1. Trang chủ
  2. » Ngoại Ngữ

A comparative assessment of prediction capabilities of modified analytical hierarchy process (MAHP) and Mamdani fuzzy logic models using NetcadGIS for forest fire susceptibility mapping

26 498 0

Đang tải... (xem toàn văn)

Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống

THÔNG TIN TÀI LIỆU

Thông tin cơ bản

Định dạng
Số trang 26
Dung lượng 1,85 MB
File đính kèm 635937145943402483.rar (1 MB)

Các công cụ chuyển đổi và chỉnh sửa cho tài liệu này

Nội dung

The main purpose of this study is to assess forest fire susceptibility maps (FFSMs) and their performances comparison using modified analytical hierarchy process (MAHP) and Mamdani fuzzy logic (MFL) models in a geographic information system (GIS) environment. This study was carried out in the Minudasht Forests, Golestan Province, Iran, and was conducted in three main stages such as spatial data construction, forest fire modelling using MAHP and MFL, and validation of constructed models using receiver operating characteristic (ROC) curve. At first, seven conditioning factors, such as altitude, slope aspect, slope angle, annual temperature, wind effect, land use, and normalized different vegetation index, were extracted from the spatial database. In the next step, FFSMs were prepared using MAHP and MFL modules in a NetcadGIS Architect environment. Finally, the ROC curves and area under the curves (AUCs) were estimated for validation purposes. The results showed that the AUCs for MFL and MAHP are 88.20% and 77.72%, respectively. The results obtained in this study also showed that the MFL model performed better than the MAHP model. These FFSMs can be applied for land use planning, management, and prevention of future fire hazards.

Trang 1

Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=tgnh20

Geomatics, Natural Hazards and Risk

ISSN: 1947-5705 (Print) 1947-5713 (Online) Journal homepage: http://www.tandfonline.com/loi/tgnh20

A comparative assessment of prediction capabilities of modified analytical hierarchy process (M-AHP) and Mamdani fuzzy logic models using Netcad-GIS for forest fire susceptibility

mapping Hamid reza Pourghasemi, Masood Beheshtirad & Biswajeet Pradhan

To cite this article: Hamid reza Pourghasemi, Masood Beheshtirad & Biswajeet Pradhan

(2016) A comparative assessment of prediction capabilities of modified analyticalhierarchy process (M-AHP) and Mamdani fuzzy logic models using Netcad-GIS for forestfire susceptibility mapping, Geomatics, Natural Hazards and Risk, 7:2, 861-885, DOI:

10.1080/19475705.2014.984247

To link to this article: http://dx.doi.org/10.1080/19475705.2014.984247

© 2014 Taylor & Francis Published online: 01 Dec 2014

Submit your article to this journal Article views: 142

Trang 2

A comparative assessment of prediction capabilities of modified analytical hierarchy process (M-AHP) and Mamdani fuzzy logic models

using Netcad-GIS for forest fire susceptibility mapping

HAMID REZA POURGHASEMIy*, MASOOD BEHESHTIRADzand

BISWAJEET PRADHAN x

yDepartment of Natural Resources and Environment, College of Agriculture, Shiraz

University, Shiraz, IranzDepartment of Natural Resources, Sirjan Branch, Islamic Azad University, Sirjan, IranxDepartment of Civil Engineering, Faculty of Engineering, Geospatial InformationScience Research Center (GISRC), University Putra Malaysia, Serdang 43400, Malaysia

(Received 31 May 2014; accepted 1 November 2014)

The main purpose of this study is to assess forest fire susceptibility maps (FFSMs)and their performances comparison using modified analytical hierarchy process(M-AHP) and Mamdani fuzzy logic (MFL) models in a geographic informationsystem (GIS) environment This study was carried out in the Minudasht Forests,Golestan Province, Iran, and was conducted in three main stages such as spatialdata construction, forest fire modelling using M-AHP and MFL, and validation

of constructed models using receiver operating characteristic (ROC) curve Atfirst, seven conditioning factors, such as altitude, slope aspect, slope angle, annualtemperature, wind effect, land use, and normalized different vegetation index,were extracted from the spatial database In the next step, FFSMs were preparedusing M-AHP and MFL modules in a Netcad-GIS Architect environment.Finally, the ROC curves and area under the curves (AUCs) were estimated forvalidation purposes The results showed that the AUCs for MFL and M-AHP are88.20% and 77.72%, respectively The results obtained in this study also showedthat the MFL model performed better than the M-AHP model These FFSMscan be applied for land use planning, management, and prevention of future firehazards

1 Introduction

Forests are major natural resources which play a crucial role in maintaining mental balance The health of forest in a given area is a true indicator of the ecologi-cal condition prevailing in that area (Saklani 2008) In general, fire is a naturalcomponent of many forest ecosystems and cannot be avoided (Dimopoulou &Giannikos2001) Forest fires cause major damages to environment, human healthand property, and endanger life (Rawat2003) Six million square kilometre of forestshas been lost around the world in less than 200 years mainly due to forest fire (Dimo-poulou & Giannikos2002) In Iran, forest fire is one of the most natural occurringhazards According to the ECE/FAO database (Economic Commission for Europe/Food and Agriculture Organization) on forest fires in 19821995, the number of

environ-*Corresponding author Email:hamidreza.pourghasemi@yahoo.com

Ó 2014 Taylor & FrancisVol 7, No 2, 861885, http://dx.doi.org/10.1080/19475705.2014.984247

Trang 3

forest fires per year is 130 and the burnt average area and its maximum is 54 km2and

330 km2, respectively (Allard2001) During the period 19911997, nearly 3063 fireshave been reported, of which 13,700 ha was burnt In the year 1998, there were 998fires reported and the burnt area was estimated at 206,713 ha covering mostly shrubs.Losses were estimated at more than 5.6 million Rials (almost USD 3200 in the year

of 1998), including 8761 tons of cattle feed lost (Allard 2001) In a recent paper,Janbaz Ghobadi et al (2012) reported that, in the last decade, 9086 ha of the forestshave been affected by forest fire Also, in Iran, it is difficult to control forest fire natu-rally; however, it is possible to map different hazard levels for minimizing fire haz-ards and avoid potential damage

In the literature, several different methods and techniques for forest fire bility and risk mapping have been proposed and tested Many studies have evaluatedforest fire using geographic information system (GIS) and remote sensing (RS) tech-nologies (Chuvieco & Congalton1989; Prosper-Laget et al.1995; Castro & Chuvieco

suscepti-1998; Jaiswal et al.2002; Erten et al.2004; Wulder & Franklin2006; Pradhan et al

2007; Razali2007; Saklani2008; Chuvieco et al.2010; Pradhan & Assilzadeh2010;Adab et al.2013; Teodoro & Duarte2013) Several studies have applied probabilis-tic-based models such as fuel moisture content (FMC), fire area simulator (FAR-SITE), and Maxent models (Chuvieco et al.2004; Garcıa et al.2008; Krasnow et al

2009; Renard et al.2012)

In the past decade, some methods, such as artificial neural networks (ANNs)(Betanzos et al.2002; Maeda et al.2009; Bisquert et al.2012; Safi & Bouroumi2013),fuzzy logic (Nadeau et al.2005; Carvalho et al.2006; Agarwal et al.2013), and adap-tive neuro-fuzzy inference system (ANFIS) (Angayarkkani & Radhakrishnan 2011),have been proposed Recently, new forest risk assessment methods, such as supportvector machine (SVM) (Cortez & Morais2007; Koetz et al.2008; Zhao et al.2011),decision tree methods (Stojanova et al.2006), and random forest (Cortez & Morais

2007; Pierce et al.2012; Leuenberger et al.2013), were employed and their ces were assessed

performan-The aim of the current research is to assess forest fire susceptibility maps (FFSMs)using modified analytical hierarchy process (M-AHP) and Mamdani fuzzy logic(MFL) models developed in Netcad GIS 6 The assessment was performed in theMinudasht forests situated in Golestan Province, Iran The main difference betweenthis research and the approaches described in the aforementioned publications isthat an M-AHP model is applied and the result is compared with MFL model in thestudy area Also, expert opinions are used in the mentioned models for defining therules and conditioning factor scores in MFL and M-AHP models, respectively Thiscontribution provides originality to this study

bound-of Minudasht ranges between temperate and semi-humid types The mean annual

Trang 4

precipitation within the study area varies from 138 to 335 mm (Shadman Roodposhti

et al.2014) Based on Iranian Meteorological Organization, maximum and minimum

of temperature were reported asC40 and ¡5C, respectively Agriculture is the main

economic activity of the region In addition, part of the Golestan National Park islocated within the county and it is known as tropical dry forests

3 Methodology

The overall methodology flow chart of the study is shown infigure 2 The flowchartconsists of three phases: (1) data integration and analysis, (2) forest fire susceptibilitymodelling using M-AHP and MFL approaches, and (3) validation of the constructedmodels using receiver operating characteristic (ROC) curve

3.1 Data integration and analysis

In general, data collection and construction of a database of effective factors in anystudy area are the most important parts of the process (Ercanoglu & Gokceoglu

2002) At first, fire occurrences and locations were collected from MODIS ate-Resolution Imaging Spectro Radiometer) satellite images (collected in year

(Moder-Figure 1 Forest fire location (black and pink dots) map in the study area Modified fromPourtaghi et al (2014) To view this figure in colour, please see the online version of the journal

Trang 5

2010), extensive field surveys, and national reports Forest fires are related to year

2010, at 1:25,000 scale (table 1) Out of 151 forest fire locations, 70% were used in themodel training and the remaining 30% were used for validation (Pourghasemi et al

2012a; Zare et al.2012; Pourghasemi et al.2014; Regmi et al.2014) For FFSM inthe study area, seven effective factors were considered These factors include altitude,slope aspect, slope angle, annual temperature, wind effect, land use, and normalizeddifferent vegetation index (NDVI) The spatial database for the study area is shown

intable 1

Figure 2 Flow chart of used methodology in the study area

Table 1 Data used for forest fire susceptibility mapping (FFSM)

Forest fire locations map Point Satellite image, aerial photos,

and extensive field surveys

1:25,000Topographic map Line and point National Cartographic

vegetation index (NDVI)

Grid National Geographic

Organization (NGO)

30m£ 30mMeteorological data Excel data Iranian Meteorological

Organization (IRIMO)



Trang 6

One of the important factors in any fire hazard rating system is topography data.

In the literatures, the impacts of elevation, slope aspect, and slope angle in fire iour have been widely reported (Chuvieco & Congalton 1989; Erten et al 2004;Renard et al.2012; Adab et al.2013)

behav-In the current research, a digital elevation model (DEM) was created by digitizingcontours (30 m interval) and survey base points The DEM map has a grid size of

30 m with 2323 columns and 1657 rows Using the DEM, altitude, slope aspect, andslope angle were extracted (figures 3(a) and (b))

Elevation is a crucial physiographic variable that is associated with temperature,moisture, and wind (Xiangwei et al.2011) Therefore, it has an important role in firespreading (Jaiswal et al.2002) The altitude map was extracted from the DEM andclassified into five classes according to equal interval classification (Pradhan et al

2007; Pourtaghi et al.2014); that is, (1)<500 m, (2) 5001000 m, (3) 10001500 m,(4) 15002000 m and (5) >2000 m (figure 3(a))

Slope aspect is another factor that correlated with the amount of received solarenergy in the area Therefore, slope aspect layer was selected as one of the forestfire-related factors and has been categorized into nine classes: (1) flat, (2) north,(3) north-east, (4) east, (5) south-east, (6) south, (7) south-west, (8) west, and(9) north-west (figure 3(b))

Also, one of the parameters that influence the fire spread rate is slope angle (Weise

& Biging1997) Fire can move more quickly up the slope and less quickly down theslope (Kushla & Ripple1997) So, the slope map of the study area is derived fromthe DEM and divided into four classes such as 05, 515, 1530, and>30

(figure 3(c)) In addition, using the meteorological database, the annual temperatureand wind effect factors were calculated (figures 4(a) and (b))

Temperature highly affects the moisture amount in forest combustion Hightemperature led to dry combustion quickly (Antoninetti et al.1993) The annualtemperature map was classified as follows: <15, 1516, 1617, 1718, and

>18C (figure 4(a)).

Wind is an important factor because it provides fresh oxygen and the flame puts anew fuel source (Rawat2003) Wind effect factor map was created based on threeinput parameters, such as DEM in grid format, wind direction (degree), and windspeed (m/s) in SAGA GIS (http://saga.sourcearchive.com/documentation/2.0.7plus-pdfsg-2/wind effect_8cpp_source.html) In the current research, the wind effect wasprepared in SAGA-GIS and classified based on the natural break classificationscheme (Pourtaghi et al 2014) into three classes such as (0.750.95), (0.951.14),and (>1.14) (figure 4(b))

The land use map was created using Landsat-7 images of 2010 In order to createthe land use map, a supervised classification using maximum likelihood algorithmwas applied A total of 370 signatures (training classes) were collected from all landuse types The signatures were collected by field survey and using GPS Out of these

370 signatures, 250 signatures were used for land use mapping and the remainingwere used for accuracy assessment Nine land use classes were drawn such as irriga-tion farming (IF), dense forest (DF), sparse forest (SF), irrigated and rain-fed mixedfarming (IRMF), rain-fed farming (RF), good range (GR), moderate range (MR),moderate forest (MF), woodlands and shrubbery (WS), and urban (residential) (U)(figure 5)

For assessment of vegetation cover, we used normalized difference vegetationindex (NDVI), which is the most commonly used index to assess live FMC (Chuvieco

Trang 7

Figure 3 Topographical parameter maps of the study area: (a) altitude, (b) slope aspect and(c) slope angle Modified from Pourtaghi et al (2014).

Trang 8

2003) The NDVI was prepared using Landsat-7 images (path 162 and row 34obtained on 13 November 2010) based on the following equation (Rouse et al.1973):

NDVI¼NIR¡ RED

where NIR (band 4) and RED (band 3) values are the infrared and red portion of theelectromagnetic spectrum, respectively In this study, the NDVI map was prepared inENVI 4.8 and divided into six classes (figure 6)

For classification of conditioning factors, different methods were used such asequal interval, natural break, and normal or common standards Finally, for applica-tion of M-AHP and MFL models, all the aforementioned conditioning factors wereconverted to a raster grid with 30 m£ 30 m pixel size in the ArcGIS 9.3 software Allthe maps are in UTM (Universal Transvers Mercator) coordinate system andWGS84 spatial reference (WGS84-UTM-Zone40N)

3.2 Statistical index

In this research, the statistical index (SI) model was applied to illustrate the tive relationship between distributions of forest fire occurrences with predictor fac-tors The SI method is a bivariate statistical analysis proposed by van Westen (1997)

quantita-A weight value for each categorical unit is defined as the natural logarithm of the

Figure 3 (Continued)

Trang 9

Figure 4 Meteorological parameter maps of the study area: (a) annual temperature and(b) wind effect (no dimension) Modified from Pourtaghi et al (2014).

Trang 10

forest fire density in the categorical unit divided by the forest fire density in the entiremap (van Westen1997; Rautela & Lakhera2000; Cevik & Topal2003; Pourghasemi,Moradi, et al 2013) This method is based on the following equation (van Westen

is the total pixels of the entire map

3.3 Modified analytical hierarchy process

The analytical hierarchy process (AHP) is a theory of measurement for consideringtangible and intangible criteria that have been applied to numerous areas, such asdecision theory and conflict resolution (Vargas 1990; Yalcin 2008; Youssef et al

2011) The AHP includes a matrix-based pairwise comparison of the contribution

of different factors on forest fire occurrence The process consists of four phases:Figure 5 Land use map of the study area Modified from Pourtaghi et al (2014)

Trang 11

(1) breaking a complex unstructured problem down into its component factors, i.e ditioning factors considered in this study; (2) combine these factors in a hierarchicalorder; (3) assign numerical values according to the relative important of each factor(pairwise comparison); and (4) synthesize the rating to determine the priorities to beassigned to these factors (Saaty & Vargas2001) The pairwise comparison is the process

con-of comparing the relative importance, preference, or likelihood con-of two elements (forexample, criteria) with respect to another element (for example, the goal) in the levelabove to establish priorities for the elements being compared (Saaty1994;table 2).One of the key points in AHP is calculation of consistency ratio (Saaty1977) Ifconsistency ratio is less of 0.1, then the mentioned matrix can be considered as anacceptable consistency (Saaty1977) Several researchers have used AHP model invarious applications and reported a reasonable accuracy (Ayalew et al.2005; Hajeeh

& Al-Othman2005; Komac2006; Yalcin2008; Esmali Ouri & Amirian2009; Wu &Chen 2009; Langenbrunner et al 2010; Abba et al 2013; Agarwal et al 2013;Kayastha et al.2013; Tierno et al 2013; Zhang et al 2013; Giri & Nejadhashemi

2014) However, it is based on expert opinions and thus may be subjected to tive limitations with uncertainty and subjectivity (Pourghasemi, Moradi, et al.2013).Thus, for solving the limitations of the conventional AHP, an M-AHP is proposed

cogni-by Nefeslioglu et al (2013) and applied in the current research The differencesbetween the M-AHP and the conventional AHP can be classified into two groups:

Figure 6 Normalized different vegetation index (NDVI) map of the study area Modifiedfrom Pourtaghi et al (2014)

Trang 12

(1) The preparation of the factor comparison matrix In this step, there are ing differences:

follow- The factor comparison matrix in the M-AHP model is not according to theexpert opinion,

 The expert viewpoints in M-AHP only used to define the maximum scoresfor each factor in order to prepare factor score matrix,

 Normalization of the factor score values according to maximum scorefactor,

 Finally, construction of factor comparison matrix based on modifiedimportance value scale (Nefeslioglu et al.2013)

(2) The evaluation of the importance distributions of the conditioning factors onthe decision points (Nefeslioglu et al.2013) In this step, at first, each factorwill be normalized based on its own maximum score Subsequently, linear dis-tance between the normalized factor score and decision points will be calcu-lated, and finally decision point comparison matrix will be prepared byconsidering modified importance value scale

In other words, it is sufficient for the expert to identify important factors at modelrunning (http://portal.netcad.com.tr/pages/viewpage.action?pageIdD113803523) Thedetails of the mentioned algorithm/tool (M-AHP) with an example on snow avalanchecan be found in Nefeslioglu et al (2013)

3.4 Mamdani fuzzy logic

The fuzzy set theory was first introduced by Zadeh (1965), and it is one of the toolsused to handle complex problems Fuzzy sets theory is a mathematical method used

to characterize and propagate uncertainty and imprecision in data and functionalrelationships (Kurtener & Badenko 2000) The mentioned theory has been com-monly used in different scientific studies and disciplines (Juang et al 1992; AlvarezGrima & Babuska 1999; Ercanoglu & Gokceoglu 2002; Nefeslioglu et al 2006;Saboya et al 2006; Gokceoglu et al.2009; Yagiz & Gokceoglu2010; Akgun et al

2012; Pourghasemi et al 2012b; Osna et al 2014) In the fuzzy set theory,

Table 2 The AHP pairwise comparison scale (Saaty1994)

elements

Two elements contribute equally

element over another

Experience and judgment favourone element over another

5 Strong importance of one element

over another

An element is strongly favoured

7 Very strong importance of one

element over another

An element is very stronglydominant

element over another

An element is favoured by at least

an order of magnitude

2, 4, 6, and 8 Intermediate values Used to compromise between two

judgments

Trang 13

membership can take on any value between 0 and 1, reflecting the degree of certainty

of membership (Zadeh1965; Pradhan2011) A membership value was chosen trary according to subjective judgement about the relative importance of the mapsand their various states (Bonham-Carter1994) A number of different types of mem-bership functions (MFs) have been proposed for fuzzy inference system These MFsare triangular, trapezoidal, sigmoidal, bell, Gaussian combination, and p-shaped(Pradhan2013; Osna et al.2014)

arbi-Meanwhile, in the literature, different fuzzy inference systems (FIS) have been posed (Mamdani, Sugeno, and Tsukamoto), but the Mamdani fuzzy is one of themost interesting methods applied in engineering geology problems (Alvarez Grima

pro-2000; Akgun et al.2012)

In the Mamdani fuzzy model, ifthen rules replace the usual set of equations used

to characterize a system (Yager & Filev1994) The Mamdani fuzzy model takes thefollowing form:

Ri: if x1 isAi1 and xj isAij; then y is Bi

fori ¼ 1; 2; :::; k and j ¼ 1; 2; :::; r; (3)where k is the number of rules,xjðj ¼ 1; 2; ; rÞ are input variables, y is the outputvariables, and Aijand Biare linguistic terms

In Mamdani model, each rule is a fuzzy relationRiðX£Y ! ½0; 1Þ which is lated using the following equation:

calcu-mRiðx; yÞ ¼ IðmAiðxÞ; mBiðyÞÞ; (4)where the operator I can be either a fuzzy implication or a conjunction operator (t-norm) (Jager1995)

There are four inference steps in Mamdani fuzzy inference system such as tion, rule assessment, aggregation, and defuzzification steps (Mamdani & Assilian

fuzzifica-1975); they are presented in equations (5)(7)

Step 1: compute the degree of fulfillmentaiof the antecedent for each rule i:

Ngày đăng: 25/04/2016, 07:43

TỪ KHÓA LIÊN QUAN

TÀI LIỆU CÙNG NGƯỜI DÙNG

TÀI LIỆU LIÊN QUAN

🧩 Sản phẩm bạn có thể quan tâm