Prediction of Refugee Population in School Districts Chunhao SHAO ...

jsnds.org
  • No tags were found...

Prediction of Refugee Population in School Districts Chunhao SHAO ...

Journal of Natural Disaster Science, Volume 33, Number 1, 2012, pp37-48Prediction of Refugee Population in School DistrictsChunhao SHAO*Takeyoshi TANAKA***Graduate School of Engineering, Kyoto University**Disaster Prevention Research Institute, Kyoto University(Received May 31, 2011 Accepted December 10, 2011)ABSTRACTThis study developed a numerical formula based on statistical data on collapsed, burntdown buildings, and refugees taking shelter in schools during the Great Hanshin Earthquakeof 1995. The data on building damage used in the study were 2004 municipalities, classifiedinto 46 junior high school districts. The independent variables were collapsed buildings,burnt down buildings, and average population of each building. The refugee populationwas regarded as a dependent variable in this study. The formula effectiveness in the juniorhigh school district was 63% and 98% in administrative wards. When predicting the refugeepopulation in schools in Kobe City using the formula, the school refugee population was80% of all refugees in Kobe City predicted by the Hyogo Prefectural Government. Thisoutcome was the same percentage of the population in school refuges in the Great HanshinEarthquake. Predicting the refugee population who took shelter in other places besidesschool refuges and controlling other effective intermediary variables to fit different localareas are issues worth pursuing in advanced research.Keyword: School refuge, School district, Refugee prediction, Collapsed building, Burntdown building1. INTRODUCTIONThe Great Hanshin Earthquake, with a seismicintensity of seven, occurred on January 17, 1995at 5:46 AM. The range of the earthquake includedKobe City and Awaji Island. Over 50% of the zonesdestroyed were urban areas with high-density populations,and people in large numbers therefore tookshelter in streets and buildings. As a result, mostpeople went to refuges designated by the local government,and most refuges in Kobe City were nearbyschool refuges including elementary schools, highschools, etc. According to the investigation outcomeconducted by Yokota (1998), the refugee populationwho took shelter in school refuges composed 60% ofall refugees in Kobe City during the Great HanshinEarthquake. Based on the phenomenon, it is importantto grasp the population in school refuges in cityplanning for disaster prevention in urban areas. Bycorrectly predicting the population in school refuges,relief supplies and rescue actions taken by officialinstitutions responsible and NPOs could becomemore effective and appropriate. Therefore, this studycorrectly predicts the refugee population in schoolrefuges by collapsed and burnt down buildings, andthe population of collapsed and burnt down buildings,to provide a positive suggestion in predicting refugeepopulation as a reference for policy-making in disas-37


C. SHAO, T. TANAKAter prevention and future urban planning.2. LITERATURE REVIEWStudies related to predicting population in schoolrefuges are few. Several important studies cited heredescribe research conducted around this issue.(1) Kimura and Hayashi et al., in the report, “Predictionof Earthquake in Kyoto City for the Third Time”(2009) and “Development of a Method of Estimatingthe Number of People at Evacuation Centers in Caseof Urban Earthquake Disaster” (2004), predictedpopulation change at refuges over ten hours by the independentvariable of seismic intensity. The formulawas expressed as:Here, PE is the total refugee population, PEW is therefugee population caused by damaged wooden buildings,PER is the refugee population caused by damagednon-wooden buildings, and PEL is the refugee populationcaused by lifeline damage.(4) Kawasaki and Nagahashi (1996) utilized the damagerate of wooden buildings to predict the rate ofrefuge. The seismic intensity of an earthquake wascontrolled to two types: the M5 and the M7 level; therefugee population was counted by the 3-m 2 unit ofspace occupied by each person in a refuge. The studyonly considered damage to wooden buildings, and didnot include other possible variables in the prediction.The equation for the rate of taking shelter is as expressedbelow.y10 = (0.0863x 2 ‒ 0.7935x + 1.8108) × population.(1)RP = (BC + 0.5 BH + 0.1 BP) ÷ BW.(4)Here, y10 is rate of refugees taking shelter at the 10 thhour after the earthquake, and x is seismic intensity.The study was conducted for the entire range ofKyoto City, and regarded seismic intensity as an independentvariable to predict the entire refugee populationin the city.(2) Sakata (1999) used the data of every administrativeward during a 10-day period to 50 days after theoccurrence of the Great Hanshin Earthquake, andregarded the decreased population in school refugesas an independent variable for building a model to explainthe change in the school refugee population. Theformula was expressed as:Y = a ·EXP (bx).(2)Here, a is a constant, b is rate of decrease in the refugeepopulation, Y is the change in refugee population,and X is the number of days after the earthquake.(3) In the Mieno study (1999), every administrativeward in Fukuoka City was taken as the researchobject, and the status of refugee house damage wasclassified into three types: collapse of wooden buildings,collapse of non-wooden buildings, and damageto the water supply system. The independent variableswere the rate of completely collapsed buildings, firespread, and damage to the water supply system. Theoutcome still did not approach the area of schooldistricts. The equation used in the study is expressedbelow:PE = PEW + PER + PEL.(3)Here, RP is the rate of people who should take shelterin refuge, BC is the number of completely collapsedbuildings, BH is the number of half-collapsed buildings,BP is the number of partly collapsed buildings,and BW is the number of undamaged whole buildings.The literature does not include any studies focusingon prediction of the refugee population in schoolinstitutions in each school district. The independentvariables used to predict the refugee population werethe seismic intensity of the earthquake, collapse rate,rate of damage to lifelines, and change in the refugeepopulation. The prediction objects ranged from administrativewards to the entire city or special areas.Compared to past studies, this study chose the basicunit of “school district” in city disaster preventionas the predicted object of official school refuges. Tograsp the relationship between residents of damagedbuildings and the refugee population, the populationsof collapsed and burnt down buildings were regardedas independent variables. The purpose of this studywas not only to predict the refugee population of aschool refuge in a broad area of each ward or the entirecity, but also to grasp information on small areassuch as one school district.3. METHODOLOGYThe methods used in this study are based on theliterature and statistic analysis. The former composesthe main database, and the latter is the analysis process.These methods are explained in detail below.38


PREDICTION OF REFUGEE POPULATION IN SCHOOL DISTRICTS3.1 Literature analysisThis study used quantification of literatureanalysis as a research method. The building damagedata used in this study were based on the document,“Final Report on the South Earthquake in HyogoCounty,” published by the Building Research Instituteof the Ministry of Construction in 1998. Statistics onrefugee population at the peak time in each schoolrefuge on January 25, 1995 are presented in a reportpublished by the Education Committee in Kobe Cityand in the research of Kashihara (1998). Householdand population statistics are provided in the city censuspublished by the General Affairs Planning Departmentof Kobe City Hall (1995). The data mentionedabove were collected for the statistical analysis.The data process steps are described below.(1) First, building collapse and destruction by firewere sorted for every municipality in the schooldistrict. The 2004 municipalities were classifiedinto 46 junior high school districts, which alsoincluded elementary school districts. School districtclassification was based on information providedby the Kobe Education Committee. Schooldistrict distribution was according to damagedmunicipalities, as shown in Fig. 1. The numberin each municipality indicates the junior highschools belonging to that municipality.(2) To understand the status of damaged buildingsand total refugee population in each schooldistrict, statistics on various types of damagedbuildings and peak population on January 25,1995 in every school were calculated accordingto the school district. Table 1 lists the manipulateddata.3.2 Process of statistical analysisThe method utilized in this study arranges roughdata through “regression analysis.” This method waschosen because regression built for prediction takesinto account the percentage of factors chosen as independentvariables (population of damaged buildings)and explains the dependant variable (refugeepopulation). The prediction quantity can predict reliefsupplies and other rescue actions needed when makingpolicies related to disaster prevention and urbanplanning. In this study, the dependent variable was therefugee population in school refuges in junior highschool districts, and the independent variables werethe number of completely collapsed buildings, numberof burnt down buildings, and average populationof buildings in the school district.Fig. 1. Distribution of school districts in damaged municipality units39


C. SHAO, T. TANAKATable 1. Data on damaged buildings and refugee population in school districtsNote:CCB: Number of completely collapsed buildings in the school districtHCB: Number of half-collapsed buildings in the school districtBB: Number of burnt down buildings in the school districtBB Mid-R: Report on middle by the Construction MinistryM.H-rise: Middle- and high-rise buildings40


PearsonProduct-MomentCorrelation(r coefficient)PREDICTION OF REFUGEE POPULATION IN SCHOOL DISTRICTSTable 2. Pearson product moment correlation among variables in the studyVariablePopulationofRefugeesin SchoolRefuges(yg)Number ofCompletelyCollapsedBuildings(xc)Numberof BurntDownBuildings(xf)NumberofBuildings(xb)Populationin theSchoolDistrict(xp)yg 1.000 .830 .317 .497 .090xc .830 1.000 .420 .540 .085xf .317 .420 1.000 .107 -.219xb .497 .540 .107 1.000 .681xp .090 .085 -.219 .681 1.000To prove the relationship between the independentand dependent variables, the Pearson productmomentcorrelation (r coefficient) was used to understandthe basic relationship among variables in thestudy. Table 2 shows a positive relationship betweenthe most independent and dependent variables. Thestrength of relationship from high to low is the numberof collapsed buildings, the number of buildings inschool districts, the number of burnt down buildings,and the school district populations. The r coefficientamong independent variables is low and middling(


C. SHAO, T. TANAKAFig. 2. Scatter plot for yg, xxp/b, and xfxp/bvariables, which is also a multicollinearity problem,the VIF value should be less than 10. The VIF valuefor the number of collapsed buildings was 1.058, forthe number of burnt down buildings, it was 1.215, andfor the average population of buildings, it was 1.087.In the equation, there was no multicollinearity problemamong the three variables.In addition to the equation for school districts, tomatch data based on administration ward, a formulafor administration wards was also developed:Here,ny a = Σ y g .i =1(10)ya is the population of refugees in school refuges ineach ward.yg is the population of refugees in school refuges injunior high school districts.n is the number of school districts in each ward.R 2 of regression for administrational wards is70%, which means that the independent variables canexplain 70% of the dependent variables. The formulawill be used in the fourth part of this article to comparethe population prediction of refugees in entireKobe City provided by Hyogo Prefecture.(2) Influence of burnt down buildings on refugeepopulation predictionFrom R 2 stated above, we find that the populationof collapsed buildings has higher prediction effectivenessfor the refugee population than the burnt downbuilding population. However, burnt down buildingsplayed an important role in the study. All school districtswere divided into two types, one being districtswith no burnt down buildings, and the other beingdistricts with burnt down buildings. A comparison ofthe effectiveness of prediction for the refugee populationby the number of collapsed buildings is shown inthe scatter plot of Fig. 3.In Fig. 3, the number of genuinely collapsedbuildings has a positive correlation with school district.This relationship is manifest in school districtswith both burnt down buildings and no burnt downbuildings. R 2 in the former is 0.628, and in the latter,it is 0.605. Therefore, the prediction of “collapsedbuildings” in burnt down districts is a little higherthan in non-burnt down districts. In this study, theprediction effectiveness (R 2 ) with the contribution ofburnt down buildings is near to 0.0539, and this valueis close to the difference in R 2 between collapsedbuildings in these two kinds of school districts shownin Fig. 3. In other words, the “number of collapsedbuildings” has a strong statistical predictive power forboth kinds of school district, and the “number of burntdown buildings” can boost the prediction effectivenessof collapsed buildings in places of conflagrationoccurrence.42


PREDICTION OF REFUGEE POPULATION IN SCHOOL DISTRICTSFig. 3. Distribution of collapsed buildings and refugee population in two types of areas4. COMPARING PREDICTION OUTCOMESBETWEEN THE STUDY AND OFFICIAL DATAThis section makes a prediction based on theformula in this study for current-day Kobe City, andthen compares this prediction with the prediction bythe Hyogo Prefectural Government based on the assumptionthat the Rokko-Awajishima fault caused theearthquake. The numerical formula used in this sectionis the prediction for administrative wards, such asequation (10) given in the third part of the article.y a = Σ n i = 1 y g .(10)Similar to the information shown in Table 3, accordingto the prediction by the Hyogo PrefecturalGovernment on the assumption that the Rokko-Awajishimafault caused the earthquake, the number ofcollapsed buildings might be 60,322, and the numberof burnt down buildings might be 21,455 during thetime interval from 4:00 AM to 5:00 AM during thewinter season. The potential refugee population inKobe City might be 128,236. Because the Hyogo PrefecturalGovernment did not state the potential refugeepopulation or building damage based on each administrativeward, the refugee population, number ofcollapsed buildings, and number of burnt down buildingsin each ward were distributed by the percentagein each ward during the Great Hanshin Earthquake.The average building population was also calculatedin the same way. Figure 4 shows a comparison of therefugee population between this study and the officialdata in each ward.On the other hand, when making predictionbased on the formula in this study, the population ofrefugees in school refuges in school districts of KobeCity was 104,375. Compared to population of refugeesin entire Kobe City predicted by Hyogo Prefecture,there was a difference of 23,861 people. As describedabove, the prediction given by the formula inthis study was 104,375 people, which is 80% of HyogoPrefecture’s assumption, 128,236 people. The reasonis that the population of refugees in this study wasfocused on people staying in school refuges. In otherwords, people who take shelter in other places, suchas parks, homes of relatives, and community centers,were not included when making the prediction. However,the percentage of refugee population in schoolrefuges during the Great Hanshin Earthquake wasover 60% of the refugee population in entire KobeCity. The percentage is close to the prediction in thisstudy when compared to the refugee population inentire Kobe City predicted by the Hyogo PrefecturalGovernment.43


C. SHAO, T. TANAKAWhen comparing each administrative ward, thereis a large difference between this study and predictionoffices of each ward for Nagata and Tarumi Wards.The difference reflects the fact that the conflagrationin these two wards would have influenced the predictionin this study, because burnt down buildings havea negative relationship with the refugee population inthis study. For this reason, the prediction made in thisstudy is lower for both Nagata and Tarumi Wards.Table 3. Prediction of collapsed buildings, burnt down buildings, and refugee population inevery ward of Kobe CityWard C.B B.B P.B P.R P.R (H.P)Higashinada 12,246 1,007 1.43 15,750 35,043Nada 11,414 1,432 1.06 16,173 20,206Hyogo 8,529 2,896 1.11 17,615 15,183Nagata 13,887 14,660 1.3 51,318 20,356Suma 6,886 1,254 0.68 12,217 12,162Tarumi 1,052 3 2.83 4,553 3,598Kita 242 3 5.92 3,925 1,356Chuo 5,676 200 1.35 8,631 19,305Nishi 390 0 1.88 4,032 1,027Kobe City 60,322 21,455 1.95 104,375 128,236Note.C.B: Number of collapsed buildings in the predictionB.B: Number of burnt down buildings in the predictionP. B: Population of damaged buildings on average in the predictionP. R: Population of refugees in the prediction of this studyP. R (H.P): Population of refugees in the prediction based on the data of Hyogo PrefectureFig. 4. Comparison of refugee populations in wards44


C. SHAO, T. TANAKA(3) Chuo WardThe refugee population predicted by this studyis 8,631 people, with 19,305 refugees predicted byHyogo Prefecture in entire Kobe City, which is a multiplicationof 15.8%, or the percentage of the refugeepopulation in Chuo Ward in the Great Hanshin Earthquake.The case of Chuo Ward is like HigashinadaWard, the extreme difference being a result of refugeesfrom other school districts or wards. Moreover,there are refugees who evacuate for other reasons, besidescollapsed and burnt down houses. In particular,there were many downtown refuge sites in Chuo Wardduring the period of the Great Hanshin Earthquakeaccommodating many refugees from other wards.The difference of 10,674 people shows the refugeepopulation from other wards and evacuation for otherreasons than collapsed and burnt down houses.CONCLUSIONIn this study, the prediction effectiveness forschool districts is 65% and 70% for the administrativewards. When considering the prediction effectivenessbetween individual variables, the number of collapsedbuildings was a powerful factor in predicting refugeepopulation, but in school districts with conflagration,it was powerless to predict the refugee population.This is because the death of residents caused by conflagrationas well there being refugees from otherschool districts influenced the relationship betweenthe independent and the dependent variables.Regarding application of the formula, this studyalso conducted a comparison of prediction outcomesfrom this study and those from the Hyogo PrefecturalGovernment. The prediction of refugee populationin this study was 80% (104,375 people) of the refugeepopulation (128,236 people) in entire Kobe Citypredicted by Hyogo Prefecture. According to the t-test, the difference approached statistical significance.Therefore, the prediction in this study was proven,and the percentage was also supported by the dataon the Great Hanshin Earthquake. An irregular differencewas found between individual wards in therefugee population. This was because the predictionby Hyogo Prefecture was for entire Kobe City, but notfor each ward. The prediction for each ward from theprediction office was based on the percentage of therefugee population of each ward in the Great HanshinEarthquake. For this reason, there was an irregulardifference in prediction between this study and that ofthe prediction office in each ward. Continuous explorationof further information about the prediction ofeach ward by other offices or other institutions wouldenable a comparison with the prediction of this studyfor a more accurate outcome.Based on the above findings, the contribution ofthis study encompasses several points. First, the predictionof refugee population in this study can includeschool districts and administrative wards with significantstatistical levels. The equation formula showsthe influence of damaged buildings on the refugeepopulation, which provides more detail than using theseismic intensity of an earthquake as an independentvariable. Finally, the prediction of school refugeepopulation enables an accurate calculation of essentialrelief supply and the adoption of more effective rescueaction as a basic reference for policy-making indisaster prevention and urban planning.Future research would be to investigate how topredict the refugee population among those who donot take shelter in school refuges and to improve theprediction effectiveness. Controlling the influence ofintermediary variables, such as refugees from otherschool districts and fatality count, could apply to predictionsmade by other cities. In the Great East JapanEarthquake on March 11, 2011, besides schools, otherrefuge sites also played an important role as refugeeaccommodation and supply distribution.Based on the survey conducted by the CabinetOffice of the Japan Government and the ReconstructionHeadquarters in Response to the Great East JapanEarthquake on June 2 nd and September 22 nd , therefugees in schools and public halls in Fukushima,Miyagi, Iwate, and other prefectures damaged inthe Earthquake were 41,143 people on June 2 nd and2,840 people on September 22 nd . The refugees takingshelter in hotels were 28,014 people on June 2 nd and4,337 people on September 22 nd . Those taking refugeat relatives’ or friends’ houses were 32,483 on June2 nd and 17,683 on September 22 nd . Refugees living intemporary housing, hospitals, and other prefabricatedhouses were 22,954 people on June 2 nd and 2,840 peopleon September 22 nd . From the whole image of thechange in refugee population in different refuge sites,compared to the Great Hanshin Earthquake, becausemany schools and designated refuge sites were de-46


PREDICTION OF REFUGEE POPULATION IN SCHOOL DISTRICTSstroyed by the tsunami caused by the earthquake, therefugee population taking shelter in hotels is greaterthan in the Great East Hanshin Earthquake. Moreover,the population in temporary housing, hospitals, andother prefabricated houses continued increasing fromJune to September. The phenomenon indicates thatrefugees selected temporary housing and other publicprefabricated houses to live in for middle- and longtermrefuge.The data on the Great East Japan Earthquakedescribed above should be regarded as an importantreference and adjusted to the equation of the formuladeveloped in the study. Finally, building a predictionmodel to fit not only west Japan, but also northeastJapan in urban areas, rural areas, and harbor villagesand providing suggestions for the planning of disasterprevention in different areas are essential in future research.REFERENCESBuilding Research Institute of Ministry of Construction,1998, Final Report on the South Earthquake in HyogoCounty (in Japanese).Building Research Institute of Ministry of Construction,1995, Final Report on the South Earthquake in HyogoCounty (in Japanese).Cabinet Office of the Japan Government, 2011, Survey onRefugee Population in Japan during Great East JapanEarthquake, accessed on October 10, 2011 (in Japanese).Crisis Management Department in Kobe City, 2010, Reporton the Damage Status and Reconstruction of Kobe Cityin the Great Hanshin Earthquake, 1-17 (in Japanese).Disaster Prevention and Crisis Management Department ofKyoto City, 2009, Prediction of Earthquake in KyotoCity for the Third Time, 38-55 (in Japanese).General Affairs Planning Department of Kobe City Hallhomepage, 1995, accessed onJanuary 15, 2011 (in Japanese).Hyogo Prefecture Homepage, 2006, accessed on January 15, 2011.Kashihara, S., Ueno, J., Morita, T. & Yokoda, T., 1998,Study on Refuges in the Great Hanshin Earthquake,265-283, Osaka University Press, Osaka (in Japanese).Kawasaki, J. and Narashino, S., 1996, A study on the Presumptionof Refugee Population and Capacity of RefugePlaces in the Earthquake Disaster: A Case Studyfor the Tsudanuma Area, Nawashino City, Chiba Prefecture,Summary Report of the Architectural Instituteof Japan, 89-90 (in Japanese).Kimura, R., Hayashi, H., Tatsuki, S. & Tamura, K., 2004,Development of a Method of Estimating the Numberof People at Evacuation Centers in Case of UrbanEarthquake Disaster, 13 th World Conference on EarthquakeEngineering, Paper No. 1306.Kobe Education Committee, 1998, Reconstruction of DamagedCampuses in the Great Hanshin Earthquake, 1-6(in Japanese).Mieno, Y., Taga, N. & Seike, K., 1999, Study on Predictionof Capacity of Refuges for the Earthquake in FukuokaCity, Report of the Kyushu Branch of the ArchitecturalInstitute of Japan, 38, 219-220 (in Japanese).Reconstruction Headquarters in Response to the Great EastJapan Earthquake, 2011,< http://www.reconstruction.go.jp/> accessed on October10 th , 2011 (in Japanese).Sakata, K., 1999, A Study on Modeling the Transition in theNumber of Refugees in Shelters after Earthquake Disasters:A Study on the Shelters in the Great Hanshin-Awaji Earthquake Disaster, Report of the Kinki Branchof the Architectural Institute of Japan, 39, 157-160 (inJapanese).Yokota, T., 1998, Study on Refuges in the Great HanshinEarthquake: Chapter 2. Occurrence of Refuge and Behaviorof Refugees, pp. 35-66, Osaka University Press,Osaka (in Japanese).47

More magazines by this user
Similar magazines