Ask anyone who has worked a major disaster what broke first, and you will almost always hear the same answer: communications. After twenty five years working in emergency management and public safety communications, I have come to see this not as a series of unlucky accidents but as a predictable pattern with predictable causes. The good news is that predictable failures can be planned for. This article is my attempt to put both halves of my professional life on one page: what goes wrong, why it goes wrong, and what an emergency manager can actually do about it, starting this month.

The oldest finding in the after action file

Read after action reports from American disasters across the last several decades and a theme repeats so consistently that it has become almost a cliche in this profession: communications failed, communications were overwhelmed, agencies could not talk to each other, the public did not get the warning. Hurricanes, wildfires, tornado outbreaks, mass casualty incidents, urban search and rescue operations. Different hazards, different states, different decades. Same finding.

The reasons cluster into a handful of recurring categories:

• Overload. Systems engineered for a normal Tuesday get hit with many times their designed load. Cellular networks congest. Radio channels stack up with traffic. Dispatch centers drown in call volume. Nothing has to break for communications to fail; demand alone can do it.
• Incompatibility. Neighboring agencies arrive to help and discover their radios cannot talk to each other. Different bands, different systems, different encryption, no shared channels programmed, or shared channels programmed but never exercised.
• Destroyed infrastructure. Towers, power, backhaul, and generator fuel are all physical things sitting in the same disaster area as everything else. The event that creates the need for communications is often the event that removes the means.
• No shared plan. Even when the technology works, incidents fall apart because nobody decided in advance who talks to whom, on what channel, for what purpose. Everyone transmits; nobody coordinates.

Notice that only one of those four is purely a technology problem. The other three are planning problems, relationship problems, and governance problems. That should be encouraging, because emergency managers are in the planning, relationship, and governance business.

The emergency manager’s communications portfolio

Many emergency managers treat communications as somebody else’s technical specialty. I understand the instinct. Radio systems are genuinely complex, and I say that as someone who wrote a very long reference work on P25 systems. But the emergency manager owns a communications portfolio whether they acknowledge it or not, and it looks like this:

• Interoperability planning. Who needs to talk to whom across agency and jurisdiction lines, and what channels, gateways, or agreements make that possible? Your state has a statewide interoperability coordinator and a statewide plan. Know them. Your region likely has a tactical interoperable communications plan or its equivalent. Know it, and know what has actually been tested versus what only exists on paper.
• Incident communications planning. The discipline of assigning specific channels to specific functions on a specific incident, written down where everyone can see it. More on this below.
• Public alerting. Warning the population is an emergency management responsibility in most jurisdictions, and it has its own technology stack and its own failure modes.
• Backup layers. What happens when the primary systems degrade? Satellite, deployable assets, cache radios, amateur radio volunteers, runners with paper. Someone has to have thought about this before the lights go out, and that someone is you.

You do not have to be able to program a radio or engineer a tower site. You do need to understand your systems conversationally, well enough to ask good questions, spot weak assumptions, and advocate for resources. I write about these systems in depth elsewhere on this site, and I would encourage any emergency manager to build at least a working vocabulary. It changes every conversation you have with your technical partners.

The incident radio communications plan and the COML

The Incident Command System gives us a simple, powerful idea: before an operational period begins, someone writes down which radio channels are assigned to which functions, and everyone works from that document. In ICS this is the incident radio communications plan, commonly known by its form number, the ICS 205. The concept matters more than the form. Command talks here. Each division or group talks there. Air operations, if any, talks somewhere else. Everyone knows where to find help and where not to step on critical traffic.

It sounds almost too simple to matter. It is the difference between an organized incident and a shouting match. Most communications chaos on scenes I have observed traces back to the absence of exactly this document, or to a document written by someone who did not know the local systems.

That is where the Communications Unit Leader, the COML, comes in. The COML is a credentialed ICS position responsible for planning and managing communications on an incident: writing the 205, managing the communications unit, coordinating channel assignments, arranging for additional resources, and troubleshooting as things evolve. The training path runs through position-specific coursework and a performance task book completed on real incidents and exercises. I have been through that training pipeline myself, and I can tell you the scarcity is real: many regions have very few qualified COMLs, and some have none. Ask yourself right now whether you know, by name, who would fill that seat on your next major incident. If you do not have an answer, you have found your first gap.

The resilience ladder: PACE in plain terms

Military planners use a framework called PACE: Primary, Alternate, Contingency, Emergency. It is the most useful communications planning tool I know, precisely because it forces you to assume failure in advance. Applied to a typical American jurisdiction, the ladder looks something like this:

• Primary. Your land mobile radio system and the commercial phone and data networks. This is daily life: dispatch, cell phones, email, computer aided dispatch.
• Alternate. A second path using the same general infrastructure. Interoperability channels, a neighboring system via mutual aid agreements, priority cellular services for public safety, conventional backup channels behind your trunked system.
• Contingency. Paths that do not depend on your local infrastructure at all. Satellite phones and satellite data terminals, deployable assets like mobile communications vehicles and portable repeaters, state or federal caches you can request.
• Emergency. The last rungs. Amateur radio operators, who bring their own equipment, their own skills, and networks that function without any commercial infrastructure. High frequency radio that needs no towers. And, honestly, runners carrying paper messages, which has closed the loop in more real disasters than most people would guess.

Two rules make PACE actually work. First, every rung must be written down, resourced, and known to the people who will use it. A satellite phone nobody can find, with a dead battery and a number nobody has, is not a contingency. It is a prop. Second, you have to define the triggers: how do you decide the primary has failed, and who makes the call to move down the ladder? In a real event, the transition is the hard part.

Three rooms, one system: the EM office, the 911 center, and the radio shop

In most jurisdictions, disaster communications lives in three rooms that do not talk to each other nearly enough. The emergency management office owns planning and coordination. The 911 center owns call taking and dispatch, and in a disaster becomes both a critical node and a critical vulnerability. The radio system administrators, whether a county radio shop, an IT department, or a regional authority, own the infrastructure itself: the sites, the channels, the programming, the maintenance contracts.

Each room holds knowledge the others need. The radio shop knows which tower sites have aging generators. The 911 center knows what call volume does to their staffing and their phone trunks. The EM office knows the hazard scenarios and the mutual aid picture. When these three only meet during an actual emergency, you get exactly the failures the after action reports keep describing.

My strong advice to emergency managers: build standing relationships with both of the other rooms, and learn your radio system at least conversationally. Know whether it is trunked or conventional, what bands it operates on, roughly how many sites it has, what its coverage weak spots are, how it is backed up, and what its known single points of failure are. You do not need to be an engineer. You need to be a competent customer and a credible advocate. When the radio shop tells the budget office that a site generator is past its service life, an emergency manager who understands why that matters is the difference between funded and deferred.

Reaching the public: alerting is your job too

Everything above concerns responders talking to responders. The other half of the portfolio is warning the public, and in most jurisdictions that responsibility lands squarely on emergency management. The modern toolkit includes several overlapping layers:

• Wireless emergency alerts pushed to cell phones in a defined geographic area through the federal alerting infrastructure.
• The Emergency Alert System reaching broadcast radio and television.
• NOAA Weather Radio, still one of the most reliable last-mile warning tools in existence, especially for people asleep at 3 a.m.
• Outdoor warning sirens, which are outdoor tools, useful for people outside and widely misunderstood as an indoor warning system.
• Opt-in notification systems, which reach only the fraction of your population that has signed up, a fraction most jurisdictions overestimate.

Each layer misses somebody. The overlapping use of several layers is the design, not a redundancy to be trimmed. And every layer is subject to the same human constraint: alert fatigue. Send vague, overly broad, or unnecessary alerts and the public learns to ignore you, which spends down a trust account you will desperately need on the worst day. Tight geographic targeting, clear plain language about what to do, and discipline about when to trigger an alert are as much a part of alerting as the technology. So is practice: the person authorized to send a wireless alert should not be reading the software interface for the first time during a tornado warning.

Pre-disaster work that pays off

Almost everything that determines communications success in a disaster happens before the disaster. The highest-return work I have seen, across many jurisdictions and many years:

• Exercise your interoperability channels. Not once, on a signed memo, but regularly, on the air, with the actual agencies you would call for help. A channel that has never carried real traffic is a rumor, not a capability.
• Build and maintain a cache of radios. Programmed, charged, inventoried, with a checkout procedure, ready to hand to mutual aid partners or volunteers whose equipment does not match yours.
• Verify backup power at critical sites. Generators at radio sites and the 911 center, tested under load, with fuel contracts and refueling plans that assume roads may be bad.
• Formalize your amateur radio relationship. Hams have been passing disaster traffic for over a century, and organized volunteer structures exist in nearly every county. As a longtime Amateur Extra licensee myself, I will tell you plainly: the time to meet your local operators, sign an agreement, define their role, and integrate them into exercises is now, not when the cell network dies.
• Write the plan down and put it in reach. Channel plans, contact rosters, PACE ladders, and alerting procedures, printed, because the day you need them may be the day the file server is dark.

An assessment you can run this month

None of this requires a grant cycle to begin. Here is a working assessment you can run in the next thirty days with your 911 director and your radio system administrator across a conference table. Answer each question honestly, in writing:

• People. Who is our COML, and who is the backup? Who is authorized and trained to send a public alert, on every shift?
• Plans. Do we have a current incident radio communications plan template with our real local channels in it? When was our regional interoperability plan last updated, and last exercised?
• PACE. Can we name our primary, alternate, contingency, and emergency paths for (a) responder communications and (b) EOC-to-field communications? Are the contingency assets physically located, charged, and tested within the last quarter?
• Infrastructure. Which radio sites, and the 911 center itself, have backup power? When was it last tested under load? What is our single worst point of failure, and does the budget office know about it?
• Interoperability. When did we last pass live traffic with each neighboring jurisdiction on a shared channel? Which mutual aid partners can we not talk to today?
• Public alerting. What percentage of our population is on the opt-in system? When did we last send a live or test alert, and what did we learn?
• Volunteers. Do we have a current, signed agreement with our amateur radio group, and were they in our last exercise?
• Degraded operations. Have we ever exercised with the primary system deliberately turned off? If not, when will we?

Every “I do not know” on that list is a finding you got for free, before the disaster delivered it the expensive way. Turn the answers into a short list of gaps, put dates and names on each one, and you have a communications improvement plan that will outperform most consultant deliverables I have read.

Takeaways

• Communications failure in disasters is a decades-long, well-documented pattern, and most of its causes are planning and relationship problems, not technology problems.
• Emergency managers own a communications portfolio: interoperability planning, incident communications planning, public alerting, and backup layers. Owning it starts with learning your systems conversationally.
• The incident radio communications plan concept, and a qualified COML to build it, are the difference between organized incidents and on-air chaos. Know your COML by name.
• Plan communications as a PACE ladder: primary, alternate, contingency, emergency. Every rung must be resourced, tested, and known, with clear triggers for moving down.
• Public alerting is a layered system by design, and alert discipline protects the public trust you will need most on your worst day.
• The highest-return work is done in blue skies: exercised interop channels, ready radio caches, tested backup power, and signed agreements with amateur radio volunteers.
• No system survives everything. Plan for degraded operations as a normal operating mode, and run the thirty-day assessment with your comms partners this month.