infrastructure · digital · family: the model was never trained on this
knock out power, lose water, lose everything
Cascading failures across interdependent infrastructure systems cannot be predicted
Problem statement
Modern infrastructure systems — power grids, water networks, transportation, telecommunications, gas pipelines — are deeply interdependent, and disruptions in one system can cascade to others. When a hurricane knocks out power, water treatment plants stop, hospitals lose function, emergency services are overwhelmed, and transportation clogs. Writing about cascading and compounding hazard events, the NHERI Science Plan says the linkages and interdependencies of these cascading effects are "poorly understood but critical to a community's long-term resilience." The Science Plan calls for "multi-scale modeling and simulation capabilities that better couple human and physical systems." On the socioeconomic side, the plan says these events "do not affect all people equally," and it lists as a research need improved methods to capture data on "the complex interactions between physical and social vulnerability."
Why this matters
According to NOAA figures quoted in the NHERI Science Plan, the U.S. experienced 18 weather and climate disasters costing at least $1 billion each in 2022, after 22 such events in 2020 and 20 in 2021; damages from the 2022 disasters totaled $165.1 billion, and since 1980 the country has sustained 341 billion-dollar disasters with a cumulative cost exceeding $2.475 trillion. These events do not affect all people equally: the plan says older adults, people living in poverty, and people of color are disproportionately likely to be injured or killed, and that the long-term, cascading and compounding effects of hazards often disproportionately impact disadvantaged communities. The ISP program does not support research whose "primary methodological contribution focuses on individual infrastructure components without a systems research perspective."
What’s been tried and why it hasn’t worked
Until recently, the NHERI Science Plan says, engineering research on the design and performance of civil infrastructure "has predominantly focused on resilience against a single natural hazard such as earthquakes or tornadoes." Among its research questions on cascading and compounding impacts, the plan asks "how can knowledge from single-hazard research be leveraged to understand cascading or compounding hazard impacts?" Recovery is a research need as well: the plan calls for identifying "the services most needed by communities for recovery" and the infrastructure systems required to provide them. The plan's examples of cascading and compounding events include design decisions at the Fukushima plant that "caused a nuclear disaster after the 2011 Tohoku earthquake and tsunami," and co-seismic landslides from the 2008 Wenchuan earthquake whose debris "later remobilized as a series of catastrophic debris flows during intense storms several years after the earthquake." The plan says the "large uncertainty of loading actions from natural hazards is particularly challenging" both for performance-based engineering of civil infrastructure systems and "for gauging the interaction between civil infrastructure and social systems." The plan says robust regional-scale simulations "require high-quality datasets to describe the built environment and population in an urban area," and under its first key research question it names better "availability of quality inventories of site conditions, civil infrastructure and societal impacts" for use in research studies. The plan asks for longitudinal studies that "elucidate the linkages and non-linear interactions between disasters and hazards."
What would unlock progress
Multi-scale modeling and simulation capabilities that "better couple human and physical systems" across scales, from households and buildings to communities, regions, and the nation; integration of socioeconomic vulnerability data into infrastructure resilience models; methods to model compound and cascading hazard events rather than single hazards; large-scale building and infrastructure inventory datasets that are openly available and standardized; and attention to recovery — identifying the services communities most need to recover and the infrastructure systems required to provide them.
Entry points for student teams
A student team could build a simplified agent-based model of two interdependent infrastructure networks (e.g., power and water) for a specific small city, calibrating coupling parameters from published post-disaster reports (which are publicly available from FEMA and state emergency agencies). This is a feasible computational modeling project. Alternatively, a team could analyze publicly available post-disaster data to empirically characterize cascading failure sequences and their correlation with census-tract-level socioeconomic data, producing an empirical interdependency map.
Genes
Structural cousins: stuck the same way, in other fields
Sources
"Infrastructure Systems and People (ISP)," NSF CMMI Division. (accessed 2026-02-10). Supplemented with NHERI Science Plan Third Edition (2023), DOI: 10.17603/ds2-abbs-0966 (), and NSF DCL for NHERI FY2026-FY2035 (NSF 24-117, ). go to source 1 ↗ go to source 2 ↗ go to source 3 ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
- The NHERI Science Plan Third Edition (2023) provides detailed analysis of the multi-hazard modeling gap and is a valuable primary source.
- NSF 24-117 (DCL for NHERI FY2026-FY2035) signals a long-term commitment to natural hazards engineering research.
- Cross-domain connection: the cascading failure problem shares deep structure with supply chain resilience problems — both involve modeling failure propagation across heterogeneous interconnected networks.
- The equity dimension (infrastructure failures disproportionately affecting vulnerable communities) connects this to environmental justice research and the `constraint:equity` tag.
- Related programs: NSF Engineering for Civil Infrastructure (ECI) and the Smart and Connected Communities (S&CC) program address adjacent aspects.
Reconciliation 2026-09-30: Fact repair after reading the whole write-up against every source it names. All three were fetched on 2026-09-30. The NSF Infrastructure Systems and People (ISP) program page (https://www.nsf.gov/funding/opportunities/isp-infrastructure-systems-people, which now redirects to the archived PD 25-342Y page and points to NSF 26-515 as the current version) was read in full. The NHERI Science Plan, Third Edition (October 2023) was read in full text, 132 pages, downloaded from the DesignSafe published project PRJ-4240 behind the DOI 10.17603/ds2-abbs-0966; its citation (Robertson, Bobet, Edge, Holmes, Johnson, LaChance and Ramirez, 2023) matches the DataCite record. NSF 24-117, the Dear Colleague Letter on NHERI support for FY 2026-FY 2035 (August 13, 2024), was read in full at nsf.gov. (1) The write-up said the U.S. had "a billion-dollar disaster event every 16 days on average" in 2020-2024 against "every 82 days in the 1980s," and that severe storms alone averaged "$37.9 billion in annual damage." None of the three sources contains these figures. The Science Plan quotes NOAA (2023): 18 billion-dollar weather and climate disasters in 2022, 22 in 2020 and 20 in 2021, $165.1 billion in damages from the 2022 events, and 341 such events since 1980 costing more than $2.475 trillion in all; the paragraph now gives those figures. (2) "No unified model exists to predict how failures cascade" and "engineers cannot simulate these cascading chains realistically" are not in the sources; the ISP page says disruptions in one system can have cascading impacts on others, and the Science Plan, writing about cascading and compounding hazard events, says the linkages and interdependencies of these cascading effects are "poorly understood but critical to a community's long-term resilience"; the text now says those two things and no longer attaches the plan's sentence to the hurricane illustration. (3) "Existing models treat each infrastructure network in isolation or use grossly simplified coupling assumptions" is not in the sources; the sentence now gives the plan's call for "multi-scale modeling and simulation capabilities that better couple human and physical systems." (4) "The socioeconomic dimension is almost entirely absent" and "routinely omitted from resilience assessments" are not in the sources; the text now gives the plan's statements that these events "do not affect all people equally" and that improved methods are needed to capture data on "the complex interactions between physical and social vulnerability." (5) "Less redundancy" as a trait of low-income communities and the claim that "infrastructure investments are misallocated" are not in the sources; the text now gives the plan's statements that older adults, people living in poverty, and people of color are disproportionately likely to be injured or killed (introduction) and that cascading and compounding effects often disproportionately impact disadvantaged communities (Appendix A3), and quotes the ISP program's own scope line that it does not support research whose "primary methodological contribution focuses on individual infrastructure components without a systems research perspective." (The plan's statement about low-income individuals in Appendix A1 is about windstorms, so it is not used as a general claim.) (6) The What's Been Tried section rested on six claims that appear in none of the sources: single-network cascade models for power grids that cannot capture cross-network coupling; interdependency models that focus on hardening rather than recovery; "most models assume single-hazard scenarios" with an earthquake-tsunami-fire chain; uncertainty treatment that is "surprisingly undeveloped" against "wide variability of outcomes"; building inventory data "largely unavailable"; and a missing "unified theoretical framework describing energy-transfer pathways, interaction modes, and expansion mechanisms." The section now says what the Science Plan says: that engineering research until recently "has predominantly focused on resilience against a single natural hazard" (the nearest thing in the sources to the old single-hazard sentence); the single-hazard-to-cascading research question (Appendix A3); recovery services as a research need (Key Research Question 3); the plan's own cascade examples (Fukushima after the 2011 Tohoku earthquake and tsunami; Wenchuan landslide debris remobilized years later); the large uncertainty of loading actions (Key Research Question 6); that robust regional-scale simulations "require high-quality datasets to describe the built environment and population in an urban area" and better availability of quality inventories (Key Research Question 1); and the call for studies that "elucidate the linkages and non-linear interactions between disasters and hazards." (7) In What Would Unlock Progress, "realistic coupling mechanisms rather than just topological connections" and "a shift from failure prevention to recovery process optimization that leverages positive interdependencies" are not in the sources; the items now follow Key Research Questions 3 and 4. Confirmed in the sources and left as written: the Source line (ISP under the CMMI division; the Science Plan's title, edition, year and DOI; NSF 24-117 as the NHERI FY 2026-FY 2035 letter); that infrastructure systems are interdependent and that disruptions in one system cascade to others (ISP); the list of power, water, transportation, telecommunications and pipelines as examples; the hurricane-to-hospitals chain as an illustration; integration of socioeconomic data and compound-hazard modeling as needs; openly available and standardized large-scale infrastructure databases (ISP names "collecting, standardizing, and sharing large-scale databases"); and the Source Note that NSF 24-117 signals long-term commitment (the letter calls NHERI "a high priority" and plans support through FY 2035). Could not be checked against these sources: the entry point's statement that post-disaster reports are publicly available from FEMA and state emergency agencies (a general statement, left as written), and the Source Note naming NSF's Engineering for Civil Infrastructure and Smart and Connected Communities programs as related (neither is named in the three sources; left as a pointer). The Source line now carries the resolving links for the two supplementary sources it already named (the DOI link for the Science Plan and the nsf.gov page for NSF 24-117); no source was added. No title changed. Genome tags untouched. A second reader went over these repairs against the same three sources on the same day and tightened the wording of items 2, 4, 5, 6 and 7 where the first repair said more than the sources do; no repair was undone.