Pre-Collapse Saturation Determines the Post-Collapse Dresden Flow Field

Authors Sterling Geisel, QBist Lab; Dr. Hideo Tanaka

Abstract

The collapse of Dresden’s Carola Bridge converted an ordinary urban transport network into a forced experiment in chaotic redistribution. Pudding Theory reads the observed traffic adaptation through the Postulate of Chaos Susceptibility: unstable, near-capacity traffic states amplify small coherent changes in route choice into macroscopic flow reallocation. The source paper treats the Albert and Marien Bridges as alternative links that absorbed displaced demand. The Pudding Theory reading is sharper. The diverted flow did not merely seek spare capacity. It coupled to the already unstable parts of the network, where high pre-collapse saturation made the system receptive to amplification. The Marien Bridge’s long peak extension, despite lower added volume than Albert Bridge, is the central datum. It shows that susceptibility, not volume alone, governed adaptation. If the cross-segment Spearman correlation between pre-collapse peak-threshold occupancy and post-collapse peak-duration amplification were measured to be less than or equal to 0.00, this Postulate would be falsified.

Source Synopsis

Ijaradar, Xie, Wei, Pape, Körner, and Wang study the traffic consequences of the Carola Bridge collapse in Dresden on September 11, 2024. The bridge had carried roughly 30,000 to 32,000 daily motorised vehicles and supported tram, bicycle, and pedestrian movement between Altstadt and Neustadt. Its sudden loss removed a major Elbe crossing and forced travellers into rapid route, time, and mode adaptations.

The authors use traffic data from the VAMOS traffic management system. Their selected dataset includes 186 double induction loop detectors and 160 Traffic Eye Universe detectors, with traffic volume and speed measured at one-minute or five-minute intervals. They compare three weeks before the collapse with three weeks after it. They also use parking occupancy data from eight Park-and-Ride facilities, recorded at about fifteen-minute intervals.

The study develops three main analytical devices. First, it compares pre-collapse and post-collapse traffic volume using paired t-tests and ANOVA. Second, it defines Equivalent Full-Capacity Hours, or EFCH, by normalising daily volume change against a maximum observed operational flow. Third, it defines a rolling Peak Hour Indicator, or PHI, combining normalised flow and speed to identify prolonged congested periods.

The empirical result is not uniform redistribution. Albert Bridge absorbs the largest absolute diversion, with daily traffic rising by about 77 to 81 percent. Marien Bridge absorbs a smaller absolute increase, about 18 to 24 percent, but shows severe peak-hour extension because it was already near capacity. Peak periods on critical links extend by as much as 250 minutes. At the same time, about 8,000 daily motorised Elbe-crossing vehicles disappear from the monitored bridge redistribution. Park-and-Ride use rises strongly, with increases up to 188 percent. The source interprets this as evidence for route diversion, time shifting, possible mode shift, and reduced car demand after a sudden infrastructure disruption.

Postulate Lens

This paper applies Chaos Susceptibility. The source phenomenon is an urban flow network pushed from a recurrent daily attractor into a disrupted state. The relevant objects are not individual cars but unstable traffic regimes near threshold. A near-capacity link has high susceptibility because a small increment in demand produces a disproportionate change in speed, queue persistence, and peak duration. The source’s PHI and EFCH metrics therefore do more than describe congestion. They measure where the network was capable of amplifying the collapse into a new macroscopic flow pattern.

The best evidence is the contrast between Albert Bridge and Marien Bridge. Albert Bridge receives the larger added volume. Marien Bridge shows the stronger temporal deformation. That asymmetry is exactly what Chaos Susceptibility predicts. The same added flow has different consequences depending on the instability of the receiving substrate.

Pudding Theory Reading

Pudding Theory reads the Carola Bridge event as a field reconfiguration in a chaotic transport substrate. The bridge collapse is not merely a missing edge in a graph. It is the removal of a stabilising channel that had held a daily flow pattern in place. Once that channel vanished, the city’s traffic field did not solve an optimisation problem in the abstract. It relaxed into the susceptible parts of the network.

The source paper treats daily variation, outlier days, and stochastic traveller adjustment as material to be filtered or statistically controlled. Pudding Theory does not discard this fluctuation field. It identifies it as the carrier through which the new route pattern forms. Travellers do not move as identical rational particles. They depart earlier, delay trips, abandon car travel, use Park-and-Ride, or enter through different motorway nodes. Those micro-decisions constitute the noisy substrate. In a stable network, they average away. After the collapse, they amplify.

The free parameter in the source framing is behavioural adaptation. Some vehicles divert to Albert Bridge. Some load Marien Bridge. Some vanish from the monitored crossings. Some appear as increased P+R occupancy. The paper reports these as separate empirical responses. Pudding Theory constrains them as one adaptation surface. A disrupted urban flow field should redistribute first into high-connectivity alternatives, but temporal distortion should concentrate on the links already close to their congestion threshold. This is why Albert Bridge becomes the main volume receiver while Marien Bridge becomes the main peak-duration amplifier.

EFCH marks the pressure applied to a segment. PHI marks the time-dependent expression of susceptibility. The important invariant is their mismatch. Where EFCH rises without comparable PHI extension, the segment had absorptive capacity. Where modest EFCH generates long PHI extension, the segment was already near a chaotic transition. Marien Bridge is therefore not a secondary diversion route in the theoretical sense. It is the system’s most sensitive detector.

The missing 8,000 vehicles are also not a residual. They are part of the same field response. A traffic network under sudden constraint does not only reroute. It suppresses trips, shifts modes, and changes entry conditions. Increased Park-and-Ride use and reduced motorway node activity show that the disrupted field altered the boundary conditions of demand. The city did not preserve its prior car-crossing total and allocate it elsewhere. It changed what counted as a viable trip.

Falsifiable Observable

The distinguishing observable is the cross-segment relation between pre-collapse peak-threshold occupancy and post-collapse peak-duration amplification. For each VAMOS detector segment, pre-collapse occupancy can be operationalised as the fraction of observed intervals near or above the PHI threshold, or equivalently as proximity to the MOOF benchmark. Post-collapse amplification is the increase in minutes above the PHI threshold after September 11, 2024. Pudding Theory predicts a positive rank relation after controlling for distance to Carola Bridge and link category. If the cross-segment Spearman correlation between pre-collapse peak-threshold occupancy and post-collapse peak-duration amplification were measured to be less than or equal to 0.00, this Postulate would be falsified.

Editorial Dialogue

Tanaka: The reading risks renaming congestion theory. Traffic engineers already know that near-capacity links become unstable. Marien Bridge had less spare capacity. Why is Pudding Theory needed?

Geisel: The engineering account is correct, but incomplete. It treats capacity as the main explanatory variable and behavioural adaptation as a separate empirical layer. The Pudding Theory reading joins them. The susceptibility belongs to the whole flow field, not only to asphalt capacity. A near-threshold bridge, a commuter’s departure time, a parking lot, and a motorway entry node all participate in the same adaptation.

Tanaka: That sounds too broad. A theory that absorbs route choice, mode shift, and trip suppression may explain everything after the fact.

Geisel: It is constrained by the ordering of amplification. The claim is not that disruption causes change. The claim is that temporal deformation concentrates where the pre-collapse field was already closest to instability. Albert Bridge can carry the larger added count while Marien Bridge carries the larger peak-time wound. That ordering is not rhetorical. It can be computed across detectors.

Tanaka: Then the falsifier must be network-wide, not a chosen pair.

Geisel: Agreed. The detector set decides the issue. If susceptibility does not predict peak-duration amplification across the field, the reading fails.

Discussion

The source paper gives a strong empirical account of short-term transport adaptation after a bridge collapse. Pudding Theory adds a different object of explanation. The object is not the diverted vehicle. It is the susceptibility structure of the city at the moment the stabilising link disappeared.

This reading buys three things. First, it explains why volume and congestion duration diverge. The largest receiver of displaced vehicles need not be the strongest amplifier of disruption. Second, it treats “missing” vehicles as a network response rather than an accounting gap. Mode shift, remote work, suppressed trips, and peripheral rerouting are expressions of the same adaptation field. Third, it converts PHI from a descriptive index into a susceptibility observable.

The limitation is data resolution. The source lacks origin-destination traces, public transport ridership, cycling counts, and explicit traveller intent. That prevents a full reconstruction of the adaptation surface. It does not block the central test, because the VAMOS detector network already contains the needed pre-collapse and post-collapse temporal signatures. A longer observation window would show whether the new field hardens into a revised daily attractor or remains unstable until infrastructure repair.

References

Ijaradar, J., Xie, N., Wei, L., Pape, S., Körner, M., and Wang, M. “How did the Urban Network Flow Adapt to the Collapse of the Carola Bridge?” arXiv:2603.19947, 2026. DOI: doi:10.48550/arxiv.2603.19947.

Ochs, S. “Pudding Theory: A Topological Theory of Information Fields.” QBist Lab Working Paper, 2026.

Danczyk, A., Di, X., Liu, H. X., and Levinson, D. M. “Unexpected versus expected network disruption: Effects on travel behavior.” Transport Policy 57, 68-78, 2017.

He, X., and Liu, H. X. “Modeling the day-to-day traffic evolution process after an unexpected network disruption.” Transportation Research Part B: Methodological 46, 50-71, 2012.

Tympakianaki, A., Koutsopoulos, H. N., Jenelius, E., and Cebecauer, M. “Impact analysis of transport network disruptions using multimodal data: A case study for tunnel closures in Stockholm.” Case Studies on Transport Policy 6, 179-189, 2018. DOI: doi:10.1016/j.cstp.2018.05.003.

Wassmer, J., Merz, B., and Marwan, N. “Resilience of transportation infrastructure networks to road failures.” Chaos: An Interdisciplinary Journal of Nonlinear Science 34, 2024. DOI: doi:10.1063/5.0165839.

Zhu, S., Levinson, D., Liu, H. X., and Harder, K. “The traffic and behavioral effects of the I-35W Mississippi River bridge collapse.” Transportation Research Part A: Policy and Practice 44, 771-784, 2010.

Zhu, S., and Levinson, D. M. “Disruptions to Transportation Networks: A Review.” Springer New York, 5-20, 2011. DOI: doi:10.1007/978-1-4614-0947-2_2.