Preliminary Identification of Road Network Vulnerability Due to Expansion of Status of Mount Merapi Eruption-Impacted Areas in Yogyakarta Special Region - Politeknik Negeri Padang

  Preliminary Identification of Road Network Vulnerability Due to Expansion of Status of Mount Merapi Eruption-Impacted Areas in Yogyakarta Special Region

  #

• # * _#

  Imam Muthohar , Hardiansyah , Budi Hartanto , Sigit Priyanto Civil and Enveronmental Engineering, Faculty Of Engineering, Universitas Gadjah Mada, Jalan Grafika No.2, Yogyakarta, 55281, Indonesia

  E-mail: imam.muthohar@ugm.ac.id

• *Civil and Enveronmental Engineering, Faculty Of Engineering, Universitas Gadjah Mada, Jalan Grafika No.2, Yogyakarta, 55281, Indonesia _#

   E-mailom Civil and Enveronmental Engineering, Faculty Of Engineering, Universitas Gadjah Mada, Jalan Grafika No.2, Yogyakarta, 55281, Indonesia E-mail: boed@ugm.ac.id

• *Civil and Enveronmental Engineering, Faculty Of Engineering, Universitas Gadjah Mada, Jalan Grafika No.2, Yogyakarta, 55281, Indonesia E-mail: spriyanto2007@yahoo.co.id

  Abstract —The eruption of Mount Merapi in Yogyakarta, Indonesia in 2010 caused many casualties due to minimum preparedness in facing

  disaster. Increasing population capacity and evacuating to safe places become very important to minimize casualties. Regional government through the Regional Disaster Management Agency has divided disaster prone areas into three parts, namely ring 1 at a distance of 10 km, ring 2 at a distance of 15 km and ring 3 at a distance of 20 km from the center of Mount Merapi. The success of the evacuation is fully supported by road network infrastructure as a way to rescue in an emergency. This research attempts to model evacuation process based on the rise of refugees in ring 1, expanded to ring 2 and finally expanded to ring 3. The model was developed using SATURN (Simulation and Assignment of Traffic to Urban Road Networks) program version 11.3. 12W, involving 140 centroid, 449 buffer nodes, and 851 links across Yogyakarta Special Region, which was aimed at making a preliminary identification of road networks considered vulnerable to disaster. An assumption made to identify vulnerability was the improvement of road network performance in the form of flow and travel times on the coverage of ring 1, ring 2, ring 3, Sleman outside the ring, Yogyakarta City, Bantul, Kulon Progo and Gunung Kidul. The research results indicated that the performance increase in the road networks existing in the area of ring 2, ring 3 and Sleman outside the ring. The road network in ring 1 started to increase when the evacuation was expanded to ring 2 and ring 3. Meanwhile, the performance of road networks in Yogyakarta City, Bantul, Kulon Progo and Gunung Kidul during the evacuation period simultaneously decreased in when the evacuation areas were expanded. The results of preliminary identification of the vulnerability have determined that the road networks existing in ring 1, ring 2, ring 3 and Sleman outside the ring were considered vulnerable to the evacuation of Mount Merapi eruption. Therefore, it is necessary to pay a great deal of attention in order to face the disasters that potentially occur at anytime.

  Keywords — Model, Evacuation, SATURN, vulnerability.

  Evacuation is a process in which people move from a dan-

  I the lives of vulnerably affected people. It is also an effort to

  NTRODUCTION gerous place to a safer place to reduce health problems and I.

  Soil fertility and natural wealth on the slope of Mount completely reduce the number of victims, so that it requires Merapi (Indonesia) are among of the reasons why people liv- various coordinated aspects in determining the policy. ing in this risky area are reluctant to move although a potential

  Transport modeling for evacuation has been developed in sev- threat of disaster is present at any time. It can be said that peo- eral studies to obtain optimum evacuation movement results, ple living in this area are greatly exposed to social vulnera- which can provide an alternative solution to disaster prob- bility and danger of disaster [1][2]. Mount Merapi is one of lems, especially in minimizing casualties, such as design of the world's most active volcanoes and has exploded more than evacuation routes to relieve congestion during evacuation 80 times and the last eruption in 2010 claimed more than 400 [8][9]. lives [3]. Increased population capacity and evacuation to safe

  Macroscopic transport modeling can be used to optimize places are crucial to save them from the danger of disaster road network performance as an evacuation route, such as an

  [4][5]. An optimum evacuation plan by involving road net- attempt to examine the correlation between evacuation pro- work of evacuation routes and safe shelters are absolutely re- cess and road network performance [10][11]. This paper aims quired to ensure the safety of refugees [6][7]. at examining the performance of road networks resulted from

II. M

  (1) where, = constant (result of regression analysis)

  ETHOD

  Demand model and supply model are transport modeling methods that can be used in evacuation model-ing. When the disaster takes place, the number of users of the road networks increases significantly be-cause there are self-rescuing ef- forts. Large number of travelers and possibility of damage to several roads due to disaster make road networks very vulner- able. Evacuation modeling using user equilibrium (UE) can analyze flow and travel times of road network due to emer- gency conditions has resulted in a conclusion that the meas- urement of evacuation performance is highly dependent on the structure of road network and the number of vehicles in the emergency planning zone [12][13]. Several evacuation models have been used to identify critical networks, by taking congestion or ratio between flow and capacity of the road into consideration [14].

  Rise of travel is a central aspect of transport modeling pro- cess. Travel is once movement made by a per-son by using one or more modes and each travel has one origin and one destination [20]. The final stage of transport studies is to make a regression technique more widely. Several travels of trip ends (dependent variable) were observed in each zone and each zone had measurable characteristics in which the rise of travel might be correlated. A regression analysis to analyze changes in road network performance burdened by evacuation process including flow and travel times as dependent varia- bles and the number of refugees in affected area as independ- ent variable is expressed in equation 1.

  = +

  The evacuation model was built and developed using SATURN (Simulation and Assignment of Traffic to Urban Road Networks) software version 11.3.12W, a network anal- ysis software developed by Institute for Transport Studies, University of Leeds. This model focused on the road networks in the Province of Yogyakarta Special Region, involving 140 centroids consisting of 73 zones based on subdistricts of 6 ex- ternal zones and 61 evacuation zones. meanwhile, the buffer node consisted of 449 nodes and involved 851 links consisting of state roads, provincial roads and regency/municipal roads.

  Impacts of changes in natural conditions and of human hands, increased intensity of disasters in recent years has become an interesting research object as a study material, especially con- cerning road vulnerability networks due to disasters. Changes in road network performance due to an event can be identified as network vulnerability and toughness and can be expressed by index [15][16][17]. A methodology of analyzing of road network vulnerabilities is developed based on consideration of socio-economic impact of network degradation and search for determining the most important location as a result of net- work failure [18]. The vulnerability of road transport system is considered not from a security standpoint, but rather as an issue of reduced accessibility for various reasons. It emphasizes on system function and not physical network, although some of the reasons causing road network discontinuity are caused by physical damage [19]. Therefore, increased network performance can be identified as network vulnerability as a results of an event or disaster..

  the expansion of the affected area of Mount Merapi which in- creased the number of refugees, as an attempt to identify ini- tial vulnerability of surrounding road networks.

• 2
• ⋯ +

  1

  ∑

  The scenario was de-signed based on disaster prone areas that had been determined by the regional government through the Re-gional Disaster Management Agency (BPBD) by dividing the areas into three parts, namely ring 1 with a distance of 10 km, ring 2 with a distance of 15 km and ring 3 with a distance of 20 km from the peak of Mount Merapi. The road networks were then drawn with the help of the ArcGis program to make the data input process on SATURN easier. The picture of the road network model using ArcGis is shown in Figure 1..

  1

  1

  ∑ The utilization of existing road network as an evacuation traffic facility can be very difficult when such road network is considered not to have adequate readiness to deal with emergency. In practice, the concept of road network perfor- mance-based evacuation transport model has several ad- vantages, such as its ability to analyze traffic performance for evacuation process on a large scale, and several results of sim- ulation in the form of evacuation time and solid path identifi- cation can produce a measurement of network vulnerability.

  ′′

  =

  ′′′

  ′

  ,

  =

  ′′

  = (2)

  ′

  2

  The Furness method generates flow from the first equilibrium zone and pull into a balanced zone, as expressed in equation

  = independent variable Travel distribution is a central aspect of transport model- ing process. Although travel, distribution and loading are of- ten discussed separately, human behavior makes the three phases correlated to each other. In regard to travel distribu- tion, it is important to know two things about the end of the travel, namely con-nected together, without determining ac- tual route and sometimes without reference to travel mode, with a travel matrix between origin and destination known.

  2

  2.

  Figure 1. Input network model Figure 2. Road network model in SATURN

  Figure 2 shows the road network models to be loaded un- Figure 1 shows road networks and rise zone and pull of travel. Evacuation zone and some regency road network were der several conditions for analyzing perfor-mance. The road network performance observed in the analysis was total vol- placed in Sleman Regency, the closest area from the peak of

  Mount Merapi, as alternative evacuation routes for refugees. ume and travel times on the road network included in ring 1, ring 2, ring 3, Sleman outside the ring, Yogyakarta City, Ban- Furthermore, each ring was limited to make simulation pro- cess and analysis of results easier. tul, Kulon Progo and Gunung Kidul.

  After the design of the road network model was com- pleted, the next step was process to input network and travel matrices into SATURN. The input of road network consisted

ESULTS AND

  ISCUSSION

  III. D R of number of nodes connecting links, free flow speed, speed

  Before the simulation process, the model was firstly vali- during capacity, capacity of road link, number of directions dated by comparing the results of daily travel model to traffic and length of link. Meanwhile, the input of travel distribution counting in the field. The results of model validation in 60 consisted of daily OD matrices and evacuation travel, i.e. national and provincial roads are shown in Figure 3. evacuation route of refugees in ring 1 which was obtained from the daily OD matrix modification added by the number

  8,000

  of refugees to be evacuated with a variation of 50%, 60%, 70%, 80%, 90% and 100%. Similarly, the expansion of evac-

  7,000

  uation to ring 2 was a combination of variation of refugees in

  r) u 6,000 o

  ring 1 add by 80%, 90%, and 100% with 50%, 60%, 70%,

  /h y = 0.5771x + 713.71

  80%, 90% and 100% of the population in ring 2. Furthermore,

  cu 5,000 R² = 0.761

  (p

  the expansion of evacuation to ring 3 was a combination of

  el d 4,000

  variation of evacuation travel in ring 1 add by 90% and 100% o and ring 2 added by 80%, 90% and 100% with 50%, 60%, f M

  3,000 o

  70%, 80%, 90% and 100% of the population moving to ring ts

  sal 2,000 3.

  Re

  Before the simulation on the disaster condition, daily

  1,000

  travel modeling was firstly made as the basic model for the next process. The basic model should have a similarity with

  3,000 6,000 9,000 12,000

  the real condition as evidenced by statistical testing with cer-

  Results of TC (pcu/hour)

  tain value limits. The following is the inputs of the road net- work model in SATURN as shown in Figure 2.

  Figure 3. Road network model in SATURN

  The result of validation show that R Square was 0.761, mean- ing the similarity of the model was 76% with actual condition. Some models could have similarity close to 100%, but it was difficult to obtain for mac-roscopic models. After the valida- tion process was accepted, the next process was performing a predefined scenario simulation. The results of the evacua- tion scenario of evacuating refuges in ring 1, which consisted of 6 variations of the model, obtained regression equations to

  Similarly, when the evacuation scenario of refugees was ex- tended to ring 3 consisting of 36 variations of the simulation model, the regression equations were obtained as shown in Table 3.

  Y = 708071.317 - 14.521 X1 - 97.660 X2 0.262 Bantul Y = 1386271.735 - 28.609 X1 - 56.920X2 0.188 Kulon Progo Y = 563039.405 - 55.887 X1 - 118.486X2 0.687 Gunung Kidul Y = 778742.681 - 103.306 X1 - 50.382X2 0.681

  TABLE 1. EQUATIONS TO DETERMINE FLOW AND TRAVEL TIMES DUE TO REFUGEES IN RING 1 Road Net- work Equation for Determining Flow R Square

  Ring 1 Y = 3,163.967 - 2.773 X1 0.132 Ring 2 Y = 62,795.522 + 115.645 X1 0.904 Ring 3 Y = 178,699.163 + 220.444 X1 0.951 Sleman (out- side the ring) Y = 501,562.304 + 174.213 X1 0.783 Yogyakarta City

  Y = 129,712.163 - 12.416 X1 0.261 Bantul Y = 214,892.707 + 4.925 X1 0.058 Kulon Progo Y = 92,570.859 - 13.229 X1 0.538 Gunung Kidul Y = 106,773.467 - 13.703 X1 0.487

  Road Net- work Equation for Determining Travel Times R Square Ring 1 Y = 18,818.489 - 27.288 X1 0.128 Ring 2 Y = 517,846.707 + 1,183.305 X1 0.891 Ring 3 Y = 1,389,174.076 + 2,436.103 X1 0.942 Sleman (out- side the ring) Y = 2,751,714.837 + 1,251.416 X1 0.765 Yogyakarta City

  Ring 1 Y = 3101.357 + 23.882X1 + 4.597X2 + 0.716X3 0.607 Ring 2 Y = 63987.326 + 584.049X1 + 87.487X2

  TABLE 3. EQUATION TO DETERMINE FLOW AND TRAVEL TIME DUE TO REFUGEES EXPANDED UNTIL RING 3 Road Net- work Equation for Determining Flow R Square

  determine flow and travel times in the observed areas as shown in Table 1.

• 117.084X3 0.962 Ring 3 Y = 177983.405 + 1628.080 X1 + 62.953 X2 + 104.676 X3 0.976 Sleman (out- side the ring) Y = 496034.779 + 2192.804 X1 - 6.300

X2- 410.418 X3

  Y = 502668.801 + 64.248 X1 + 75.369X2 0.497 Yogyakarta City Y = 129797.382 - 1.241 X1 - 16.488 X2 0.273 Bantul Y = 215192.971 - 6.305 X1 - 6.469 X2 0.196 Kulon Progo Y = 92906.276 - 9.713 X1 - 16.029 X2 0.732 Gunung Kidul

  “did not know”. Therefore, when the percentage of the refugees was analyzed using the equations, the total travel flow and travel times in the observed areas were obtained as shown in Figures 4 and 5.

  The results of interviews conducted on those who lived in the disaster-prone areas of Mount Merapi, Sleman, Province of Yogyakarta Special Region indicated that those who would perform evacuation travel with vehicles were 91% of the total population, while the rest traveled by foot and answered

  The equations were then used to determine flow and travel time of the road network in the observed are-as based on the percentage of population moving in each disaster prone area of ring 1, ring 2 and ring 3..

  Kidul Y = 774275.900 + 13.097 X1 - 197.545 X2 - 577.877 X3 0.879

  Kulon Progo Y = 559436.915 + 84.013 X1 - 164.953 X2 - 515.076 X3 0.912 Gunung

  X2 - 889.794 X3 0.761 Bantul Y = 1377501.929 + 1030.629 X1 - 365.223 X2 - 982.230 X3 0.838

  400.400 X2 - 3034.874 X3 0.935 Yogyakarta City Y = 699346.581 + 840.826 X1 - 347.512

  Ring 3 Y = 1381063.484 + 17475.264 X1 + 1035.347 X2 + 909.207 X3 0.974 Sleman (out- side the ring) Y = 2708837.771 + 16027.897 X1 +

  X2 + 26.744 X3 0.635 Ring 2 Y = 530929.269 + 6348.368 X1 + 995.581 X2 + 1226.795 X3 0.964

  Road Net- work Equation for Determining Travel Times R Square Ring 1 Y = 18483.315 + 261.375 X1 + 23.437

  199.425X2 0.888 Sleman (out- side the ring)

  Bantul Y = 213804.088 + 153.412 X1 - 55.926 X2 - 151.268 X3 0.839 Kulon Progo Y = 92384.347 + 7.300 X1 - 22.916 X2 - 75.769 X3 0.923

  0.94 Yogyakarta City Y = 128447.795 + 132.961 X1 - 54.416 X2 - 141.633 X3 0.761

  Y = 707,362.750 - 68.605 X1 0.227 Bantul Y = 1,384,230.913 + 36.679 X1 0.064 Kulon Progo Y = 560,867.402 - 97.408 X1 0.586 Gunung Kidul Y = 778,016.087 - 113.059 X1 0.507

  When the evacuation scenario of the refugees was expanded to ring 2 consisting of 18 variations of simu-lation model, the regression equations were obtained as shown in Table 2.

  TABLE 2. EQUATION TO DETERMINE FLOW AND TRAVEL TIME DUE TO REFUGEES EXPANDED UNTIL RING 2 Road Net- work Equation for Determining Flow R Square

  Ring 1 Y = 3166.328 + 7.268 X1 + 1.850 X2 0.394 Ring 2 Y = 63112.574 + 188.149 X1 + 72.476X2 0.899 Ring 3 Y = 177104.773 + 111.053 X1 +

  699.767 X2 0.339 Yogyakarta City

  Y = 1371677.072 + 1331.666 X1 + 2281.422 X2 0.889 Sleman (out- side the ring) Y = 2754782.232 + 156.050 X1 +

  Y = 523523.971 + 1990.701 X1 + 712.565 X2 0.889 Ring 3

  work Equation for Determining Travel Times R Square Ring 1 Y = 19032.167 + 110.003 X1 + 18.129X2 0.529 Ring 2

  Y = 106940.399 - 12.269 X1 - 7.112 X2

  Gunung Kidul Y= 106397.352 + 0.778 X1 - 23.295 X2 - 75.485 X3 0.889

0.68 Road Net-

  Figure 4. Total flow in the observation areas Figure 5. Total travel time in the observation areas Figures 4 and 5 show the changes in road network perfor- mance due to the evacuation process compared with road net- work performance on daily travel. Accordingly, it can be seen that the total flow and travel times of each road network ob- served in all scenarios produced changes in values. The values increased in some observation areas, but decreased in other areas.

  The results of flow calculation analysis show that the val- ues increased in the road networks in ring 2, ring 3 and Sleman outside the ring. The values of flow in the road net- work of ring 2 consisting of daily travel scenario, evacuation scenario of refugee in ring 1, evacuation scenario of refugees until ring 2 and refugee evacuation scenario until ring 3 were 64,184; 73,319; 86,829; 135,752 pcu/hour respectively. The values of flow in the road network of ring 3 were 179,913; 198,760; 205,358; 341,393 pcu/hour respec-tively. Mean- while, the values of flow in the road network of Sleman out- side the ring were 504,959; 517,416; 515,374; 657,659 pcu/ hour respectively.

  Based on the results of the discussion, it can be concluded that:

  A CKNOWLEDGMENT

  5. The results of the analysis have proved that not all road networks in the Province of Yogyakarta Special Region are vulnerable to the disaster of Mount Merapi eruption.

  4. The preliminary identification of vulnerability had deter- mined that the road networks existing in ring 1, ring 2, ring 3 and Sleman outside the ring were considered vul- nerable to the evacuation of Mount Merapi eruption.

  3. The performance of road networks existing in Yogya- karta City, Bantul, Kulon Progo and Gunung Kidul sim- ultaneously decreased when the evacuation zone was ex- panded.

  2. The road network in ring 1 started to increase when the evacuation was expanded to ring 2 and ring 3.

  1. The expansion of refugee evacuation status from ring 1, ring 2 and until ring 3 increased the flow and travel times in the road networks include in ring 2, ring 3, and Sleman outside the ring.

  ONCLUSIONS

  The results of the analysis of travel time calculation show that there were changes in the form of in-creased values of travel time due to the evacuation scenario of the refugees compared with the daily travel scenario in the road network of ring 2, ring 3 and Sleman outside the ring. The values of travel times in the road network of ring 2 consisting of daily travel scenario, evacuation scenario of refugees in ring 1, evac-uation scenario of refugees until ring 2, and evacuation scenario of refugees until ring 3 were 533,607; 625,527; 769,521; 1,310,867 seconds respectively. The increased val- ues of travel time in the road network of ring 3 were 1,403,447; 1,610,859; 1,700,468; 3,148,267 seconds respec- tively. Meanwhile, the increased values of total travel time in the road network of Sleman outside the ring were 2,774,307; 2,865,594; 2,832,662; 3,927,639 seconds respectively.

  IV. C

  Figure 6 shows that in the implementation of scenario of ref- uges in ring 2 and ring 3, road networks exist-ing in ring 1, ring 2, ring 3 and Sleman outside the ring were identified to be vulnerable. Meanwhile, due to the evacuation scenario of ring 1, the road networks identified to be vulnerable were only in ring 2, ring 3 and Sleman outside the ring. However, the road networks existing in ring 1 could be concluded to vul- nerable because it had been identified in the scenario of ring 2 and ring 3.

  Figure 6. Road networks identified to be vulnerable

  Therefore, the road networks were initially identified to ex- perience vulnerability due to the increased val-ues of flow and travel time as shown in Figure 6.

  Furthermore, flow and travel times decreased from daily travel to evacuation travel as the evacuation status was ex- panded to ring 3, i.e. the road network existing in Yogyakarta City, Bantul, Kulon Progo and Gunung Kidul. The implemen- tation of the evacuation scenario from ring 1, expanded to sce- nario ring 2 and until the scenario ring 3 made the values of flow and travel times continuing to decrease, indicating that some travels in these areas were delayed due to the disaster of Mount Merapi.

  Other obtained results show that there were changes in flow and travel times of the road network in ring 1 due to the influ- ence of daily travel scenario, evacuation scenario of refugees in ring 1, evacuation sce-nario of refugees in ring 2 and evac- uation scenario of refugees until ring 3, namely 3,236 ; 2,912; 3,996; 5,758 pcu/hour and 19,832; 16,335; 30,692; 46.835 seconds respectively. These results indicate that the values of flow and travel times of the road network increase when the evacuation scenario of refugees was expanded until ring 2 and ring 3. This indicates that there was a phenomenon of inter- ference in the road network.

  would like express my sincere gratitude to The Ministry of Reseach, Technologyand Higher Education, which provided scholarship for taking Doctoral Program at Universitas Gad- jah Mada, Head of the Doctoral Program of Civil Engineer- ing, Universitas Gadjah Mada, Supervisor who assisted in the preparation of this research article, and Co-Supervisor who assisted in.

• – 369, 2016.

• –542,

  2004.

  H. Fu, A. J. Pel, and S. P. Hoogendoorn, “Optimal traffic management to ensure emergency evacuation compliance,” 2013 10th IEEE Int. Conf. NETWORKING, Sens. Control , pp. 532

  [18] M. a P. Taylor, S. V. C. Sekhar, and G. M.

  Geogr. , vol. 14, no. 3, pp. 215 –227, 2006.

  D. M. Scott, D. C. Novak, L. Aultman-Hall, and F. Guo, “Network Robustness Index: A new method for identifying critical links and evaluating the performance of transportation networks,” J. Transp.

  81, pp. 4 –15, 2015. [17]

  A. Reggiani, P. Nijkamp, and D. Lanzi, “Transport resilience and vulnerability : The role of connectivity,” Transp. Res. Part A, vol.

  [16]

  C. Balijepalli and O. Oppong, “Measuring vulnerability of road network considering the extent of serviceability of critical road links in urban areas,” J. Transp. Geogr., vol. 39, pp. 145–155, 2014.

  [15]

  E. Leal, D. Oliveira, and W. Porto, “Determining critical links in a road network : vulnerability and congestion indicators,” Procedia - Soc. Behav. Sci. , vol. 162, no. Panam, pp. 158 –167, 2014.

  [14]

  Hoogendoorn, “Link-level vulnerability indicators for real-world networks,” Transp. Res. Part A, vol. 46, no. 5, pp. 843–854, 2012.

  a. G. Hobeika and C. Kim, “Comparison of traffic assignments in evacuation modeling,” IEEE Trans. Eng. Manag., vol. 45, no. 2, pp. 192 –198, 1998. [13] V. L. Knoop, M. Snelder, H. J. Van Zuylen, and S. P.

  [12]

  Multiobjektif,” J. Transp., vol. 16, no. 3, pp. 231–240, 2016. [11]

  [10] P. S. Hardiansyah, Suparma BS, Muthohar I, “Perencanaan Transportasi untuk Evakuasi Fokus pada Pengungsi Model dengan Pendekatan

  A. Stepanov and J. M. Smith, “Multi-objective evacuation routing in transportation networks,” Eur. J. Oper. Res., vol. 198, no. 2, pp. 435 –446, 2009. [9] Z. Zhang, S. a. Parr, H. Jiang, and B. Wolshon, “Optimization model for regional evacuation transportation system using macroscopic productivity function,” Transp. Res. Part B Methodol. , 2015.

  [8]

  [7] Y.-C. C. Y.-C. Chiu, “Traffic scheduling simulation and assignment for area-wide evacuation,” Proceedings. 7th Int. IEEE Conf. Intell. Transp. Syst. (IEEE Cat. No.04TH8749) , pp. 537

  A Stat. Mech. its Appl. , vol. 428, pp. 396 –409, 2015.

  [5] T. Hayashi, “Disaster Prevention Education in Merapi Volcano Area Primary Schools : Focusing on Students ’ P erception and Teachers ’ Performance,” Procedia Environ. Sci., vol. 20, pp. 668– 677, 2014. [6] J. Wang, L. Zhang, Q. Shi, P. Yang, and X. Hu, “Modeling and simulating for congestion pedestrian evacuation with panic,” Phys.

  E. Tyas, W. Mei, A. Fajarwati, S. Hasanati, and I. Meilyana, “Resettlement following the 2010 Merapi Volcano eruption,” Procedia - Soc. Behav. Sci. , vol. 227, no. November 2015, pp. 361

  [4]

  — A ‘ 100-year ’ event,” J. Volcanol. Geotherm. Res. , vol. 241 –242, pp. 121–135, 2012.

  Budisantoso, F. Costa, S. Andreastuti, F. Prata, D. Schneider, L. Clarisse, H. Humaida, S. Sumarti, C. Bignami, J. Griswold, S. Carn, C. Oppenheimer, and F. Lavigne, “The 2010 explosive eruption of Java †^TM s Merapi volcano

  [3] P. Jousset, J. Pallister, M. Boichu, M. F. Buongiorno, A.

  Sci. , vol. 227, no. November 2015, pp. 387 –394, 2016.

  63 –77, 2016. [2] M. B. Rahman, I. Susana, and S. Purwo, “Community resilience : learning from Mt Merapi eruption 2010,” Procedia - Soc. Behav.

  R EFERENCES [1] S. J. Ki, “Social vulnerability at a local level around the Merapi volcano,” Int. J. Disaster Risk Reduct., vol. 20, no. October, pp.

• –537,

  2013.

  D’Este, “Application of accessibility based methods for vulnerability analysis of strategic road networks,” Networks Spat. Econ., vol. 6, no. 3–4, pp. 267– 291, 2006. [19] K. Berdica, “An introduction to road vulnerability: What has been done, is done and should be done, ” Transp. Policy, vol. 9, no. 2, pp. 117 –127, 2002. [20] Salter R. J.,Hounsell NB., Higway Traffic Analysis and Design, Third Edition . 1996.