2024年俄亥俄州日食全州事件建模.pdf

《2024年俄亥俄州日食全州事件建模.pdf》由會員分享，可在線閱讀，更多相關《2024年俄亥俄州日食全州事件建模.pdf（32頁珍藏版）》請在三個皮匠報告上搜索。

1、Statewide Event Modeling for the 2024 Eclipse in OhioPresenters Rebekah Straub Ohio Department of Transportation Roberto Miquel Whitman,Requardt&Associates,LLP Jonathan Avner Whitman,Requardt&Associates,LLP Greg Giaimo WSPConference on Innovations in Travel Analysis and PlanningBackground There will

2、 be a total solar eclipse passing through Ohio in 2024.ODOT wants to be able to assist in planning for and positioning resources on eclipse day to facilitate smooth traffic operations.Due to the extent of the geographic impact of the eclipse on Ohio,the Ohio Statewide Model is the only tool of suita

3、ble scale for this analysis.Goal was to create an Eclipse Day event model for Ohio using data collected from the 2017 eclipse in Kentucky and TennesseeConference on Innovations in Travel Analysis and Planning2017 Eclipse Total solar eclipse occurred on August 21,2017 70-mile-wide path of totality(ar

4、ea from which a total solar eclipse can be observed)Stretched from Oregon to South Carolina Passed through Kentucky and Tennessee Eclipse day specific changes to traffic caused disruptions in traffic flows on key facilities entering and leaving the path of totalityConference on Innovations in Travel

5、 Analysis and PlanningMap of 2017 Path of TotalityConference on Innovations in Travel Analysis and Planning2024 Eclipse Total solar eclipse will occur on April 8,2024 Will stretch from Maine to Texas Will pass through Ohio covering approximately half of the state Cleveland,Dayton,and Akron are all i

6、n the path of totality.The northern edges of Columbus and Cincinnati are also in the path Other key population centers outside of Ohio are also in the path of totality,which may reduce trips entering Ohio on eclipse day(Indianapolis,Erie,Buffalo).Conference on Innovations in Travel Analysis and Plan

7、ningMap of 2024 Path of TotalityConference on Innovations in Travel Analysis and PlanningClassification of Trips II trips:Trips beginning and ending in the path of totality IE trips:Trips traveling from the path of totality EI trips:Trips traveling to the path of totality EE trips:Trips traveling th

8、rough or avoiding the path of totality Excluded trips:Trips interacting with the path of totality in areas outside of KY/TN or OhioConference on Innovations in Travel Analysis and PlanningKY/TN Analysis ZonesConference on Innovations in Travel Analysis and PlanningOhio Statewide Model TAZsConference

9、 on Innovations in Travel Analysis and PlanningExploring Travel Patterns Considerable noise in the data Difficulty in discerning signal from noise Focused on trips where changes were undeniable Changes in trip lengths seen as key to understanding patterns II trips and EE trips showed small but notic

10、eable reductions in trip lengths People in path of totality take the day off?Reduce their travel or stay near viewing areas?IE trips and EI trips show dramatic increases in trip lengths People making non-typical trips from further away to see the eclipse?IE/EI trips became focus for model approachCo

11、nference on Innovations in Travel Analysis and PlanningAverage Trip Lengths KY/TNConference on Innovations in Travel Analysis and PlanningTravel Time(Min)ResidentsVisitorsCombinedRegularEclipseRegularEclipseRegularEclipseOverall181723212119II141311101312IE3542721164663EI394274904955EE171522192118Lon

12、ger travel times for IE/EI trips on eclipse day than regular day imply that new trips are being made by people to the path of totality from farther away for the sole purpose of observing the eclipse.For example,traveling from Columbus to Nashville to view the eclipse when one otherwise would not.Tra

13、nslating TLDs The Ohio statewide model has different trip lengths than the ones observed in KY and TN.Unable to use KY and TN eclipse day trips lengths directly.Needed to translate KY and TN behaviors to relate to Ohio behaviors.Calculate the difference in regular day TLDs and eclipse day TLDs for K

14、Y/TN.Calculate the difference in regular day TLDs bin by bin between KY/TN and Ohio.Factor the differences in regular day TLDs between KY/TN and Ohio by the differences between regular day and eclipse day for KY/TN.New eclipse day Ohio TLDs to be used for calibrating eclipse gravity models for IE an

15、d EI.Conference on Innovations in Travel Analysis and PlanningTLD-IIConference on Innovations in Travel Analysis and PlanningTLD-IEConference on Innovations in Travel Analysis and PlanningTLD-EIConference on Innovations in Travel Analysis and PlanningTLD-EEConference on Innovations in Travel Analysi

16、s and PlanningIdentifying Changes in Trip Ends Looked at changes in trip ends between regular day and eclipse day.A lot of variability,especially close to and within the path of totality.Looked at relationships between population and other SE characteristics in trip end changes.Used R to try to iden

17、tify any trends.Changes in trip ends positively correlated with changes in population.Also looked at average travel times to population.Wanted to understand role of proximity.As TAZs are further from the path of totality,more novel eclipse day trips are created.Conference on Innovations in Travel An

18、alysis and PlanningChanges in Trip Ends by PopulationConference on Innovations in Travel Analysis and PlanningIndependent Variables Analyzed Desire to see what if any characteristics explain where changes in demand would occur.Had to assess changes in origins and changes in destinations for IE trips

19、.Initial variables were based on attribute already present in the model:Population Employment Hotel Rooms IE destinations are outside of the path of totality.How do TAZs outside of the path of totality interact with the path of totality?Conference on Innovations in Travel Analysis and PlanningAverag

20、e Weighted Travel Time Example Travel Times in Minutes for TAZ 3 Minimum time=50 Average time=(100+75)/2=87.5 Average weighted time=(100*100)+(75*200)/(100+200)=83.3Conference on Innovations in Travel Analysis and PlanningAnalysis of IE Trip End ChangesConference on Innovations in Travel Analysis an

21、d PlanningTrip End Allocation Unable to predict how many eclipse travelers will come to Ohio Too many unknowns Many distant population centers have viewing areas closer than Ohio Data are not suitable for creating a robust trip generation model Opting for an allocation of a control total approach Us

22、ing analysis of the data to understand trip end trends and allocate accordingly Allocate trip ends using a parabolic curve to capture the general trends observed in the data for IE/EI trips Ratio based suppression of II tripsConference on Innovations in Travel Analysis and PlanningDiurnal Distributi

23、on of Trips Observed changes in diurnal distribution between regular and eclipse days.Changes most notable in IE and EI trips.Compared KY/TN distributions to Ohio model distributions.Applied Ohio average diurnal distributions reported in SDE manual to Ohio model periods for passenger vehicles.Applie

24、d NCHRP 765 diurnal distributions for trucks to Ohio model periods for trucks.Used ratio of changes between regular and eclipse days in KY/TN to create Ohio eclipse diurnal distributions.Needed to shift the changes to account for the fact that 2024 totality will occur later in the day than 2017 tota

25、lityConference on Innovations in Travel Analysis and PlanningDiurnal Distribution of II TripsConference on Innovations in Travel Analysis and PlanningTotalityTotalityDiurnal Distribution of IE TripsConference on Innovations in Travel Analysis and PlanningTotalityTotalityDiurnal Distribution of EI Tr

26、ipsConference on Innovations in Travel Analysis and PlanningTotalityTotalityDiurnal Distribution of EE TripsConference on Innovations in Travel Analysis and PlanningTotalityTotalityTruck Trips Also looked at truck trips.Trip lengths barely changed between regular and eclipse days.Some changes to diu

27、rnal distributions.Drivers scheduling their mandated rest to avoid eclipse traffic?Decided not to alter truck demand for eclipse model but will shift diurnal distribution for model assignment.Conference on Innovations in Travel Analysis and PlanningEclipse Assignment Created a sequential static assi

28、gnment.24 one-hour assignments Volume in excess of capacity is carried over to the next hour Can measure anticipated congestion hot-spots,but cannot address operational elements such as queuing Best analyzed by comparing changes to congestion between a“regular”day and the eclipse scenarioConference

29、on Innovations in Travel Analysis and PlanningNext Steps Continue to refine assignment process Develop scenarios for testing eclipse day impacts Run and analyze scenarios Report findingsConference on Innovations in Travel Analysis and PlanningQuestions?Conference on Innovations in Travel Analysis and Planning