Mid-South Regional Travel Surveys & Model Update

Transcription

1 Mid-South Regional Travel Surveys & Model Update July 23, 2014 Presented to the Mid-South Travel Survey and Model Update Steering Committee By: Thomas Rossi, Anurag Komanduri, Daniel Beagan, and Brent Selby, Cambridge Systematics Subconsultants: Abt SRBI, Dikita Engineering and Neel-Schaffer

2 Introductions Meeting Agenda Household survey status Bike-ped data update Freight data and analysis approach Modeling, data and zones Transit survey status update Model development general plan Project schedule/status

3 Introductions Introductions of Steering Committee Consultant Team Cambridge Systematics prime, modeling lead & survey analysis Dikita Engineering transit onboard survey lead Abt SRBI household survey lead Neel-Schaffer support on freight and non-motorized surveys

4 Household Travel Survey Data retrieval complete Data delivered to CS Nearly 4,900 households Corrective actions to maximize completes from H-T-R households Diary data undergoing QAQC Identify usable records Prepare for survey expansion Use expanded data in model estimation

5 Household Travel Survey Final survey dataset provided by SRBI CS staff performing QA/QC 100 standard checks Preliminary data analysis routines Critical Description Frequency % Previous % Status 1 Missing Geocodes, Home 0 0% 0 0% CLEAN 1 Sign Inconsistency in Geocodes, home 0 0% 0 0% CLEAN 1 Geocodes Out of the Region, home 0 0% 1 0% CLEANED 0 Length HHID must be constant. 0 0% 0 0% CLEAN 0 Number of Workers > HH Members % 0 0% CLEAN 1 HH Missing Income 297 6% 769 8% To be imputed 0 HH Has Only Over/Under $50K Income 313 6% 659 7% To be imputed 1 HH ID in HH File Not in Person file 0 0% 0 0% CLEAN 1 Number of People in Person File Must Equal HHsize 1 0% 0 0% To be corrected/dropped 1 HHs in the HH File Not in Trip File 0 0% 0 0% CLEAN 0 Number of Spouses > 1 3 0% 2 0% To be corrected 1 Trips by HH in Trip File not equal to HHTRIPS 0 0% 0 0% CLEAN 1 HH not retrieved % % Improved

6 Household Travel Survey Critical Description Frequency % Previous % Status 1 Missing Geocode, origin % % To investigate 1 Missing Geocode, destination % % To investigate 1 Sign Inconsistency in Geocode, origin 1 0.0% 1 0.0% To be dropped 1 Sign Inconsistency in Geocode, destination 0 0.0% 0 0.0% CLEAN 0 Geocode Out of the Region, origin % % To investigate 0 Geocode Out of the Region, destination % % To investigate 1 All HHs in the Trip File Should be in HH File 0 0.0% 0 0.0% CLEAN 1 Duplicate Trip Record 0 0.0% 0 0.0% CLEAN 0 Final Trip Record not Marked as Last 0 0.0% 0 0.0% CLEAN 0 Non-traveler not at home % 0 0.0% To investigate 1 Name/Origin Activity/Type Don't Match Previous Desti 0 0.0% 0 0.0% CLEAN 1 Origin Lat/Long Don't Match Previous Destination's 0 0.0% % CLEANED 1 Same O or D (Name & Address) as Previous Record % % To investigate 1 O and D (Name and Address) are the Same % % To investigate 0 O and D are Both Home Activity % % To investigate 0 Day Does not Start at Home or Work % % To investigate 0 Day Does not End at Home or Work % 0 0.0% To investigate 1 Missing Time Information % % Big improvement 1 Arrival Time is After Departure Time 0 0.0% 0 0.0% CLEAN 0 Unusual Activity Durations % % To investigate 0 Redundant Reports of Non-Transfer Modes 0 0.0% 0 0.0% CLEAN 0 Child is listed as driving 0 0.0% 0 0.0% CLEAN 1 Origin Matches Destination by Lat/Long % % Improved

7 Household Travel Survey GPS analysis underway 50 household sample Convert GPS traces into stops Differentiate between trip ends and traffic stops Use speed measures to identify stops Home Location Diary vs. GPS comparison Are short trips under-represented in diary survey? Is the GPS database missing some trips? Can we identify mode and purpose? How can we best use land-use files? Data processing heuristics being developed

8 Bike-Ped Data Bike-ped counts completed by Neel-Schaffer in April Information from all locations processed in early July QA/QC underway likely use in validation of model

9 Online bike-ped survey monitored by Abt-SRBI HTS Recruits providing addresses (~5400) Choice sample provided by MPO Data collection closed in mid-june 1,150+ online completes 650 from MPO mailing list 500 from HTS participants Update bike-ped plan Support model validation Bike-Ped Data

10 Freight Data Model-related freight data provided by TDOT Preliminary TRANSEARCH data from IHS/Global Insight Processed Truck GPS data from ATRI Finalizing employment database by TAZ Model development will commence once emp data are approved Travel Market I-I I-E E-I Type of Truck Service Freight Freight Dataset Used ATRI (All trucks included) TRANSEARCH (Only freight trucks included) E-E Freight TRANSEARCH (Only freight trucks included)

11 Modeling Data Finalized TAZ system Input from MDoT & ADoT Input from West Memphis MPO Consistent with ACS boundaries Shared with TDoT for approval 1,729 zones in total Updating highway and transit networks 2010 base year Socioeconomic data being finalized ACS data population and household InfoGroup data employment by NAICS Population Density Worker Density

12 Transit Onboard Survey Survey expansion complete Usable record: 3,227 Expanded totals: 28,781 Avg. Expansion Weight: 8.82 Developing analytical reports Identify key variables to highlight Support modeling + Title VI Data visualization techniques TripPurp AM Early AM Peak Midday PM Peak PM Late JTW 81.7% 53.1% 41.7% 54.1% 46.1% HBO 4.6% 17.9% 24.9% 13.2% 15.7% HBSch 9.0% 8.5% 1.4% 6.5% 1.2% HBU 1.7% 9.1% 7.6% 7.2% 3.1% NHB-W 2.3% 4.0% 7.2% 5.9% 12.0% HBShop 0.0% 2.2% 6.7% 3.7% 4.9% HBSocRec 0.4% 1.9% 3.4% 4.7% 8.1% NHB-NW 0.0% 3.2% 7.0% 4.5% 8.9% HBAir 0.3% 0.3% 0.2% 0.1%

13 Model Development Plan Model Component Description Data Source Vehicle Availability Household level logit model Household survey Internal Trip Productions Cross-classification by trip purpose Household survey Internal Trip Attractions Linear regression by trip purpose Household survey Special Generators Asserted totals Outside data Journey to Work Stops Multinomial logit model Household survey EE and EI Trip Generation Linear regression model Household survey Destination Choice Multinomial logit model Household survey Intermediate Stop Destination Choice Multinomial logit model Household survey Time of day Fixed factors by trip purpose Household survey Mode choice Truck trip generation Nested logit model Linear regression model Household survey Transit survey TRANSEARCH + ATRI Freight survey Truck trip distribution Gravity model TRANSEARCH + ATRI

14 Next Steps HH survey processing QA/QC, expansion, analysis GPS processing QA/QC, research findings Transit survey processing Title VI, modeling Model networks QA/QC, skims for model estimation, 2010 transit network, bike network Model input data employment and demographics review TDOT approval

15 1. Project Management 2. Travel Demand Model Consultation 3. Research Design, Pretest, and Refinement 4. Household Travel Survey Data Collection 5. Freight Survey Task 6. Transit Onboard Survey 7. Bicycle and Pedestrian Survey 8. Data Weighting and Expansion 9. Travel Surveys Final Report/ Data Delivery 10.Travel Demand Model Update Jun 1 Approximate Month (revised) Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Interim Findings Deliverable Project Oversight Steering Committee Meetings 15

16 Questions/Comments and Discussion