Project Meeting 2017.05.04 - bellevuewa/BKRCast GitHub Wiki

Calibration / Validation Setup

• Latest model run summarized for calibration/validation summaries
• Model run uses the 400% BKR sampling strategy
• Summarized SoundCast, BKR with LODES, and BKR with ESD
• Most summaries are comparing results for the entire PSRC region; not just BKR

General Results

• Synthetic population matches well, except for too many univ students
• Maybe # of univ students is ok within BKR?
• Tours by purpose looks ok; work-based is high (and for SoundCast as well)
• Tours by person type looks ok
• Trips by purpose needs some work
• There is a trip length issues (see below)
• Auto ownership and sufficiency look good
• Mode shares look good
• Transit is low but that's probably acceptable due to the survey being high, according to PSRC and Sound Transit
• Time-of-day patterns look good

Trip lengths

• There is not enough 3-6 mile trips and too many 6-12 mile trips for work, school, and other
• This may be due to large zones outside BKR
• We need summaries for within and outside BKR
• Initial review suggests this is only outside BKR, but we need to better understand

Next Steps

• Summarize highway assignment screenlines and transit boardings by line
• @Stefan to provide broadings to COB and COB to join to the BKRCast networks
• Investigate trip length issue
• Split summaries into BKR/non-BKR
• Summarize trip lengths by mode
• Review terminal time settings - BKR are longer in CBD areas, maybe this is part of the issue?