Agent-based simulator of dynamic flood-people interactions

Georges Kesserwani; Mohammad Shirvani; Paul Richmond

REVIEW 4 major objections 6 minor 7 references

Agent-based simulator of dynamic flood-people interactions

T0 review · 4 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A simulator in which floodwater and pedestrians exchange information shows that crowds can change the flood risk they face, and that the same tool can size sandbag-barrier crews.

desk verdict A real GPU-coupled flood-pedestrian simulator whose headline 'dramatic' feedback effect rests on an uncalibrated roughness multiplier; useful as an engineering demonstration, not as evidence. read the letter →

arxiv 1908.05232 v2 pith:QCUFMLJF submitted 2019-08-08 physics.soc-ph cs.CE

Mohammad Shirvani , Georges Kesserwani , Paul Richmond This is my paper · ORCID

classification physics.soc-phcs.CE

keywords two-wayflood-peoplecouplingagent-basedmodelshallowwaterequationsfloodevacuationpedestriandynamicshazardratesandbagbarrierGPUsimulation

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes a computer simulator in which floodwater and people are both represented as interacting agents, so that the flood changes how people move and people in turn change the flood's local depth and speed. The authors argue that such two-way coupling matters: in a synthetic flood of a crowded shopping centre, letting evacuees affect local hydrodynamics through increased terrain roughness changed the flood-risk states of the evacuees themselves. They also show that the same simulator can be used to plan a pre-flood sandbag barrier, estimating how many responders and how many sandbag layers are needed to keep downstream water risk low. If the coupling rules are right, flood evacuation planning should treat crowds not merely as passive receptors but as active modifiers of floodwater.

What carries the argument

The mechanism that carries the argument is a grid of navigation agents sandwiched between the flood agents and the pedestrian agents. Each navigation agent receives a flood message, converts depth and velocity into a Hazard Rate, sends that rate to the pedestrians above it, and then receives messages from those pedestrians about how many are present or what sandbagging action they are performing. It updates local terrain parameters---$n_M = n_M + N_p n_M$ for evacuee crowding, or an increment of $z$ by one sandbag thickness for responders---and sends the updated terrain back to the flood agent. This intermediate grid makes the two-way feedback local, simultaneous, and independent of the numerical scheme, which is a first-order finite-volume solver of the shallow water equations with wetting and drying.

What would settle it

A direct test would be a flume experiment measuring water depth and velocity in a shallow current with and without a stationary or moving crowd, compared with the roughness augmentation $n_M = n_M + N_p n_M$; if measured depths and velocities do not follow that relation, the coupling's effect on evacuee risk states is not supported. A second check would measure actual human walking speeds across the HR ranges of the risk table.

Watch

Extended reading notes

Core claim

The central claim is that a single simulator can capture the two-way interaction between flood dynamics and pedestrian behaviour, and that this interaction materially changes predicted risk. On the flood-to-people side, each pedestrian receives a Hazard Rate $HR = (V + 0.5) \times h$ from the water and moves with a state-dependent speed: brisk walk below HR 0.75, slow walk below 1.5, slower below 2.5, and no walking above. On the people-to-flood side, groups of evacuees locally increase the Manning roughness $n_M$ by an amount proportional to the number of evacuees present, while responders progressively raise terrain height by dropping layers of sandbags. In the shopping-centre demonstration, ignoring the evacuee roughness feedback changes the spatial pattern of high- and medium-risk states: crowds in high-HR zones appear to raise the hazard around them, while crowds in medium-HR zones appear to reduce it. In the intervention scenario, a three-layer (0.75 m) sandbag barrier reached with at least 100 responders is predicted to be sufficient and safe within the 12-hour warning window.

Load-bearing premise

The whole two-way effect rests on uncalibrated rules: that each evacuee adds one unit of Manning roughness to the local terrain and that the flood-risk walking speeds follow the tabulated Hazard Rate thresholds; if the true hydrodynamic effect of a crowd is different, the paper's main finding that people dramatically affect flood impact would not hold.

Editorial extensions

If this is right

Evacuation models that ignore two-way coupling may misstate where and when pedestrians become trapped, because crowd density itself can shift local hazard.
The roughness feedback can either amplify or damp local risk depending on state: gathering in high-HR zones worsens surroundings, while gathering in medium-HR zones appears to buffer them.
The simulator gives concrete planning numbers for sandbag barriers, including deployment time, barrier height, and residual risk reduction, such as a 91.2% drop in maximum HR after one layer and negligible gain beyond three layers.
The method is general enough to be re-run on real urban sites once inflow conditions and exit locations are specified.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the crowd-roughness relation were calibrated empirically, the same coupling architecture could be applied to other crowded settings, such as stadiums or transport hubs, where gathering changes inundation paths.
The directional effect of crowds may depend on flow regime: roughness that slows water could raise depth upstream of a crowd and lower speed downstream, which is a testable prediction of the model.
A sensitivity analysis varying the exponent or prefactor of the roughness feedback would reveal how much of the reported risk-state change is due to the assumed rule rather than to the hydrodynamics.
The responder results suggest that beyond a saturation point, extra manpower no longer shortens deployment time, a pattern worth testing against real sandbagging logistics.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

You should know two things about this paper. First, it is one of the few flood evacuation models that actually couples people back into the hydrodynamics: evacuees increase Manning's roughness and responders raise terrain with sandbags. Second, the 'dramatic' effect of evacuees on flood risk states reported in Scenario 1 is largely a consequence of the assumed roughness rule nM = nM + Np*nM, not of any measured physics. With baseline nM = 0.01, one pedestrian doubles local roughness and the 20-agent cap gives a 21-fold increase; because Manning resistance scales with n^2, the coupling is very strong. The paper flags this rule as assumed (footnote 1) and calls for validation in the conclusions, so the limitation is explicit, but the headline result remains a test of that rule rather than a validated phenomenon.

What is genuinely new: the two-way coupling itself, implemented on FLAMEGPU with discrete flood agents and continuous pedestrian agents communicating through navigation agents. Previous ABMs like FloodPEDS and LSM treat flood as one-way forcing. The hydrodynamic solver is a standard finite-volume SWE scheme, verified against two dam-break tests (results not shown, which is a minor but real gap). The sandbag deployment scenario is more solid: terrain-height increments are a direct and physical way to represent barriers, and the 36-run manpower/thickness sweep gives practically useful outputs like deployment times and HR reduction.

The soft spots are real but not fatal to the paper's core purpose. The roughness feedback is uncalibrated and no sensitivity analysis is provided, so the quantitative differences in Figures 6–7 (5–8% wider high-HR area, 25% wider low-HR area) could shrink or reverse under a weaker or density-dependent crowd-resistance model. The walking speed table is also assumed, but that's a common practice in this literature. The case study is entirely synthetic, which is fine for demonstration but not for prediction. The paper is honest about all of this.

In sum: this is a useful methods contribution for anyone working on coupled flood-human simulation, especially on GPUs. It is not a validated hazard assessment tool. I would send it to peer review because the integration is novel and the authors have made the software and data available, but the revision must include a sensitivity analysis of the roughness feedback and a softened interpretation of Scenario 1. If the editors want a desk decision, conditional acceptance with major revision seems right.

For my own work: I probably wouldn't cite it as evidence for people-flood feedback, but I might cite it as an implementation example of two-way ABM coupling. I'd bring it to a reading group only if the topic was ABM coupling rather than flood risk. Serious thinker: yes, the reasoning is clear and the limitations are acknowledged.

Referee Report

4 major / 6 minor

Summary. The paper presents a coupled agent-based simulator that links a hydrodynamic ABM, based on an explicit finite-volume solver of the shallow water equations, with a pedestrian ABM built on social-force walking rules, all implemented on the FLAMEGPU GPU platform. Two-way coupling is achieved via intermediate navigation agents: evacuees alter local Manning roughness as a function of their local count, while responders construct sandbag barriers by incrementing terrain height. The simulator is demonstrated on a synthetic shopping-centre case with two scenarios: evacuation during a flood and pre-flood sandbag deployment. The authors report that evacuee feedback can 'dramatically affect' flood risk states, and that the responder simulations can guide decisions on manpower, barrier height, and risk reduction.

Significance. If the results hold, the paper offers a useful methodological advance: a GPU-accelerated, two-way coupled flood-pedestrian simulation framework, which goes beyond the one-way coupling used in most existing flood-evacuation ABMs. The authors are transparent about the assumptions underlying their behavioural rules, they provide software accessibility information, and the hydrodynamic solver is based on a published and well-established scheme. However, the central quantitative claim about evacuees 'dramatically' affecting flood risk states rests on an uncalibrated roughness feedback rule, and the claimed verification of the hydrodynamic component is not shown. These issues prevent the results from being taken as evidence of a validated interaction mechanism, though the framework itself is a plausible and potentially valuable building block.

major comments (4)

[Section 2.4, Eq. (nM = nM + Np*nM)] The local terrain-roughness feedback rule nM = nM + Np*nM is the sole mechanism by which evacuee presence affects flood hydrodynamics in Scenario 1, and Figures 6-7 attribute the 'dramatic' differences in evacuee risk states to this rule. However, the rule is explicitly assumed (footnote 1) and no sensitivity analysis or calibration is provided. Because Manning friction enters the momentum equation through a term that scales with n^2, the cap of Np=20 raises the local resistance by a factor of up to 441 relative to the base value, so the reported 5-8% and 25% differences in hazard-rate areas test the prescribed forcing rather than an empirically supported crowd-resistance effect. The central claim of Section 3.2 therefore needs either a sensitivity sweep over the feedback parameter, a physically motivated alternative formulation, or a more guarded wording that distinguishes the demonstration from a validated finding.
[Section 2.2] The verification of the hydrodynamic ABM is reported only in one sentence ('the hydrodynamic ABM (results not shown) reproduced the same predictions as the sequential counterpart'), without any comparison plot, error metric, or benchmark table. Since the credibility of the coupled simulator's flood physics depends on this solver, the authors should include the verification results, for example as an appendix with comparisons to the cited dam-break tests of Wang et al. (2011) and Huang et al. (2013).
[Section 3.2 and Table 1] The walking speed states (1.8, 0.9, 0.45, and 0.0 m/s) assigned to the four HR ranges are assumed on the basis of the UK Environment Agency thresholds, as acknowledged in footnote 1, but the speed values themselves are not derived from empirical data and no sensitivity analysis is reported. The evacuation-time and risk-state statistics in Figures 6-7 may be sensitive to these choices, so the authors should either cite empirical studies of walking speed in floodwater or demonstrate that the qualitative conclusions are robust across a plausible range of speed values.
[Section 3.3] The sandbag pickup and drop-off times are stated as 'half a minute' and described as 'specified' without a reference, yet the deployment-time estimates in Figure 8 and the resulting manpower recommendations depend directly on these values. The authors should cite a source for these action times, or perform a sensitivity analysis to show that the conclusion that a 10% responder group can deploy a three-layer barrier safely within 12 hours is robust to plausible variations in handling times.

minor comments (6)

[Section 3.3 (text after Figure 9)] The phrase 'tree-layer thickness' appears to be a typo for 'three-layer thickness'.
[Section 3.3 (caption of Figure 10)] The phrase 'in line with increased thickness' should be 'in line with increased thickness' or, more idiomatically, 'as the thickness increases'.
[Section 2.1] The phrase 'like-for-like distribution' is vague; it should be clarified that the navigation and flood grids are collocated with identical cell sizes and alignments.
[Section 3.1, Figure 5(b)] The sentence beginning 'With the hydrographs associated with...' is difficult to follow because it lists three conditions together; splitting it into separate statements would improve readability.
[Abstract and Section 2.4] The abstract refers to a 'terrain roughness feedback factor' but this term is not defined in the methodology; the authors should either introduce this term explicitly in Section 2.4 or avoid it in the abstract.
[Software accessibility] The software and data links (DAFNI and the project website) would benefit from a persistent identifier, such as a DOI or versioned release, to support reproducibility.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the two-way coupling is implemented from explicit assumptions rather than fitted to outputs.

full rationale

The paper does not reduce any of its conclusions to its inputs by construction. The crowd-roughness feedback nM = nM + Np*nM, the walking-speed reductions in Table 1, and the sandbag-deployment procedures are all prescribed modelling assumptions, not parameters calibrated to the simulated results. The HR thresholds are taken from the UK Environment Agency and are used consistently as both input labels and output metrics, which is a functional definition rather than a circular derivation. The central claim that evacuees can 'dramatically affect' flood impact on their own states is a simulation outcome produced by the explicitly stated feedback rule; the rule itself is not derived from that outcome, and the paper openly flags the assumptions as assumptions and lists validation as future work. Self-citations to FLAMEGPU and earlier pedestrian/hydrodynamic implementations support the software platform and numerical components, but the paper's flood-people interaction claim does not depend on any self-cited uniqueness theorem or imported ansatz. The hydrodynamic solver is cross-verified against independent dam-break test cases. The main weakness is that the roughness-feedback rule is uncalibrated and potentially strong, but that is a correctness and validation risk, not circularity.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central claim rests on assumed behavioral rules (roughness feedback, walking speed reductions) and standard domain models (SWE, social force). The free parameters are chosen or assumed, not fitted to the target results. No new physical entities are introduced; the navigation agents are computational constructs rather than physical postulates.

free parameters (4)

Walking speed states in floodwater = 1.8, 0.9, 0.45, 0.0 m/s for low, medium, high, highest hazard
Assumed in Section 2.3 and Table 1, acknowledged in footnote 1; not empirically calibrated, affect evacuation timing and risk states.
Terrain roughness feedback factor = nM = nM + Np*nM, max Np = 20 so nM max about 0.21
Ad hoc rule in Section 2.4 to represent crowd presence; no experimental basis, central to the 'dramatic' finding.
Sandbag pickup and drop-off time = 0.5 minutes each
Specified in Section 2.4 without justification; directly controls deployment time in Scenario 2.
Initial Manning's n = 0.01
Represents clear cement (Chow 1959), but is a chosen input affecting flood propagation.

assumptions (4)

domain assumption Shallow water equations with Manning friction govern flood dynamics
Used in Section 2.2; standard for flood modeling, but an approximation.
domain assumption Social force model describes pedestrian movement and evacuation
Adopted from Helbing and Molnar 1995; valid for dry domains, extended here to floodwater via assumed speed reductions.
domain assumption Hazard Rate HR = (V + 0.5) * h maps to risk states
From UK Environment Agency (2006) and Kvočka et al. 2016; used as input in Table 1.
standard math The finite volume scheme is first-order accurate and stable under CFL
Section 2.2; standard scheme from Wang et al. 2011.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Agent-based simulator of dynamic flood-people interactions." pith.science (2026). https://pith.science/paper/QCUFMLJF

@misc{pith2026190805232,
  author       = {Pith},
  title        = {Pith review of: Agent-based simulator of dynamic flood-people interactions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QCUFMLJF}},
  note         = {Machine review of arXiv:1908.05232}
}

read the original abstract

This paper presents a new simulator for dynamic modelling of interactions between flooding and people in crowded areas. The simulator is developed in FLAMEGPU (a Flexible Large scale Agent-based Modelling Environment for the GPU), which allows to model multiple agent interactions while benefitting from the speed-up of GPUs. Flooding variables including terrain data are represented by a hydrodynamic Agent-Based Model (ABM) that is based on a non-sequential implementation of a robust Finite Volume (FV) solver of the Shallow Water Equations (SWEs). People movements are represented by a pedestrian ABM adopting force-based walking rules. The hydrodynamic ABM is coupled to the pedestrian ABM according to risk-to-life thresholds reported by the UK Environment Agency (EA). A hypothetical case study of a crowded shopping centre is proposed and used to assess the dynamic coupling ability of the simulator. Flooding into the shopping centre is induced based on realistic inflow conditions, and the simulator is applied considering two scenarios: evacuation without advanced warning and intervention with an advanced warning of 12 hours. Results show that the simulator can produce detailed statistics of spatiotemporal people states during evacuation, and is useful to plan safe and effective people intervention to deploy a sandbag-based temporary barriers.

Figures

Figures reproduced from arXiv: 1908.05232 by the authors.

**Figure 1.** The process for generating and running an agent-based simulation program on FLAMEGPU (http://www.flamegpu.com/home). 2. Methodology FLAMEGPU is a platform for the simulation of multiple agent interactions on CUDA Cores of a Graphical Processing Unit (GPU) (Richmond et al. 2009; Chimeh & Richmond 2018). It allows to create and run a CUDA simulation program by processing three inputs as shown in [PITH_FULL_IMAGE:figu… view at source ↗

**Figure 2.** A flood agent (‘dark blue’) elevating its states of floodwater variables, U, from time iteration n to n + 1. The process is done simultaneously for all flood agents, facilitated by the messages (‘white message icons’) the flood agent receives to access the states of floodwater variables of its four neighbours. An explicit shock-capturing scheme is selected (Wang et al. 2011), and its formulation is reduced to first-… view at source ↗

**Figure 3.** Dynamic messaging between a flood agent and pedestrian agents (evacuees) facilitated by the navigation agent located at the same position (left). Messaging procedure for pedestrian agents (responders) deploying a sandbag barrier (right): red navigation agent represents a ‘sandbag storage’ destination and grey navigation agents represent the deployment destination (Section 2.4). 2.4 Interactions between the flood, na… view at source ↗

Figures from the paper (7 more)

**Figure 4.** Figure 4: Schematic description of the hypothetical shopping centre (Section 3) with the two scenarios: (a) ‘during-flood evacuation’ and (b) ‘pre-flood intervention’. 3. Demonstration on a synthetic case study A synthetic test is proposed to evaluate the coupled ABMs for modell…

**Figure 5.** Figure 5: Inflow hydrographs, panel (a), and flood conditions defined by four different (Qpeak, tinflow) pairing. The flooding conditions are defined by fixing the volume of water that will be released while proportionally doubling the discharge peak (Qpeak) and halving the dura…

**Figure 6.** Figure 6: A stack chart illustrating the ‘flood risk states’ ( [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]

**Figure 7.** Figure 7: Spatial flood maps alongside the distribution of evacuees at (a) t = 3.6 min and (b) t = 6.3 min: Left and right columns contain the plots produced by the run ‘without’ and ‘with’ the effects of people on local flood hydrodynamics, respectively. Time histories of the o…

**Figure 8.** Figure 8: Simulated times vs. responders’ group size for deploying up to six-layer thick (sandbag) barrier: ‘red line’ indicates flooding start time below which is safe to deploy (area shaded in ‘green’) or otherwise unsafe (area shaded in ‘red’). This aspect is more closely exp…

**Figure 9.** Figure 9: Centrelines of 2D water depth maps along y-axis after the deployment the sandbag barrier (red dashed line) considering up to six layers of sandbag thickness. 3.3 Simulation of Scenario 2 (pre-flood intervention) The coupled ABMs simulator is re-applied for Scenario 2, …

**Figure 10.** Figure 10: Cumulative percentage of maximum HR reduction in line with increased thickness of the barrier in terms of number of sandbag layers. Therefore, [PITH_FULL_IMAGE:figures/full_fig_p019_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

7 extracted references · 7 canonical work pages

[1]

Abebe, Y.A., Ghorbani, A., Nikolic, I., Vojinovic, Z., & Sanchez, A. (2019). A coupled flood- agent-institution modelling (CLAIM) framework for urban flood risk anagement. Environmental modelling & software, 111, 483-492. Aerts, J.C., Botzen, W.J., Clarke, K.C., Cutter, S.L., Hall, J.W., Merz, B., Michel-Kerjan, E., Mysiak, J., Surminski, S., & Kunreuther...

work page 2019
[31]

Richmond, P., Coakley, S., & Romano, D. M. (2009). A high performance agent based modelling framework on graphics card hardware with CUDA. In Proceedings of the 8th International Conference on Autonomous Agents and Multiagent Systems, pp. 1125-1126. 10-15 May 2009, Budapest, Hungary. 26 Sen Nag, O. (2018). The Largest Shopping Malls in the United Kingdom ...

work page 2009
[77]

M., Simm, J

Lumbroso, D. M., Simm, J. D., Davison, M., White, K. D., & Durden, S. (2015). Use of agent- based modelling to validate hurricane evacuation planning. In Coastal structures and solutions to coastal disasters joint conference (COPRI), pp. 321-329. 9–11 September 2015, Boston, USA. Lumbroso, D., & Davison, M. (2018). Use of an agent‐based model and Monte Ca...

work page 2015
[193]

S., Taylor, H

Becker, J. S., Taylor, H. L., Doody, B. J., Wright, K. C., Gruntfest, E., & Webber, D. (2015). A review of people’s behavior in and around floodwater. Weather, Climate, and Society, 7(4), 321-

work page 2015
[275]

Liu, X., & Lim, S. (2016). Integration of spatial analysis and an agent-based model into evacuation management for shelter assignment and routing. Journal of spatial science, 61(2), 283-298. Liu, Y., Okada, N., Shen, D., & Li, S. (2009). Agent-based flood evacuation simulation of life- threatening conditions using vitae system model. Journal of Natural Di...

work page 2016
[332]

22 Bernardini, G., Postacchini, M., Quagliarini, E., Brocchini, M., Cianca, C., & D'Orazio, M. (2017). A preliminary combined simulation tool for the risk assessment of pedestrians’ flood-induced evacuation. Environmental modelling & software, 96, 14-29. Bernardini, G., Quagliarini, E., D'Orazio, M., & Brocchini, M. (2020). Towards the simulation of flood...

work page 2017
[4282]

Helbing, D., Farkas, I., & Vicsek, T. (2000). Simulating dynamical features of escape panic. Nature, 407(6803), 487-490. Hellevang, K. (2011). Sandbagging for flood protection AE-626 (Revised). National Institute of Food and Agriculture, U.S. Department of Agriculture, Agreement No. 2009-41210-05965. Huang, Y., Zhang, N., & Pei, Y. (2013). Well-balanced f...

work page 2000

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Abebe, Y.A., Ghorbani, A., Nikolic, I., Vojinovic, Z., & Sanchez, A. (2019). A coupled flood- agent-institution modelling (CLAIM) framework for urban flood risk anagement. Environmental modelling & software, 111, 483-492. Aerts, J.C., Botzen, W.J., Clarke, K.C., Cutter, S.L., Hall, J.W., Merz, B., Michel-Kerjan, E., Mysiak, J., Surminski, S., & Kunreuther...

work page 2019

[2] [31]

Richmond, P., Coakley, S., & Romano, D. M. (2009). A high performance agent based modelling framework on graphics card hardware with CUDA. In Proceedings of the 8th International Conference on Autonomous Agents and Multiagent Systems, pp. 1125-1126. 10-15 May 2009, Budapest, Hungary. 26 Sen Nag, O. (2018). The Largest Shopping Malls in the United Kingdom ...

work page 2009

[3] [77]

M., Simm, J

Lumbroso, D. M., Simm, J. D., Davison, M., White, K. D., & Durden, S. (2015). Use of agent- based modelling to validate hurricane evacuation planning. In Coastal structures and solutions to coastal disasters joint conference (COPRI), pp. 321-329. 9–11 September 2015, Boston, USA. Lumbroso, D., & Davison, M. (2018). Use of an agent‐based model and Monte Ca...

work page 2015

[4] [193]

S., Taylor, H

Becker, J. S., Taylor, H. L., Doody, B. J., Wright, K. C., Gruntfest, E., & Webber, D. (2015). A review of people’s behavior in and around floodwater. Weather, Climate, and Society, 7(4), 321-

work page 2015

[5] [275]

Liu, X., & Lim, S. (2016). Integration of spatial analysis and an agent-based model into evacuation management for shelter assignment and routing. Journal of spatial science, 61(2), 283-298. Liu, Y., Okada, N., Shen, D., & Li, S. (2009). Agent-based flood evacuation simulation of life- threatening conditions using vitae system model. Journal of Natural Di...

work page 2016

[6] [332]

22 Bernardini, G., Postacchini, M., Quagliarini, E., Brocchini, M., Cianca, C., & D'Orazio, M. (2017). A preliminary combined simulation tool for the risk assessment of pedestrians’ flood-induced evacuation. Environmental modelling & software, 96, 14-29. Bernardini, G., Quagliarini, E., D'Orazio, M., & Brocchini, M. (2020). Towards the simulation of flood...

work page 2017

[7] [4282]

Helbing, D., Farkas, I., & Vicsek, T. (2000). Simulating dynamical features of escape panic. Nature, 407(6803), 487-490. Hellevang, K. (2011). Sandbagging for flood protection AE-626 (Revised). National Institute of Food and Agriculture, U.S. Department of Agriculture, Agreement No. 2009-41210-05965. Huang, Y., Zhang, N., & Pei, Y. (2013). Well-balanced f...

work page 2000