"Why didn't people react to the warning?"

This is perhaps one of the most important questions emerging from last week's glacial rock avalanche and flood. A simple answer I keep hearing is: "Bro, because every single person thought it was the same as last year." That answer sounds simple, but it reveals something much deeper about how people interpret disaster warnings.

People did not necessarily ignore warnings. They have interpreted them through what they already know. They have experienced floods before. They have received warnings before. They have seen rivers rise before. And, often, those previous experiences have not resulted in catastrophic consequences. And often, those warnings had not resulted in anything catastrophic. So when another warning arrived, it was natural to think "This will be another flood like the ones we have seen before."

This is where the problem becomes more complicated. We learn from experience. It helps us recognise danger and decide what to do. But experience can also become a trap when the nature of the hazard changes. If yesterday's disaster becomes the template for understanding tomorrow's disaster, we may fail to recognise when something fundamentally different is happening. This is what seems to have happened in Nepal Flood.

This may be one of the biggest challenges facing early warning systems in a changing climate. We are becoming increasingly capable of detecting hazards. We have better satellites, better rainfall monitoring, better forecasting, better hydrological models and better communication technologies. The international commitment to Early Warning for All has also brought much-needed attention and investment to warning systems around the world. But perhaps we need to ask another question:

"Are we warning people about what is happening-or are we helping them understand what is different about what is happening?"

A warning is not simply information moving from authorities to the public. People understand warnings through their own experiences, what they know about hazards, and whether they trust the information they receive. This means that even an accurate warning may not lead to action if people understand the situation differently from what the warning intended.

When the hazard changes, risk also changes

As the climate and mountain environment change, hazards may not just become more frequent or more severe. They may also behave differently. Events that has not been anticipated will happen, or the event once happened separately may interact with one another, creating situations that we have not experienced before.

Yet much of the way we prepare for disasters is based on what has happened in the past. Communities remember previous floods. Experts use historical data to understand future hazards. Governments develop plans based on previous events. This makes sense when the future looks like the past. But what happens when it doesn't? This is where a simple distinction becomes important: Past event no longer will be reference for future forecasts. Or in other words- all these events can not be managed using standard disaster risk management logic, it needs to embrace "Uncertainty".

Risk and uncertainty are not the same thing

We are generally comfortable dealing with risk. If we have enough past information, we can estimate what might happen. For example, we may know that a river has reached a particular level several times before, or that there is a certain likelihood of flooding in a particular area. We may not know exactly what will happen, but we have enough experience to make a reasonable estimate. But, Uncertainty is different.

The event that happened in Nepal was in "Uncertainty Event". Because, We do not know enough to make a reliable estimate. We may know that something is changing, but not how it will develop, when it will happen, or what its consequences might be. Sometimes, we may not even know all the ways in which the hazard could unfold. That is what actually happened last week in Nepal. In these situations, the problem is not simply that we need precise risk estimates. Sometimes, we simply do not know enough yet. Think about driving along a road you know well. You have driven it many times. You know the corners, the traffic and what to expect. You can make decisions because your experience gives you a reasonable idea of what lies ahead. Now imagine that the road itself is changing while you are driving. Your previous experience is still useful-but it is no longer enough. That is the challenge we increasingly face with hazards.

What does this mean for early warning?

The problem is therefore not simply that people did not receive enough information. It may be that the information was interpreted using a mental picture of the hazard that no longer matched reality. This changes how we should think about early warning. We often ask: Was the warning accurate? Was it delivered on time? Did people receive it? These questions are important. But perhaps, we should also ask:

Did people understand what was different about this event?

If people hear "flood warning" and think about the floods they experienced last year, they may respond based on last year's experience. But if the conditions have changed, the same word "Flood" will describe a very different situation. This is why high risk events should be replaced by Uncertain events, and it should become part of how we think about early warning. We should not pretend that uncertainty can always be removed. Instead, we need to help people understand and interpret the event itself and communicate clearly about - what we know, what we do not know, and what is different this time. This does not mean making warnings more complicated. In fact, it may means making them simpler and more meaningful. Instead of only saying:

"A high flood risk is expected in the down stream."

We might need to communicate:

"This situation is different from previous floods. In upstream, We are seeing conditions we have not experienced before. We are not certain exactly how high will be flood, but the consequences could be serious. Here is what you should do now. Go as far as you can from Flood"

That is a different kind of warning. It does not pretend to have certainty where there is none. But it also does not leave people to interpret an unfamiliar situation using only familiar experience.

We cannot warn about everything

There is another reality we need to acknowledge. Early warning has limits. A warning cannot stop a mountain from collapsing. It cannot stop a sudden surge of water, rock and debris. And if there is almost no time between a hazard emerging and its impact, even the best warning system may leave people with very little time to act. This does not make early warning useless. It means we need to be honest about what early warning can and cannot do. We need better monitoring and forecasting. We need better understanding of rapidly changing environments. We need stronger communication between institutions. We need preparedness and anticipatory action. But technology alone will not solve the problem if our assumptions remain rooted in the hazards of the past. We can install more sensors, build better models and collect more data. But if we continue to assume that tomorrow's hazard will behave like yesterday's, we may still miss what is changing in front of us.

The real challenge

Perhaps, then, the lesson from Nepal is not simply that people did not listen to the warning.

Perhaps we need to ask three different questions:

Did we detect what was happening?

Did we understand what the signals meant?

And did the warning help people understand what was different about this event?

And when the world is changing, learning from the past without recognising its limits can create a dangerous assumption: Tomorrow will behave like yesterday.

The people who thought, "It happened last year; it will probably be the same this year," may therefore have been responding to more than a warning. They were responding to what their experience had taught them a flood looked like. The challenge for disaster risk management is to help people recognise when that experience is no longer enough. Because sometimes the most important message would not be simply:

"Something is going to happen."

It is:

"We do not know exactly what is going to happen, but we know enough that you need to act now to save your life"

Dr. Dhungana is the Disaster Risk and Uncertainty Communication Researcher