Pith. sign in

REVIEW 4 major objections 6 minor 63 references

A Method for Rapid Area Prioritisation in Flood Disaster Response

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read PrioReMap turns a flood polygon into rescue priority zones.

desk verdict The novelty is in the assembly, not the parts, and the unpublished expert CPT is the one thing that has to be fixed before the prioritisation map can be trusted. read the letter →

arxiv 2506.18423 v1 pith:LEPSMA5O submitted 2025-06-23 stat.AP

Moritz Schneider , Lukas Halekotte , Tina Comes , Frank Fiedrich This is my paper · ORCID
classification stat.AP
keywords SpatialPrioritisationFloodManagementDisasterResponseDecisionSupportBayesianNetworkGISRisktoPeople
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

The paper introduces PrioReMap, a method for turning a flood-extent map into a short list of areas that need help first. It claims that a GIS-informed Bayesian network built during the preparedness phase, then fed only with the current flood polygon, can produce rapid and transparent prioritisation during an ongoing flood. Instead of looking at the probability of the most severe risk state alone, the method combines the full four-state probability distribution into one scalar score and clusters the resulting scores into priority classes. A case study for Cologne identifies 202 of 3740 hex tiles as high-priority areas. If this works, responders get a consistent, explainable ranking that reduces cognitive load under time pressure.

What carries the argument

The load-bearing object is the tile-specific Bayesian network whose target node is Risk of People in Need of Assistance, with states None, Low, Medium, and High. Its three direct parents are Density of Exposed Buildings, Presence of Exposed Care Facilities, and Accessibility of Unexposed Areas, where the latter is itself a deterministic combination of Accessibility of Immediate Unexposed Areas and Accessibility of Remote Unexposed Areas. The network is duplicated per hex tile and driven by GIS models: building and care-facility overlays give hard evidence, local flood coverage gives soft evidence for immediate accessibility, and a Dijkstra routing on the flooded road network gives hard evidence for remote accessibility. The engine of the prioritisation is the Probability Distribution Criticality (PDC) score, $PDC=\sum_i w_i P(s_i)$, which folds the entire posterior over the four risk states into a single number; k-means clustering then turns that number into priority classes. All of this is prepared before the flood so that during the event only the flood polygon has to be updated and propagated through the fixed models.

What would settle it

Re-run the Cologne case study with a second set of 96 CPT values elicited from another expert panel under the same three qualitative principles, and compare the High Priority tile sets; a large change in the set indicates the method is not stable to the unstated probabilities. Alternatively, compare the four priority classes against actual rescue-demand records or damage reports from a comparable real flood to see whether the top classes capture where help was really needed.

Watch

Extended reading notes

Core claim

The central claim is that area prioritisation in a flood can be reduced to a fixed Bayesian network plus a scalar weighting rule, so that the response phase needs only a new flood layer and no re-modelling. The target node, Risk of People in Need of Assistance, has four states (None, Low, Medium, High) and is influenced by the density of exposed buildings, exposed care facilities, and accessibility of unexposed areas, split into immediate and remote. Each hex tile runs its own copy of the network, with leaf nodes fed by GIS overlays (buildings and care facilities vs. flood extent, local flood coverage as soft evidence, and road-network routing to safe destinations as hard evidence). The posterior distribution is compressed by the Probability Distribution Criticality value, $PDC=\sum_{i=1}^{4} w_i P(s_i)$ with weights $0$, $1/3$, $2/3$, $1$, and k-means clustering assigns tiles to High Priority, Priority, Exposed, or Safe. The discovery, as the paper states it, is that this pipeline delivers transparent, reproducible recommendations without live damage observations.

Load-bearing premise

The 96 probabilities in the conditional probability table for the target node are expert judgement, not data, and the numeric values are not given; if different experts filled in the same qualitative rules, the priority recommendations could change materially.

Editorial extensions

If this is right

  • Responders can obtain a recommendation from just a flood-extent polygon, because the GIS models and the network are prebuilt and the only runtime input is the current flood layer.
  • Using the full probability distribution means two tiles with the same High-risk probability can still be ranked differently if their remaining probability mass sits in different states, as the paper's motivating example shows for Medium versus Low.
  • High-priority areas can appear away from the river's course: the Cologne case study finds a nine-tile high-priority cluster roughly 2.5 km from the Rhine, driven by exposure and accessibility rather than proximity to the channel.
  • Three leaf-node constellations produce the High Priority class: an exposed care facility, a high density of exposed buildings, or medium exposed-building density combined with limited immediate and remote accessibility.
  • Changing the percentile thresholds for building density or the PDC weights shifts the number of tiles in each class, giving end users a transparent parameter to align the map with their operational preferences.

Reading between the lines

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

  • Because the 96 CPT probabilities are not published and no sensitivity analysis is run, the stability of the priority classes across plausible expert judgements is untested; re-eliciting those numbers would be the natural next experiment.
  • The equal-spacing PDC weights treat the gap from None to Low as no more important than the gap from Medium to High; a utility-shaped weighting that penalises High risk disproportionately could change tile rankings, especially near the class boundaries.
  • The architecture implies a time-series extension: feeding successive flood snapshots through the prebuilt network would produce priority maps that show how the highest-risk areas move as the flood progresses, and those maps could be validated against actual rescue requests.
  • Since remote accessibility is computed on a road network with flooded segments removed, the method will misjudge tiles where flood depth, not just extent, blocks roads; if depth data became available cheaply, both the routing model and the accessibility node would need refinement.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes PrioReMap, a decision support method for area prioritisation in flood disaster response. The method combines a GIS-informed Bayesian network (BN) with a state-weighting scheme and clustering to translate the probability distribution of a target node, Risk of People in Need of Assistance, into four prioritisation categories (High Priority, Priority, Exposed, Safe). The BN is constructed in the preparedness phase, and during an event, tile-specific GIS models (exposed building density, exposed care facilities, immediate and remote accessibility) provide evidence to the BN. The prioritisation uses a Probability Distribution Criticality (PDC) value (Eq. 1) that weighs the target-node states, followed by k-means clustering. The method is illustrated on a simulated 500-year flood scenario in Cologne, Germany, using OpenStreetMap and HQ500 flood extent data. The paper's central claim is that PrioReMap provides rapid and transparent recommendations by shifting preparation to the preparedness phase and by presenting a concise output map.

Significance. The paper addresses a real and under-served operational need: rapid, transparent area prioritisation during an ongoing flood, where existing GIS-BN models focus on strategic planning rather than response. The conceptual contribution is valuable, particularly the deliberate separation of preparedness-phase model construction from response-phase recommendation, and the attempt to base recommendations on the whole probability distribution rather than only the most severe state. The method is presented with a detailed, realistic case study that demonstrates the full data processing pipeline, including flood extent, building and care-facility overlays, road accessibility routing, and tile-level BN inference. The authors are transparent about many limitations, including data quality and the static nature of the flood snapshot. However, the evidence provided is not sufficient to establish the central claim of effective prioritisation. The conditional probability table of the target node, which is the load-bearing component of the BN, is not published, and no sensitivity analysis, baseline comparison, or validation against observed outcomes is provided.

major comments (4)
  1. [§4.1, Fig. 5, Table 2] The CPT of the target node Risk of People in Need of Assistance is described only by qualitative principles (i)–(iii), and the 96 numerical probabilities are not published. Since the PDC in Eq. (1) is a direct weighted sum of the target-node posterior probabilities, every prioritisation recommendation in Fig. 11 is a function of these unstated expert numbers. Two experts who both respect the stated principles could produce materially different CPTs, leading to different recommendations. The deterministic rule (iii) alone forces High whenever an exposed care facility is present, which accounts for 9 of the 32 tiles with P(High)=1 in the case study. This is a reproducibility and robustness gap. I recommend publishing the full CPT as supplementary material and adding a sensitivity analysis that shows how the prioritisation map changes under alternative, plausible CPTs consistent with the stated principles.
  2. [§5 and §6] The central claim that PrioReMap provides effective area prioritisation is not tested against observed outcomes or compared with simpler baselines. The case study demonstrates that the method produces a map, but it does not show that the map is more accurate, faster, or more useful than, for instance, a direct overlay of the input GIS layers or a simple additive index of building density, care-facility presence, and accessibility. Without such a comparison, or at least a structured evaluation by domain experts, the added value of the BN and the weighting/clustering steps remains unquantified. The discussion claims that the method addresses cognitive load and improves consistency, but no measurements or user study support these claims.
  3. [§4.3, Eq. (1)] The PDC weighting scheme is presented with fixed weights [0, 0.33, 0.66, 1] with the remark that they 'can easily be adjusted', but no analysis is provided of how the prioritisation classes depend on these weights. The equal-distance assumption is a modelling choice, not a result. Similarly, the number of k-means clusters (three active clusters plus the Safe category) is justified only 'for demonstration purposes'. Since the final categories are the direct output that a responder would act on, the method should either report the stability of the clusters under perturbations of the weights and the cluster count, or provide a principled way to set these parameters.
  4. [§5.1.1, §4.2] Several GIS thresholds are chosen without clear rationale or sensitivity checks: the 90th and 75th percentile cut-offs for building density, the hexagon tile size (0.114 km²), and the number and placement of routing destination locations. The authors acknowledge that percentile thresholds influence results, but they do not quantify the influence. Given that the case study is the only demonstration of the method, the robustness of the prioritisation map to these parameters should be assessed, for example by varying the percentile thresholds and tile size within reasonable ranges and reporting the resulting changes in the numbers of high-priority and priority tiles.
minor comments (6)
  1. [Abstract] The phrase 'aims to providing' is a grammatical error; it should be 'aims to provide'.
  2. [References] The reference for Pearl (1985) appears twice: once correctly as 'Judea Pearl' and once typoed as 'Judea Peal'. The latter should be corrected and merged.
  3. [Throughout] The paper has several formatting artefacts, e.g. 'A M ETHOD' and 'PR IOREMAP' in the title/header, and inconsistent spacing in equations and captions. These should be cleaned up.
  4. [§3, Fig. 2b] The caption of Fig. 2b says 'geo-data that is not bound to tight to the boundary of the cell'; the word 'tight' should be removed and the sentence rephrased.
  5. [§4.1] Table 2 is difficult to read because the column headers are very wide and the table layout is not clear. A transposed layout with node names in the first column and states as separate rows would be more readable.
  6. [§5.1.1] In the sentence about the Accessibility of Immediate Unexposed Areas, the description of 1713 tiles as 'only partly accessible' is ambiguous because the node is actually representing the percentage of non-flooded area, not a binary accessibility state. Clarify the mapping from this percentage to the soft evidence used in the BN.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: PrioReMap's recommendations follow transparently from its stated model assumptions, GIS inputs, and Eq. (1); the unpublished CPT is a reproducibility concern, not a circular derivation.

full rationale

The derivation chain is self-contained in the sense required for circularity analysis. The BN target-node distribution is computed from the stated CPT principles (Section 4.1), the four GIS leaf-node models (Section 4.2), and standard BN inference; the recommendations are then a transparent function of the target-node posterior via Eq. (1), PDC = sum_i w_i P(s_i), with fixed weights [0, 0.33, 0.66, 1], followed by k-means clustering of PDC values (Section 4.3). No parameter is fitted to the case-study outcomes and then renamed as a prediction; the percentile thresholds for Density of Exposed Buildings and the CPT rules are declared model inputs, and the paper explicitly notes that threshold choices influence results and can be adjusted. The qualitative CPT principles (i)-(iii) are assumptions, not conclusions derived from the output, so the fact that exposed care facilities force High risk is a stated modelling rule rather than a circular 'finding'. The self-citations (Schneider et al. 2025a,b) do not carry the central claim: the 2025b accessibility model is one input component from a separate publication, and the paper does not invoke any uniqueness theorem or ansatz from that work to justify the overall PrioReMap structure. The paper's own limitation statements (Section 6: 'While this approach introduces threshold-based classifications that certainly influence the results' and 'deriving a single value ... is a sensitive task') are acknowledged robustness concerns; they do not reveal a circular step. The absence of published CPT numbers and of a sensitivity analysis is a real reproducibility and robustness gap, but it is not evidence that any equation or recommendation reduces to its own input by construction.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The central claim rests on several hand-chosen parameters and unvalidated expert assumptions. The PDC weights, percentile thresholds, tile size, number of k-means clusters, and unspecified CPT probabilities all directly determine the output maps. The causal structure of the Bayesian network is an expert assumption, not learned from data, and no independent evidence is provided for these choices.

free parameters (6)
  • PDC weights = None=0, Low=0.33, Medium=0.66, High=1
    Chosen by hand in Section 4.3 to equalize criticality distances; no empirical justification.
  • Building density percentile thresholds = 90th and 75th percentiles
    Set in Section 4.2 to define High and Medium states; the paper acknowledges these influence results.
  • Tile size (hexagon coverage) = 0.114 km^2 (420m max width)
    Selected for the case study in Section 5.1.1; no sensitivity analysis.
  • Number of k-means clusters = 3
    Fixed in Section 4.3; no justification or stability check.
  • CPT probabilities = 96 entries, unspecified numerically
    Described qualitatively in Section 4.1; the exact values are not disclosed, making them de facto free parameters.
  • Routing destination locations = Unspecified multiple locations
    Required for the remote accessibility model in Section 4.2; not enumerated.
assumptions (6)
  • standard math Bayesian probability propagation is valid.
    Assumed implicitly when using the Bayesian network for inference.
  • domain assumption Risk of people in need depends on flooded building density, exposed care facilities, and accessibility as encoded in the BN.
    The causal structure in Section 4.1 is expert-defined and not learned from data.
  • domain assumption Flood extent without depth is a sufficient hazard proxy for rapid response.
    Stated in Section 4.2; the authors justify this by time constraints and acknowledge limitations.
  • domain assumption OpenStreetMap data are sufficiently accurate and complete for the case study.
    Used throughout Section 5; the authors note in the Discussion that official data would be preferable.
  • domain assumption The HQ500 simulated flood layer is representative of a rapid-mapping flood extent during an ongoing event.
    Used as the hazard input in Section 5; the authors state it has the same structure as rapid mapping products.
  • ad hoc to paper K-means clustering of PDC values yields meaningful prioritisation classes.
    Introduced in Section 4.3; no evidence that cluster boundaries correspond to operational categories.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Method for Rapid Area Prioritisation in Flood Disaster Response." pith.science (2026). https://pith.science/paper/LEPSMA5O

@misc{pith2026250618423,
  author       = {Pith},
  title        = {Pith review of: A Method for Rapid Area Prioritisation in Flood Disaster Response},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LEPSMA5O}},
  note         = {Machine review of arXiv:2506.18423}
}
read the original abstract

In flood disasters, decision-makers have to rapidly prioritise the areas that need assistance based on a high volume of information. While approaches that combine GIS with Bayesian networks are generally effective in integrating multiple spatial variables and can thus reduce cognitive load, existing models in the literature are not equipped to address the time pressure and information-scape that is typical in a flood. To address the lack of a model for area prioritisation in flood disaster response, we present a novel decision support system that adheres to the time and information characteristics of an ongoing flood to infer the areas with the highest risk. This decision support system is based on a novel GIS-informed Bayesian network model that reflects the challenges of decision-making for area prioritisation. By developing the model during the preparedness phase, some of the most time-consuming aspects of the decision-making process are removed from the time-critical response phase. In this way, the proposed method aims to providing rapid and transparent area prioritisation recommendations for disaster response. To illustrate our method, we present a case study of an extreme flood scenario in Cologne, Germany.

Figures

Figures reproduced from arXiv: 2506.18423 by the authors.

Figure 1
Figure 1. Low-dimensional BN model composed of the directed acyclic graph (left side) and CPT (right side). [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Example GIS-data on a 5x5 matrix. Fig. 2a shows (homogeneous) building density geo-data in each cells with three configurations (high density in dark-gray, medium density in light-gray, and low density in white) processed as hard evidence. Fig. 2b shows geo-data that is not bound to tight to the boundary of the cell processed as soft evidence. considered more critical, despite having the same probability for the mos… view at source ↗
Figure 3
Figure 3. Heatmaps presenting the state probabilities of the [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Summary of the PrioReMap method. We assume that the risk of people in need of assistance depends on the number of people who are exposed to the hazard, which we represent via the proxy variable Density of Exposed Buildings in an area (Ehrlich et al., 2021) (first paren…
Figure 5
Figure 5. Figure 5: Bayesian network to infer the Risk of People in Need of Assistance. The BN is composed of six variables (two conditional nodes and four marginal nodes) that show two to four states (see [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Example hexagon including neighbouring hexagons. [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Case study area. 12 [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 8
Figure 8. Figure 8: Histogram of exposed buildings per tile. [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: Results of the GIS models. 14 [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]
Figure 10
Figure 10. Figure 10: Results of the GIS-informed BN model. tiles show a probability distribution without extreme values, resulting from a mix of critical and less critical node states, e.g., tiles with a medium density of flooded buildings that remain accessible, or tiles with a low densi…
Figure 11
Figure 11. Figure 11: PrioReMap of the case study [PITH_FULL_IMAGE:figures/full_fig_p017_11.png]
Figure 12
Figure 12. Figure 12: Tile similarities based on leaf node states. [PITH_FULL_IMAGE:figures/full_fig_p017_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

63 extracted references · 51 canonical work pages

  1. [1]

    Collaboration is key: Exploring the 2021 flood response for critical infrastructures in germany

    Florence Catherine Nick, Nathalie Sänger, Sophie van der Heijden, and Simone Sandholz. Collaboration is key: Exploring the 2021 flood response for critical infrastructures in germany. International Journal of Disaster Risk Reduction, 91: 0 103710, June 2023. ISSN 2212-4209. doi:10.1016/j.ijdrr.2023.103710

    work page arXiv 2021

  2. [2]

    A probabilistic modeling and simulation framework for power grid flood risk assessment

    Panagiotis Asaridis, Daniela Molinari, Francesco Di Maio, Francesco Ballio, and Enrico Zio. A probabilistic modeling and simulation framework for power grid flood risk assessment. International Journal of Disaster Risk Reduction, 120: 0 105353, April 2025. ISSN 2212-4209. doi:10.1016/j.ijdrr.2025.105353

    work page arXiv 2025

  3. [3]

    A comprehensive analysis method of spatial prioritization for urban flood management based on source tracking

    Wenchao Qi, Chao Ma, Hongshi Xu, Kai Zhao, and Zifan Chen. A comprehensive analysis method of spatial prioritization for urban flood management based on source tracking. Ecological Indicators, 135: 0 108565, February 2022. ISSN 1470-160X. doi:10.1016/j.ecolind.2022.108565

    arXiv 2022

  4. [4]

    Nirupama

    Costas Armenakis and N. Nirupama. Prioritization of disaster risk in a community using gis. Natural Hazards, 66 0 (1): 0 15--29, April 2012. ISSN 1573-0840. doi:10.1007/s11069-012-0167-8

    work page doi:10.1007/s11069-012-0167-8 2012

  5. [5]

    Fusion of geospatial information from remote sensing and social media to prioritise rapid response actions in case of floods

    Marc Wieland, Sebastian Schmidt, Bernd Resch, Andreas Abecker, and Sandro Martinis. Fusion of geospatial information from remote sensing and social media to prioritise rapid response actions in case of floods. Natural Hazards, January 2025. ISSN 1573-0840. doi:10.1007/s11069-025-07120-7

    work page doi:10.1007/s11069-025-07120-7 2025

  6. [6]

    An integrated bayesian networks and geographic information system (bns-gis) approach for flood disaster risk assessment: A case study of yinchuan, china

    Yuwen Lu, Guofang Zhai, and Shutian Zhou. An integrated bayesian networks and geographic information system (bns-gis) approach for flood disaster risk assessment: A case study of yinchuan, china. Ecological Indicators, 166: 0 112322, September 2024. ISSN 1470-160X. doi:10.1016/j.ecolind.2024.112322

    work page arXiv 2024

  7. [7]

    A resilience glossary shaped by context: Reviewing resilience-related terms for critical infrastructures

    Andrea Mentges, Lukas Halekotte, Moritz Schneider, Tobias Demmer, and Daniel Lichte. A resilience glossary shaped by context: Reviewing resilience-related terms for critical infrastructures. International Journal of Disaster Risk Reduction, 96: 0 103893, October 2023. ISSN 2212-4209. doi:10.1016/j.ijdrr.2023.103893

    arXiv 2023

  8. [8]

    Flood hazard mapping methods: A review

    Rofiat Bunmi Mudashiru, Nuridah Sabtu, Ismail Abustan, and Waheed Balogun. Flood hazard mapping methods: A review. Journal of Hydrology, 603: 0 126846, December 2021. ISSN 0022-1694. doi:10.1016/j.jhydrol.2021.126846

    work page arXiv 2021

Show all 63 references
  1. [9]

    A data-driven approach to analyse the co-evolution of urban systems through a resilience lens: A helsinki case study

    Ylenia Casali, Nazli Yonca Aydin, and Tina Comes. A data-driven approach to analyse the co-evolution of urban systems through a resilience lens: A helsinki case study. Environment and Planning B: Urban Analytics and City Science, 51 0 (9): 0 2074--2091, February 2024. ISSN 239...

    2024 doi

  2. [10]

    A systematic review on approaches and methods used for flood vulnerability assessment: framework for future research

    Sufia Rehman, Mehebub Sahana, Haoyuan Hong, Haroon Sajjad, and Baharin Bin Ahmed. A systematic review on approaches and methods used for flood vulnerability assessment: framework for future research. Natural Hazards, 96 0 (2): 0 975--998, March 2019. ISSN 1573-0840. doi:10.100...

    2019 doi

  3. [11]

    Emergency response inference mapping (erimap): A bayesian network-based method for dynamic observation processing

    Moritz Schneider, Lukas Halekotte, Tina Comes, Daniel Lichte, and Frank Fiedrich. Emergency response inference mapping (erimap): A bayesian network-based method for dynamic observation processing. Reliability Engineering & System Safety, 255: 0 110640, March 2025 a . ISSN 0951...

    2025

  4. [12]

    The rhythm of risk: Exploring spatio-temporal patterns of urban vulnerability with ambulance calls data

    Mikhail Sirenko, Tina Comes, and Alexander Verbraeck. The rhythm of risk: Exploring spatio-temporal patterns of urban vulnerability with ambulance calls data. Environment and Planning B: Urban Analytics and City Science, 52 0 (4): 0 863--881, August 2024. ISSN 2399-8091. doi:1...

    2024 doi

  5. [13]

    Improving situation awareness in crisis response teams: An experimental analysis of enriched information and centralized coordination

    Bartel Van de Walle, Bert Brugghemans, and Tina Comes. Improving situation awareness in crisis response teams: An experimental analysis of enriched information and centralized coordination. International Journal of Human-Computer Studies, 95: 0 66--79, November 2016. ISSN 1071...

    2016 doi

  6. [14]

    Ai for crisis decisions

    Tina Comes. Ai for crisis decisions. Ethics and Information Technology, 26 0 (1), February 2024. ISSN 1572-8439. doi:10.1007/s10676-024-09750-0

    2024 doi

  7. [15]

    Linking urban structure types and bayesian network modelling for an integrated flood risk assessment in data-scarce mega-cities

    Veronika Zwirglmaier and Matthias Garschagen. Linking urban structure types and bayesian network modelling for an integrated flood risk assessment in data-scarce mega-cities. Urban Climate, 56: 0 102034, July 2024. ISSN 2212-0955. doi:10.1016/j.uclim.2024.102034

    2024

  8. [16]

    Assessing urban flood disaster risk using bayesian network model and gis applications

    Zening Wu, Yanxia Shen, Huiliang Wang, and Meimei Wu. Assessing urban flood disaster risk using bayesian network model and gis applications. Geomatics, Natural Hazards and Risk, 10 0 (1): 0 2163--2184, January 2019. ISSN 1947-5713. doi:10.1080/19475705.2019.1685010

    2019

  9. [17]

    Dependent infrastructure service disruption mapping (disruptionmap): A method to assess cascading service disruptions in disaster scenarios

    Moritz Schneider, Lukas Halekotte, Andrea Mentges, and Frank Fiedrich. Dependent infrastructure service disruption mapping (disruptionmap): A method to assess cascading service disruptions in disaster scenarios. Scientific Reports, 15 0 (1), February 2025 b . ISSN 2045-2322. d...

    2025 doi

  10. [18]

    An integrated gis-bbn approach to quantify resilience of roadways network infrastructure system against flood hazard

    Mrinal Kanti Sen and Subhrajit Dutta. An integrated gis-bbn approach to quantify resilience of roadways network infrastructure system against flood hazard. ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part A: Civil Engineering, 6 0 (4), December 2020. ISSN...

    2020 doi

  11. [19]

    Where do the numbers come from?

    M.J. Druzdzel and L.C. van der Gaag. Building probabilistic networks: " Where do the numbers come from?" guest editors' introduction. IEEE Transactions on Knowledge and Data Engineering, 12 0 (4): 0 481--486, July 2000. ISSN 1558-2191. doi:10.1109/TKDE.2000.868901

  12. [20]

    Roadmap towards responsible ai in crisis resilience management, 2022

    Cheng-Chun Lee, Tina Comes, Megan Finn, and Ali Mostafavi. Roadmap towards responsible ai in crisis resilience management, 2022

    2022

  13. [21]

    Bayesian networks: A model of self-activated memory for evidential reasoning

    Judea Pearl. Bayesian networks: A model of self-activated memory for evidential reasoning. In Proceedings of the 7th conference of the Cognitive Science Society, University of California, Irvine, CA, USA, pages 15--17, June 1985

    1985

  14. [22]

    A dynamic emergency decision-making method based on group decision making with uncertainty information

    Jing Zheng, Yingming Wang, Kai Zhang, and Juan Liang. A dynamic emergency decision-making method based on group decision making with uncertainty information. International Journal of Disaster Risk Science, 11 0 (5): 0 667--679, October 2020. ISSN 2192-6395. doi:10.1007/s13753-...

    2020 doi

  15. [23]

    Integrated bayesian networks with gis for electric vehicles charging site selection

    Yan Zhang, Bak Koon Teoh, and Limao Zhang. Integrated bayesian networks with gis for electric vehicles charging site selection. Journal of Cleaner Production, 344: 0 131049, April 2022. ISSN 0959-6526. doi:10.1016/j.jclepro.2022.131049

    2022

  16. [24]

    Stewart, Oz Sahin, and Abel Silva Vieira

    Shahid Ali, Rodney A. Stewart, Oz Sahin, and Abel Silva Vieira. Spatial bayesian approach for socio-economic assessment of pumped hydro storage. Renewable and Sustainable Energy Reviews, 189: 0 114007, January 2024. ISSN 1364-0321. doi:10.1016/j.rser.2023.114007

    2024

  17. [25]

    A bayesian network model for suitability evaluation of underground space development in urban areas: The case of changsha, china

    Zhiwen Xu, Suhua Zhou, Chao Zhang, Minghui Yang, and Mingyi Jiang. A bayesian network model for suitability evaluation of underground space development in urban areas: The case of changsha, china. Journal of Cleaner Production, 418: 0 138135, September 2023. ISSN 0959-6526. do...

    2023

  18. [26]

    Spatial bayesian belief networks as a planning decision tool for mapping ecosystem services trade-offs on forested landscapes

    Julen Gonzalez-Redin, Sandra Luque, Laura Poggio, Ron Smith, and Alessandro Gimona. Spatial bayesian belief networks as a planning decision tool for mapping ecosystem services trade-offs on forested landscapes. Environmental Research, 144: 0 15--26, January 2016. ISSN 0013-935...

    2016 doi

  19. [27]

    Application of bayesian networks for fire risk mapping using gis and remote sensing data

    Wisdom Mdumiseni Dlamini. Application of bayesian networks for fire risk mapping using gis and remote sensing data. GeoJournal, 76 0 (3): 0 283--296, May 2010. ISSN 1572-9893. doi:10.1007/s10708-010-9362-x

    2010 doi

  20. [28]

    Grêt-Regamey and D

    A. Grêt-Regamey and D. Straub. Spatially explicit avalanche risk assessment linking bayesian networks to a gis. Natural Hazards and Earth System Sciences, 6 0 (6): 0 911--926, October 2006. ISSN 1684-9981. doi:10.5194/nhess-6-911-2006

    2006 doi

  21. [29]

    A spatial bayesian-network approach as a decision-making tool for ecological-risk prevention in land ecosystems

    Kai Guo, Xinchang Zhang, Xi Kuai, Zhifeng Wu, Yiyun Chen, and Yi Liu. A spatial bayesian-network approach as a decision-making tool for ecological-risk prevention in land ecosystems. Ecological Modelling, 419: 0 108929, March 2020. ISSN 0304-3800. doi:10.1016/j.ecolmodel.2019.108929

    2020

  22. [30]

    Sousa, and José C

    Erica Arango, Monica Santamaria, Maria Nogal, Hélder S. Sousa, and José C. Matos. Flood risk assessment for road infrastructures using bayesian networks: case study of santarem - portugal. Acta Polytechnica CTU Proceedings, 36: 0 33--46, August 2022. ISSN 2336-5382. doi:10.143...

    2022 doi

  23. [31]

    Key disaster-causing factors chains on urban flood risk based on bayesian network

    Shanqing Huang, Huimin Wang, Yejun Xu, Jingwen She, and Jing Huang. Key disaster-causing factors chains on urban flood risk based on bayesian network. Land, 10 0 (2): 0 210, February 2021. ISSN 2073-445X. doi:10.3390/land10020210

    2021 doi

  24. [32]

    A spatial bayesian network model to assess the benefits of early warning for urban flood risk to people

    Stefano Balbi, Ferdinando Villa, Vahid Mojtahed, Karin Tessa Hegetschweiler, and Carlo Giupponi. A spatial bayesian network model to assess the benefits of early warning for urban flood risk to people. Natural Hazards and Earth System Sciences, 16 0 (6): 0 1323--1337, June 201...

    2016 doi

  25. [33]

    Identification of issues in disaster response to flooding, focusing on the time continuity between residents’ evacuation and rescue activities

    Akira Matsuki and Michinori Hatayama. Identification of issues in disaster response to flooding, focusing on the time continuity between residents’ evacuation and rescue activities. International Journal of Disaster Risk Reduction, 95: 0 103841, September 2023. ISSN 2212-4209....

    2023

  26. [34]

    Identifying decision support needs for emergency response to multiple natural hazards: an activity theory approach

    Viktor Sköld Gustafsson, Tobias Andersson Granberg, Sofie Pilemalm, and Martin Waldemarsson. Identifying decision support needs for emergency response to multiple natural hazards: an activity theory approach. Natural Hazards, 120 0 (3): 0 2777--2802, November 2023. ISSN 1573-0...

    2023 doi

  27. [35]

    Generating conditional probabilities for bayesian networks: Easing the knowledge acquisition problem, 2008

    Balaram Das. Generating conditional probabilities for bayesian networks: Easing the knowledge acquisition problem, 2008

    2008

  28. [36]

    Open and consistent geospatial data on population density, built-up and settlements to analyse human presence, societal impact and sustainability: A review of ghsl applications

    Daniele Ehrlich, Sergio Freire, Michele Melchiorri, and Thomas Kemper. Open and consistent geospatial data on population density, built-up and settlements to analyse human presence, societal impact and sustainability: A review of ghsl applications. Sustainability, 13 0 (14): 0...

    2021 doi

  29. [37]

    Yared Abayneh Abebe, Maria Pregnolato, and Sebastiaan N. Jonkman. Flood impacts on healthcare facilities and disaster preparedness – a systematic review. International Journal of Disaster Risk Reduction, 119: 0 105340, March 2025. ISSN 2212-4209. doi:10.1016/j.ijdrr.2025.105340

    2025

  30. [38]

    An integrated metric for rapid and equitable emergency rescue during urban flash flooding events

    Yitong Li, Yifan Wang, and Jie Gong. An integrated metric for rapid and equitable emergency rescue during urban flash flooding events. International Journal of Disaster Risk Reduction, 118: 0 105209, February 2025. ISSN 2212-4209. doi:10.1016/j.ijdrr.2025.105209

    2025

  31. [39]

    Hexagonal discrete global grid systems for geospatial computing

    Kevin Sahr. Hexagonal discrete global grid systems for geospatial computing. Archiwum Fotogrametrii, Kartografii i Teledetekcji, 22: 0 363--376, 2011

    2011

  32. [40]

    Detection and mapping of the geomorphic effects of flooding using uav photogrammetry

    Jakub Langhammer and Tereza Vacková. Detection and mapping of the geomorphic effects of flooding using uav photogrammetry. Pure and Applied Geophysics, 175 0 (9): 0 3223--3245, April 2018. ISSN 1420-9136. doi:10.1007/s00024-018-1874-1

    2018 doi

  33. [41]

    Jacobs, Chloe M

    Ramesh Sivanpillai, Kevin M. Jacobs, Chloe M. Mattilio, and Ela V. Piskorski. Rapid flood inundation mapping by differencing water indices from pre- and post-flood landsat images. Frontiers of Earth Science, 15 0 (1): 0 1--11, April 2020. ISSN 2095-0209. doi:10.1007/s11707-020-0818-0

    2020 doi

  34. [42]

    Rapid mapping in support of emergency response after earthquake events

    Stephanie Wegscheider, Tobias Schneiderhan, Alexander Mager, Hendrik Zwenzner, Joachim Post, and Günter Strunz. Rapid mapping in support of emergency response after earthquake events. Natural Hazards, 68 0 (1): 0 181--195, July 2013. ISSN 1573-0840. doi:10.1007/s11069-013-0589-y

    2013 doi

  35. [43]

    Evaluating the robustness of bayesian flood mapping with sentinel-1 data: A multi-event validation study

    Florian Roth, Mark Edwin Tupas, Claudio Navacchi, Jie Zhao, Wolfgang Wagner, and Bernhard Bauer-Marschallinger. Evaluating the robustness of bayesian flood mapping with sentinel-1 data: A multi-event validation study. Science of Remote Sensing, 11: 0 100210, June 2025. ISSN 26...

    2025

  36. [44]

    Naser Lessani, and Zhenlong Li

    Temitope Akinboyewa, Huan Ning, M. Naser Lessani, and Zhenlong Li. Automated floodwater depth estimation using large multimodal model for rapid flood mapping. Computational Urban Science, 4 0 (1), May 2024. ISSN 2730-6852. doi:10.1007/s43762-024-00123-3

    2024 doi

  37. [45]

    Planet dump retrieved from https://planet.osm.org

    OpenStreetMap contributors . Planet dump retrieved from https://planet.osm.org . https://www.openstreetmap.org , 2017

    2017

  38. [46]

    High-resolution flood numerical model and dijkstra algorithm based risk avoidance routes planning

    Bingyao Li, Jingming Hou, Xinghua Wang, Yongyong Ma, Donglai Li, Tian Wang, and Guangzhao Chen. High-resolution flood numerical model and dijkstra algorithm based risk avoidance routes planning. Water Resources Management, 37 0 (8): 0 3243--3258, April 2023. ISSN 1573-1650. do...

    2023 doi

  39. [47]

    Utkarsh Gangwal, A. R. Siders, Jennifer Horney, Holly A. Michael, and Shangjia Dong. Critical facility accessibility and road criticality assessment considering flood-induced partial failure. Sustainable and Resilient Infrastructure, 8 0 (sup1): 0 337--355, November 2022. ISSN...

    2022

  40. [48]

    Aspects of the January 1995 flood in Germany

    Andreas Fink, Uwe Ulbrich, and Heinz Engel. Aspects of the January 1995 flood in Germany . Weather, 51 0 (2): 0 34--39, February 1996. ISSN 1477-8696. doi:10.1002/j.1477-8696.1996.tb06182.x

    1995

  41. [49]

    assessment of economic flood damage

    B. Merz, H. Kreibich, R. Schwarze, and A. Thieken. Review article "assessment of economic flood damage". Natural Hazards and Earth System Sciences, 10 0 (8): 0 1697--1724, August 2010. ISSN 1684-9981. doi:10.5194/nhess-10-1697-2010

    2010 doi

  42. [50]

    Spontaneous volunteers and the flood disaster 2021 in germany: Development of social innovations in flood risk management

    Marina Bier, Ramian Fathi, Christiane Stephan, Anke Kahl, Frank Fiedrich, and Alexander Fekete. Spontaneous volunteers and the flood disaster 2021 in germany: Development of social innovations in flood risk management. Journal of Flood Risk Management, July 2023. ISSN 1753-318...

    2021 doi

  43. [51]

    Francesca Müller, Marina Bier, Samuel Tomczyk, Anke Kahl, and Frank Fiedrich. The issue of overload: A mixed methods analysis of activity-related stress and the use of mental health and psychosocial support by spontaneous volunteers during the 2021 flood in germany. Internatio...

    2021

  44. [52]

    INSPIRE Dataset Feed: Flood risk map Layer NRW - Low probability (HQ500) , [accessed 09.04.2024]

    State Agency for Nature, Environment and Consumer Protection of North Rhine-Westphalia . INSPIRE Dataset Feed: Flood risk map Layer NRW - Low probability (HQ500) , [accessed 09.04.2024]. URL https://www.gis-rest.nrw.de/atomFeed/rest/atom/182925c1-879f-4054-bd69-b6f28e05b270.html

    2024

  45. [53]

    Critical infrastructure cascading effects

    Alexander Fekete. Critical infrastructure cascading effects. Disaster resilience assessment for floods affecting city of Cologne and Rhein - Erft - Kreis . Journal of Flood Risk Management, 13 0 (2): 0 e312600, 2020. ISSN 1753-318X. doi:10.1111/jfr3.12600

    2020 doi

  46. [54]

    Alta de Waal and Johan W. Joubert. Explainable Bayesian networks applied to transport vulnerability. Expert Systems with Applications, 209: 0 118348, December 2022. ISSN 0957-4174. doi:10.1016/j.eswa.2022.118348

    2022

  47. [55]

    Hassall, Gordon Dailey, Joanna Zawadzka, Alice E

    Kirsty L. Hassall, Gordon Dailey, Joanna Zawadzka, Alice E. Milne, Jim A. Harris, Ron Corstanje, and Andrew P. Whitmore. Facilitating the elicitation of beliefs for use in Bayesian Belief modelling. Environmental Modelling & Software, 122: 0 104539, December 2019. ISSN 1364-81...

    2019

  48. [56]

    Morris, Jeremy E

    David E. Morris, Jeremy E. Oakley, and John A. Crowe. A web-based tool for eliciting probability distributions from experts. Environmental Modelling and Software, 52: 0 1--4, February 2014. doi:10.1016/j.envsoft.2013.10.010

    2014 doi

  49. [57]

    Bayesian networks: A model of self-activated memory for evidential reasoning

    Judea Peal. Bayesian networks: A model of self-activated memory for evidential reasoning. In Proceedings of the Annual Meeting of the Cognitive Science Society, volume 7, 1985

    1985

  50. [58]

    Do we practice what we preach? the dissonance between resilience understanding and measurement

    Lukas Halekotte, Andrea Mentges, and Daniel Lichte. Do we practice what we preach? the dissonance between resilience understanding and measurement. International Journal of Disaster Risk Reduction, 118: 0 105265, February 2025. ISSN 2212-4209. doi:10.1016/j.ijdrr.2025.105265

    2025

  51. [59]

    Systematic literature review of data quality within openstreetmap

    Jasmeet Kaur, Jaiteg Singh, Sukhjit Singh Sehra, and Hardeep Singh Rai. Systematic literature review of data quality within openstreetmap. In 2017 International Conference on Next Generation Computing and Information Systems (ICNGCIS), pages 177--182. IEEE, December 2017. doi:...

    2017 doi

  52. [60]

    A new method for the assessment of spatial accuracy and completeness of openstreetmap building footprints

    Maria Antonia Brovelli and Giorgio Zamboni. A new method for the assessment of spatial accuracy and completeness of openstreetmap building footprints. ISPRS International Journal of Geo-Information, 7 0 (8): 0 289, July 2018. ISSN 2220-9964. doi:10.3390/ijgi7080289

    2018 doi

  53. [61]

    Quality of crowdsourced geospatial building information: A global assessment of openstreetmap attributes

    Filip Biljecki, Yoong Shin Chow, and Kay Lee. Quality of crowdsourced geospatial building information: A global assessment of openstreetmap attributes. Building and Environment, 237: 0 110295, June 2023. ISSN 0360-1323. doi:10.1016/j.buildenv.2023.110295

    2023

  54. [62]

    Recommended practice: Flood mapping and damage assessment using sentinel-2 (s2) optical data, 2025

    UN-SPIDER . Recommended practice: Flood mapping and damage assessment using sentinel-2 (s2) optical data, 2025. URL https://www.un-spider.org/advisory-support/recommended-practices/recommended-practice-flood-mapping-and-damage-assessment. Accessed: 2025-04-29

    2025

  55. [63]

    INSARAG Guidelines Volume II, Manual B: Operations

    UN-OCHA. INSARAG Guidelines Volume II, Manual B: Operations. United Nations Office for the Coordination of Humanitarian Affairs, Geneva, Switzerland, 2020. URL https://www.insarag.org

    2020

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.