The EMP order tested continuity plans that count on distance
Executive Order 13865 told agencies that support National Essential Functions to plan for electromagnetic pulses. An event that can affect large areas at once undercuts continuity plans whose resilience comes from the distance between sites.
Looking Back revisits past weeks one at a time, up to the most recent, and asks what it meant for the banks, health systems and public agencies we serve. The series was written in 2026, with the benefit of hindsight, and each article says which week it looks back on.
Two-minute summaryThe dek and the first sentence of each section, quoted from the article
Executive Order 13865 told agencies that support National Essential Functions to plan for electromagnetic pulses. An event that can affect large areas at once undercuts continuity plans whose resilience comes from the distance between sites.
- The week On March 26, 2019, the President signed Executive Order 13865, Coordinating National Resilience to Electromagnetic Pulses.
- Places: when the alternate site is inside the same footprint A continuity plan for an essential function names where the function will run if its home is lost.
- Parts: repair assumes a supply that is somewhere else A continuity plan can say a great deal about relocation and much less about repair.
- People: the staff who carry the plan are in the footprint too Every continuity plan names people.
- Warning and word: how the agency learns, and how it tells The fourth dependency is information.
- Objections: probability, cost and whose mission this is Three objections to this analysis deserve an answer.
- Testing a plan against a wide-area event A plan that has been read against these four dependencies still needs a test.
- How the four dependencies fit together Seen together, the four dependencies form one picture.
- With hindsight The first step in the order’s chain produced a public product quickly.
The week
On March 26, 2019, the President signed Executive Order 13865, Coordinating National Resilience to Electromagnetic Pulses. It was published in the Federal Register on March 29, 2019, at 84 FR 12041. The order opens with a statement of purpose: an electromagnetic pulse, or EMP, has the potential to disrupt, degrade and damage technology and critical infrastructure systems, and human-made or naturally occurring EMPs can affect large geographic areas.
The order defines two kinds of event. A high-altitude EMP is a human-made pulse that occurs when a nuclear device is detonated at approximately 40 kilometers or more above the surface of the Earth. A geomagnetic disturbance is a natural pulse driven by a temporary disturbance of the Earth’s magnetic field resulting from interactions with solar eruptions. The order says both can affect large geographic areas. Earlier policy already addressed the natural case: Executive Order 13744, Coordinating Efforts To Prepare the Nation for Space Weather Events, was signed on October 13, 2016.
The policy section says the federal government must provide warning of an impending EMP; protect against, respond to and recover from its effects; and prevent adversarial events through deterrence, defense and nuclear nonproliferation. Coordination sits with the Assistant to the President for National Security Affairs, who must report annually on progress, gaps in capability and how to address them. The Director of the Office of Science and Technology Policy coordinates research through the National Science and Technology Council.
The roles section gives tasks to the Secretaries of State, Defense, the Interior, Commerce, Energy and Homeland Security and to the Director of National Intelligence. Two tasks reach every agency of interest to this article. The heads of all Sector-Specific Agencies must, with the Secretary of Homeland Security, enhance information sharing with private-sector counterparts. And the heads of all agencies that support National Essential Functions must ensure that their all-hazards preparedness planning sufficiently addresses EMPs, including through mitigation, response and recovery.
The implementation section sets a chain of deadlines. Within 90 days, the Secretary of Homeland Security was to identify and list the national critical functions and associated priority critical infrastructure that, if disrupted, could reasonably result in catastrophic national or regional effects. Within one year of that list, the Secretary was to assess which of those systems, networks and assets are most vulnerable to EMPs. Later steps hang from those: a review of test data, a cross-sector plan to close the gaps, pilot tests of engineering approaches and a mitigation plan. Within 180 days of the order, FEMA was to review and update federal response plans to account for EMPs. Within 180 days of the completion of that review, agencies that support National Essential Functions were to update operational plans documenting their procedures and responsibilities to prepare for, protect against and mitigate the effects of EMPs. The order also asked for quadrennial risk assessments on EMPs, with the first due within one year. It closes with the standard provision that it is to be implemented consistent with applicable law and subject to the availability of appropriations.
Much of Executive Order 13865 concerns the threat itself: warning of a high-altitude nuclear detonation, forecasting solar storms, research on how pulses couple to the power grid, and hardening of military installations. Those tasks go to the Departments of Defense, Commerce, Energy and Homeland Security. This article is about a narrower instruction in the order that landed on civilian agencies: agencies that support National Essential Functions had to make sure their all-hazards preparedness planning sufficiently addressed EMPs, and later had to update their operational plans to say how.
Our argument is that the instruction exposed an assumption built into continuity planning. A continuity plan gains much of its strength from distance. The alternate site is far from the primary one. The backup data center is in another region. Staff can work from other places. Suppliers are spread around the country. Distance works because the events a plan expects, such as a fire, a flood, a local power cut or a building closure, are local. The order describes an event that can affect large geographic areas at once. Against that event, distance buys less than the plan assumes.
The order’s structure made the agency task easy to postpone. Its implementation section runs largely as a chain in which later steps start from earlier ones: a list of national critical functions, then an assessment of what is most vulnerable, then a review of test data, then a plan to close the gaps, then pilots. The agency step sat near the end of the response branch, after FEMA’s review of federal response plans. An agency could reasonably wait for the national work before acting. We think the part of the agency task that concerns its own continuity plan did not need to wait, because it depends on the agency’s own records rather than on national findings.
We split what distance protects into four dependencies, and add a fifth section for the objections:
- Places: alternate sites and regions that are assumed to fail independently.
- Parts: replacement equipment and repair capacity that are assumed to arrive from somewhere unaffected.
- People: staff and contractors who are assumed to be available, reachable and able to travel.
- Warning and word: the channels through which the agency learns what is happening and tells others.
- Objections: low probability, high cost and the view that this is someone else’s mission.
The four dependencies do not overlap: each is a different thing the plan assumes it can get from elsewhere. Together they are the four things this article tests a continuity plan against. We describe what we think an agency should check in each, and we say where the order or its later implementation speaks to it.
Places: when the alternate site is inside the same footprint
A continuity plan for an essential function names where the function will run if its home is lost. The alternate location is chosen to be far enough away that the same event will not reach it. That choice is a judgment about the size of the events the plan expects. A plan sized for a hurricane may put the alternate site a few hundred miles inland. A plan sized for an earthquake may put it in another state.
What the order changed about footprint
The order’s definitions are blunt about scale. It says both high-altitude EMPs and geomagnetic disturbances can affect large geographic areas. The order does not give a radius, and we will not supply one. The point for an agency is that an EMP scenario is the case where the footprint is unknown in advance and may include both the home site and the alternate.
Cloud hosting raises the same question in another form. An agency that runs an essential application in two cloud regions has bought independence against a regional outage. Whether that independence holds against a wide-area event depends on where the regions are, how they are powered and what links connect them. Those are facts the agency may not hold. The contract, the provider’s published documentation and the agency’s own architecture records are where to look, and gaps should be recorded as gaps.
Common components as well as common geography
DHS made a further point in its 2020 status report on the order. It said it was developing new concepts and approaches to evaluate how simultaneous common component failures could cause cascading failures across critical infrastructures. Common components are a different axis from geography. Two sites hundreds of miles apart may use the same model of uninterruptible power supply, the same network gear and the same building controls. If an event damages a class of component wherever it is exposed, the two sites fail together however far apart they are.
An agency can test for this with records it already has, or should have. The asset inventory for the primary and alternate sites can be compared by make and model for the components that keep the function running: power, cooling, network, compute and the controls that manage them. Where the same component appears at both sites, the plan has a common dependency. It may be acceptable. It should at least be known.
Questions for the continuity manager
- For each essential function, list the alternate locations that would remain usable if an event affected a large region including the primary site.
- Mark the critical components that are the same make and model at the primary and alternate sites.
- Note whether the plan includes any location chosen for its independence from the regional grid, such as a site with its own generation and fuel.
These checks do not require an engineering study of EMP effects. They require the continuity plan and the asset inventory to be read together, which is a task an agency can do in-house.
Data copies and where they live
Data is a part of a continuity plan that plainly depends on distance. Backups are copied to another site or another region so that one event cannot destroy both the original and the copy. The same reasoning applies here. If the original and every copy sit within a footprint that a wide-area event could reach, distance has given less protection than the plan assumes.
There is a second point about data that has nothing to do with geography. A copy is useful only if the agency can restore from it. Restoring needs working equipment, the right software, the keys to decrypt the copy and someone who knows the procedure. A copy held by an outside provider also needs a working connection to that provider and a contract that still operates. We think an agency should know, for each essential function, which copy it would restore from in a wide-area event, what that restore needs and whether each need is met by something outside the footprint.
Offline copies deserve a separate line in the plan. A copy stored on media that is disconnected from networks is protected from some threats that reach connected systems. Whether it is protected from an electromagnetic event depends on where and how it is stored, which is a question for facilities and records staff, and the answer should be written down rather than assumed.
Parts: repair assumes a supply that is somewhere else
A continuity plan can say a great deal about relocation and much less about repair. The plan moves the function, runs it elsewhere and returns when the home site is fixed. The fixing is left to facilities teams, contractors and suppliers. The plan assumes those suppliers can ship replacement parts and send technicians from places the event did not touch.
What the DHS report named
The 2020 status report is direct on this point. In describing its work on criticality models, DHS said simultaneous common component failures could hinder response and recovery operations due to a lack of repair parts and personnel. That sentence names two of this article’s four dependencies. It describes a scenario in which the people who would normally fix the agency’s equipment are fixing everyone else’s, and the warehouse that would normally ship the part has shipped its stock to someone with a higher priority.
An agency cannot solve national supply. It can decide what it holds itself. For the components identified in the previous section, the question is whether the agency or its contractor keeps spares on hand, where they are stored and whether the stored spares would themselves survive the event. A spare stored in the same room as the equipment it replaces shares that equipment’s exposure.
Contracts that assume normal conditions
Maintenance and support contracts set response times. Those times are written for normal conditions and may be suspended by force majeure clauses in exactly the event this article is about. We think an agency should know, for each essential function, which contracts it would rely on for repair and what those contracts say about wide-area events. The reading is a task for the contracting officer and counsel, not a renegotiation. It tells the continuity manager how much of the repair plan is a promise and how much is a hope.
Restoration order
When parts are scarce, someone decides which system gets them first. A plan that has already ranked its essential functions can extend that ranking to the components each function needs. The ranking is useful in any shortage, whatever its cause. It turns a scramble into a list.
Fuel and the generator assumption
Where backup power at an agency site means generators, it also means fuel. A plan that relies on a generator relies on a fuel delivery contract once the tank runs low. In a local outage, the supplier delivers from outside the affected area. In a wide-area outage, the supplier’s own operations may be affected, and every customer with a generator is asking for fuel at once.
The facts an agency needs here are simple: how long each site can run on the fuel it holds, what the delivery contract promises and whether the generator’s own controls are electronic equipment that could be affected by the event. Those facts set how long the agency can count on a site before it must move the function or stop it. They belong in the continuity plan beside the list of functions the site supports.
- Name the components whose loss would stop an essential function until a replacement arrives.
- Record where the spares for those components are stored, and whether they share the equipment’s location.
- Read the repair contracts the agency would rely on for what they say about wide-area events.
People: the staff who carry the plan are in the footprint too
Every continuity plan names people. It names the emergency relocation group, the devolution team, the system administrators who restore services and the managers who make decisions. The plan assumes those people can be reached and can get to where they are needed, or can work from where they are.
Telework is a distance strategy
Remote work can spread staff across a wide area, and a continuity plan may count on that spread. If one office closes, staff work from home. That strategy depends on home power, home internet and the agency’s remote access services. A wide-area event can affect all three at once, and the staff’s own households will need attention before work does.
The order itself does not address staffing. The DHS status report names personnel alongside parts as a constraint on recovery. We think the useful step for an agency is to identify which essential functions could continue with a reduced team using local resources: paper procedures, a site with its own power, and decisions delegated in advance to whoever is present.
Delegations that do not need a network
Orders of succession and delegations of authority exist so that decisions continue when leaders are unavailable. They are written on the assumption that the successor can be told they are now in charge. If communications fail across a region, a successor may not learn that the person above them is unreachable. A delegation that takes effect on a stated condition, such as loss of contact for a set period, works without a message. A plan can set that condition in advance, and counsel can check that the delegation is valid as written.
Knowledge held by few people
Restoring a system after a broad failure may depend on knowledge held by a few administrators or a vendor’s engineers. We think an agency should know where that knowledge is written down, and whether the written version is available without the systems it describes. A restoration runbook stored on the file server it explains how to restore would be a gap in any outage. The gap is larger in a wide-area event because the people who remember the steps may not be reachable either.
Contractors and shared services
An agency’s continuity plan may rely on people it does not employ. Contractors run help desks, data centers and building systems. Shared service providers run payroll, finance and human resources systems for several agencies at once. Those providers have their own continuity plans, written for their own priorities. In a wide-area event, a shared service provider may be restoring service for many customers, and the order in which it does so is its decision unless a contract or an agreement says otherwise.
We think an agency should know which of its essential functions depend on a contractor’s or a shared provider’s staff, and what the agreement says about priority in a broad outage. Where the agreement is silent, the agency can at least record that its function is one of several competing for the same people, and plan for a longer wait.
Records of what was done
Agencies keep records of their decisions, and those duties continue in an emergency. A team working on paper with degraded communications still needs to record what it decided, on what information and on whose authority. We think a plan can include a simple paper log for that purpose, stored with the paper procedures. It costs almost nothing and it makes the later review, by the agency, its inspector general or Congress, possible.
- Identify the essential functions that could run with the staff likely to be available locally, using local power and paper.
- Check that delegations of authority take effect on a condition that does not require a message to be received.
- Keep restoration procedures for critical systems in a form that does not depend on those systems.
Warning and word: how the agency learns, and how it tells
The fourth dependency is information. An agency needs to know what has happened, how far it reaches and how long it may last. It needs to tell its staff, its partners and the public what it is doing. Both directions depend on communications infrastructure that a wide-area event may degrade.
Warning comes from outside the agency
The order assigns warning to others. It tells the Secretary of Defense to improve the ability to rapidly characterize, attribute and provide warning of EMPs. It tells the Secretary of Commerce to provide timely and accurate operational observations, analyses, forecasts and other products for natural EMPs. It tells the Secretary of Homeland Security to provide timely distribution of information on EMPs and credible associated threats to federal, state and local governments, critical infrastructure owners and operators and other stakeholders.
For a natural event, warning can arrive before the effects. A solar eruption can be seen before the disturbance reaches Earth: NOAA’s account of the 1859 Carrington Event says the flare was observed the day before its effects hit Earth. That makes the agency’s receipt of a forecast a decision point: what does the agency do in the hours between a severe space weather warning and its arrival? We think a plan can name a few actions for that window, such as confirming fuel for generators, staging staff, checking that backups are current and moving key decisions to people at sites with their own power. Each action depends on someone receiving the warning and having authority to act on it.
Word goes out through channels that may fail
The order tells the Secretary of Homeland Security to maintain survivable means to provide necessary emergency information to the public during and after EMPs. That duty is national. An agency has its own audience: its staff, the people who depend on its services and the partners who rely on its data. An agency’s notification system may be a software service that depends on networks and on outside providers. A plan can name at least one way to reach staff that does not depend on those services, and can say where staff should go or what they should do if they hear nothing.
Sharing with the private sector
The order tells the heads of all Sector-Specific Agencies, in coordination with the Secretary of Homeland Security, to enhance and facilitate information sharing with private-sector counterparts to enhance preparedness for the effects of EMPs, to identify and share vulnerabilities and to work collaboratively to reduce them. For an agency that is a Sector-Specific Agency, that is a standing duty. For other agencies, it is a reminder that their own suppliers and service providers hold part of the picture of what a wide-area event would do to the agency’s functions.
State and local partners
A federal agency’s essential functions may depend on state and local governments, and the reverse is true as well. Benefit payments flow through state systems. Public health data flows from local health departments. Emergency management runs through state and local offices that will be dealing with their own effects of the same event. The order tells the Secretary of Homeland Security to distribute information on EMPs and credible threats to state and local governments as well as federal ones. An agency that depends on a state partner can ask, ahead of time, what that partner’s plan assumes about the agency, and record the answer beside its own plan.
- Name who receives a severe space weather warning and what they are authorized to do in response.
- Name the channel that would reach staff if the agency’s network and its notification service were both unavailable.
- List the suppliers the agency would need to hear from first, and how it would reach them.
Objections: probability, cost and whose mission this is
Three objections to this analysis deserve an answer.
The event is unlikely
A high-altitude nuclear detonation over the country is a scenario of war. A geomagnetic disturbance severe enough to damage infrastructure across large areas is rare. CISA’s own page on EMP describes extreme EMP incidents as low probability and high consequence scenarios that challenge effective policymaking. An agency could reasonably decide not to spend much on either.
Our answer is that the checks above cost little and pay off in smaller events. Knowing which components are common to the primary and alternate sites helps in any failure of that component. A ranked restoration list helps in any shortage. Delegations that take effect without a message help in any communications outage. The EMP scenario is useful as a test because it removes all four dependencies at once, and the gaps it reveals are gaps for other events too.
Hardening is expensive
Protecting equipment against EMP effects can involve shielding, filtering and specialized design. Such work is expensive, and for an agency office building it is not the first thing we would fund. The order recognizes cost. It speaks of sustainable, efficient and cost-effective approaches and makes implementation subject to the availability of appropriations. The checks in this article are planning and records work. They do not require hardening anything, and they tell an agency where any hardening money would do the most good if it ever came.
Someone else owns this
The order assigns the large tasks to Defense, Energy, Homeland Security and the intelligence community. A civilian agency could read it as someone else’s work. Two provisions say otherwise. Section 5(i) puts the planning duty on the heads of all agencies that support National Essential Functions. Section 6(e)(ii) required those agencies to update operational plans documenting their procedures and responsibilities to prepare for, protect against and mitigate the effects of EMPs. Those are agency tasks. How far a given agency’s plan went is a question for its continuity manager and its records.
What the planning duty asked for
It is useful to be precise about what the order did and did not ask of these agencies. It did not set a technical standard for their facilities. It did not require them to harden anything. It asked that their preparedness planning sufficiently address EMPs, including through mitigation, response and recovery, and that their operational plans document procedures and responsibilities. That is a planning and documentation requirement. It can be met with the kind of work this article describes, and the evidence that it was met is a document an inspector general or an auditor could ask to see.
The order also left the agency some sequence to work with. Its own update came after FEMA’s review of federal response plans, which was due within 180 days of the order. An agency could reasonably wait for that review before writing its own plan, provided it kept a record of what it was waiting for and when the review arrived.
The policy may change
A final objection is that executive orders change with administrations, so work done under one may be wasted under the next. That is a fair description of how policy moves, and the hindsight section shows that continuity and critical infrastructure policy did move after 2019. Our answer is that the work described here does not depend on the order’s survival. An agency that knows its common dependencies, holds critical spares, has self-starting delegations and has tested a function without the network is better prepared for any broad outage, under any policy. What does depend on the policy is the paperwork that shows compliance, and that is why we suggest keeping a dated record of what the agency was asked to do and what it did.
Testing a plan against a wide-area event
A plan that has been read against these four dependencies still needs a test. An exercise is where assumptions show up, because participants are asked to act on them. We think a tabletop exercise built around a wide-area event can be run by an agency’s own continuity staff, without specialist knowledge of EMP effects, if it is designed around the dependencies rather than the physics.
Setting the scenario
The scenario does not need to say what caused the event. It needs to say what is unavailable. A useful starting point is a set of conditions: commercial power is lost across a region that includes the primary site and the nearest alternate; some electronic equipment at both sites has failed; public networks are degraded; staff can be reached only by the channel the plan names for that case; and suppliers report that parts and technicians are committed elsewhere. The exercise then asks the team to run one essential function under those conditions.
For a natural event, the scenario can start earlier, with a severe space weather warning received some hours before the effects. That version tests the decision points in the warning window: who receives the warning, who can act on it and what they do.
What to watch for
The exercise is most useful when it records where the plan points to something the scenario has removed. A step that says to contact the vendor, when the vendor cannot be reached. A step that says to fail over to the alternate site, when the alternate is in the footprint. A decision that needs a person who cannot be told they are now in charge. Each of these is a finding, and each maps to one of the four dependencies.
Findings should be written as gaps with owners, not as criticisms of the plan. Some will be accepted as risks the agency chooses to carry. Others will lead to small changes: a spare part bought, a delegation reworded, a paper procedure printed and stored at the alternate site.
Running one function without the network
The most demanding test is to run part of one essential function without the agency’s network for a short period, using paper and local resources. The aim is to learn whether the paper procedure exists, whether staff can follow it and how the work done on paper gets back into the systems once they return. That last step, reconciliation, is easy to leave out of a plan and hard to improvise afterward.
The test does not need to be large. One function, one site, a few hours. Its record is evidence that the agency’s planning addressed a scenario in which its systems were unavailable, which is the scenario the order asked agencies to address.
How the four dependencies fit together
Seen together, the four dependencies form one picture. Places, parts, people and information are each something the plan expects to draw from elsewhere when the home site fails. A local event leaves each source intact. A wide-area event may remove all four at once, and the plan’s fallback for one may depend on another. The alternate site needs staff. The staff need to be told to go. The restored system needs a part. The part needs a technician who can travel.
For that reason we think the order’s instruction to agencies is best read as a test of the plan’s structure rather than a request for a new annex. An EMP annex that sits beside an unchanged continuity plan adds paper. A plan that has been re-read against a wide-area event, with its common dependencies listed and its delegations made self-starting, is a better plan for every event.
The order’s own structure suggests the same reading. Its implementation chain starts with a list of what matters, the national critical functions and the infrastructure behind them, and moves to vulnerability, then test data, then mitigation. An agency can follow the same order on a small scale: list its essential functions, list the dependencies each needs, mark which dependencies a wide-area event could remove together, and decide what to do about the worst of them.
Evidence an agency can show
For an inspector general or an oversight committee, the evidence of this work is a short set of documents. A continuity plan that names its wide-area assumptions. A comparison of components at primary and alternate sites. A spares list with storage locations. A restoration ranking. Delegations with self-starting conditions. A test record for at least one function run without the network. These records need no classified information or specialist engineering. All of them can be produced by an agency’s own continuity, facilities and IT staff.
We would add one more record: a dated note of what the agency did in response to Section 6(e)(ii), and when. Policy changes over time, as the hindsight section shows. A record of what the agency was told to do and what it did is the basis for deciding what to do next.
Where to start
An agency with limited time can start with one essential function, the one whose interruption would do the most harm to the public. For that function, it can write down the four lists: the places it can run, the parts it cannot run without, the people it needs and the channels it depends on for warning and word. Then it can mark, on each list, the items a single wide-area event could remove together. That marked page is a better guide to spending than a general EMP assessment, because it names specific things the agency controls.
The second function can follow the same pattern, and the third. Patterns will repeat. The same generator contract, the same network provider and the same small team of administrators may appear under several functions. Those repeats are where a single change helps the most, and they are the items we would bring to the agency’s senior leadership first.
With hindsight
The first step in the order’s chain produced a public product quickly. CISA’s page on National Critical Functions says that in April 2019 it published its initial set of functions, later complemented by definitions for each. The set is organized into four areas covering connections, distribution, management and supplies. The DHS status report dated August 17, 2020 says that a month after the signing of the order, the National Risk Management Center, in coordination with Sector-Specific Agencies, established and released a preliminary set.
The next steps went more slowly. The same report says the order required prioritization of systems, networks and assets related to the newly developed functions, that the complexity of the task required new models and concepts, and that while the development of these models took longer than initially envisioned, additional time was needed to validate the concepts. It said the models had matured to a point that enabled prioritization of a limited set of systems for EMP vulnerability assessments, and that DHS was continuing to refine them to support the initial quadrennial EMP risk assessment and vulnerability test plans. It named energy and communications as the department’s priority. It said vulnerability testing of prioritized components would be conducted as funding becomes available.
On the planning duty this article is about, the report gave an answer. It said that in 2020 FEMA, in consultation with the federal interagency, determined that the Power Outage Incident Annex would serve as a baseline for emergency response planning associated with EMP, and that DHS and other federal departments and agencies had updated or were updating operational plans to align with the annex’s planning guidance. It also said DHS had developed initial planning guidance to assess operational plans so that federal mission essential functions would continue in a post-EMP environment, and that FEMA was developing an interagency EMP exercise to be conducted in fiscal year 2021, designed to challenge planning assumptions used by federal planners. We have not found a public record of that exercise’s results, and we do not report one.
Congress then wrote much of the order into law. Section 1740 of the National Defense Authorization Act for Fiscal Year 2020, Public Law 116-92, approved December 20, 2019, is headed Electromagnetic pulses and geomagnetic disturbances. It amended section 320 of the Homeland Security Act of 2002 to give the Secretary of Homeland Security and the FEMA Administrator duties on distributing information, response and recovery, research and a quadrennial EMP and GMD risk assessment, including a national exercise, not later than December 21, 2020, to test the preparedness and response of the nation to the effect of an EMP or extreme GMD event. Subsection (c) reached the agencies this article is about: not later than March 20, 2020, each agency that supports a national essential function was to prepare updated operational plans documenting its procedures and responsibilities for preparing for, protecting against and mitigating the effects of EMPs and GMDs. That put a fixed date in statute beside the order’s chained timing. The section also says it is not to be construed to affect the authority of the executive branch to implement Executive Order 13865. CISA’s EMP page says the order’s tenets are codified in Section 1740 and align with Executive Order 13744 on space weather, and as checked on November 14, 2026, it still describes the order and the department’s EMP strategy.
Nature supplied a test of the natural case. NOAA’s account of the Gannon solar storm of May 8-11, 2024 says at least eight Earth-directed coronal mass ejections erupted between May 8 and 11 and caused G5 geomagnetic storm conditions, the very top of the scale. It calls the storm quite possibly the strongest geomagnetic storm recorded this century and describes increased drag on satellites that led large numbers of them to maneuver at the same time. NOAA’s account does not describe a wide-area failure of ground infrastructure, and we do not claim one. The event showed that a severe natural storm can affect systems far apart at once.
Policy then moved. Executive Order 14239 of March 18, 2025, Achieving Efficiency Through State and Local Preparedness, directed the Assistant to the President for National Security Affairs to publish a National Resilience Strategy within 90 days; to review critical infrastructure policies, including National Security Memorandum 22, and recommend changes to shift from an all-hazards approach to a risk-informed approach; to review national continuity policies, including National Security Memorandum 32 of January 19, 2025; and to review national preparedness and response policies to move away from an all-hazards approach. It told the Secretary of Homeland Security to propose, within 1 year, changes to the policies behind the federal functions framework, which the order says includes the National Essential Functions and the National Critical Functions. Its purpose section names space weather among the risks that sound local decisions and investments are designed to address. The order does not list Executive Order 13865 among the policies to be reviewed, and it does not revoke or amend it. We do not report here what those reviews produced. What held up, as we read it, is the narrow point. The order’s instruction to agencies was planning work, the department responsible said that work had started, and the questions it raises about places, parts, people and information remain useful whichever risk framework replaces all-hazards planning.
Spotted an error? Tell us and we will check it and correct the article where needed.