 919.37	-0.092 26	-0.090 96
	1 400~1 500	66	2 599.04	-0.297 23	-0.263 83		0.2~0.3	72	1 885.13	0.110 92	0.108 71
	1 500~1 600	70	1 850.39	0.101 35	0.099 80		0.3~0.4	40	687.70	0.531 54	0.426 89
	1 600~1 700	64	577.99	1.175 33	0.715 75		>0.4	19	451.67	0.207 50	0.194 02
	1 700~1 800	14	210.37	0.666 20	0.503 56
	1 800~1 950	3	101.49	-0.145 34	-0.139 39

评价因子	评价因子分级	灾害点/个	类别面积/km²	I	C_F	评价因子	评价因子分级	灾害点/个	类别面积/km²	I	C_F
坡度/(°)	0~5	12	5 011.41	-2.658 6	-0.932 18	地层	全新统	9	2 271.03	-2.154 76	-0.887 59
	5~10	27	596.50	0.280 8	0.253 46		更新统	189	3 706.01	0.400 04	0.341 38
	10~15	13	264.36	0.363 6	0.315 65		新近系	3	57.03	0.431 04	0.362 56
	15~20	31	133.18	1.918 3	0.883 34		古近系	2	348.50	-1.784 48	-0.836 92
	20~25	43	262.37	1.567 5	0.819 44		白垩系	28	331.56	0.904 40	0.616 28
	25~30	63	124.53	2.694 6	0.965 43		侏罗系	0	0.85	0	-1.000 00
	30~35	17	181.26	1.009 3	0.658 02		三叠系	0	19.68	0	-1.000 00
	>35	25	184.01	1.379 9	0.774 88		二叠系	0	11.53	0	-1.000 00
坡向	平地	17	528.11	-0.060 08	-0.060 25		奥陶系	0	4.73	0	-1.000 00
	北向	69	905.96	0.801 12	0.570 68		中元古界	0	6.70	0	-1.000 00
	东北	25	648.00	0.120 99	0.117 99	距河流距离/m	0~200	129	165.12	3.129 14	0.990 09
	东向	12	651.21	-0.617 92	-0.469 59		200~400	36	155.36	1.913 78	0.882 65
	东南	27	836.37	-0.057 23	-0.057 47		400~600	11	156.31	0.722 06	0.532 45
	南向	40	983.37	0.173 90	0.165 27		600~800	12	148.45	0.860 67	0.597 55
	西南	15	818.46	-0.623 37	-0.472 52		800~1 000	5	151.03	-0.032 03	-0.032 60
	西向	10	889.92	-1.112 54	-0.678 91		>1 000	38	5 981.34	-1.682 81	-0.819 35
	西北	16	496.22	-0.058 42	-0.058 64	距道路距离/m	0~100	8	147.50	0.461 62	0.382 83
坡高/m	0~10	27	5430.71	-1.927 98	-0.858 83		100~200	5	117.78	0.216 63	0.201 66
	10~20	31	655.24	0.324 99	0.287 29		200~300	10	129.05	0.818 39	0.578 64
	20~30	37	306.56	1.261 51	0.742 14		300~400	9	109.17	0.880 33	0.606 07
	30~40	41	181.10	1.890 53	0.879 06		>400	199	6 254.12	-0.071 68	-0.071 45
	>40	95	184.01	2.714 90	0.966 84	NDVI	<0.1	9	813.75	-1.128 42	-0.684 02
高程/m	1 294~1 400	14	1418.34	-1.242 18	-0.718 34		0.1~0.2	91	2 919.37	-0.092 26	-0.090 96
	1 400~1 500	66	2 599.04	-0.297 23	-0.263 83		0.2~0.3	72	1 885.13	0.110 92	0.108 71
	1 500~1 600	70	1 850.39	0.101 35	0.099 80		0.3~0.4	40	687.70	0.531 54	0.426 89
	1 600~1 700	64	577.99	1.175 33	0.715 75		>0.4	19	451.67	0.207 50	0.194 02
	1 700~1 800	14	210.37	0.666 20	0.503 56
	1 800~1 950	3	101.49	-0.145 34	-0.139 39

评价因子	VIF		评价因子	VIF		评价因子	VIF		评价因子	VIF
评价因子	I模型	CF模型	评价因子	I模型	CF模型	评价因子	I模型	CF模型	评价因子	I模型	CF模型
坡度	5.689	4.058	坡高	7.351	5.786	地层	1.144	1.200	距道路距离	1.062	1.096
坡向	1.132	1.053	高程	1.242	1.614	距河流距离	1.145	1.177	NDVI	1.080	1.020