Here's Why Hurricanes Are So Difficult to Predict

Tony Butt

by on

Updated 12d ago

Hurricanes are some of the most extreme natural phenomena on the planet. As we are seeing right now with Dorian, they can cause severe destruction of human habitats.

But these systems are also some of the most difficult things to predict; which makes things even worse for communities living on vulnerable coastlines. So what makes them so difficult?

Related content: Hurricane Dorian Forecast

You may recall, we even made a movie about getting the world's best surfers into the best waves on the planet.

There are several variables that are constantly changing throughout the life of a hurricane, for example:

The windspeed, including the strength of the gusts. (Remember, MSW has comprehensive wind charts, see HERE.)
The area covered by the storm.
Whether it is intensifying or weakening, and how fast it is doing so.
The trajectory of the system.
The speed of movement of the system.

Each one of these variables is just as important as the others, and they are all interdependent. For example, the trajectory of the storm determines whether it will end up being influenced by surrounding factors that either intensify or weaken it. And whether it intensifies or weakens will influence whether it grows or shrinks.

Here is our OBX hurricane forecast for tomorrow, Friday September 6. As you can see, the predictability scale on the right hand side is dipping from 80 to the plus side of 50. This is because some elements of the forecast are still uncertain due to the changing nature of hurricanes.

Here is our OBX hurricane forecast for tomorrow, Friday September 6. As you can see, the predictability scale on the right hand side is dipping from 80 to the plus side of 50. This is because some elements of the forecast are still uncertain due to the changing nature of hurricanes.

Hurricanes tend to be steered by larger atmospheric systems such as troughs, ridges, highs and lows. The intensity of the hurricane is influenced by things such as sea surface temperature (warmer sea = more energy = stronger storm), and the variation in wind with height (called shear: less shear = more convection = stronger storm). These ‘forcing mechanisms’ are themselves very difficult to forecast. Therefore, predicting the behaviour of an even smaller and more volatile system that depends on them is doubly difficult.

Atmospheric forecasts are done by the most powerful computers in the world. They make an approximation of the atmosphere using millions of discrete grid points, calculating things like pressure, temperature, windspeed and humidity at each of these points. The points in the model are typically a few kilometres apart (the resolution of the model).

Therefore, systems that are much bigger than the grid resolution are easier to predict that systems that are nearly the same size as the grid. For example, a giant, sluggish high pressure, thousands of kilometres wide will be easier to forecast than a tight, unstable hurricane a few tens of kilometres wide. In other words, the hurricane could ‘slip through the cracks’ in the model.

Forecast: Central Florida Hurricane Surf Report

The Earth’s systems are non-linear by nature. They are intertwined with feedback loops that make them fiendishly difficult to understand, let alone predict. With any natural forecast model, such as the atmosphere, small differences in the input lead to large differences in the output. These differences get proportionally larger and larger as the forecast length increases. A point is always reached whereby the inaccuracies in the input cause such a divergence in output results that you cannot predict any further into the future.

Different regions of the world refer to hurricanes in different ways. They're the same, whether a typhoon or cyclone.

Different regions of the world refer to hurricanes in different ways. They're the same, whether a typhoon or cyclone.

Hurricanes are more vulnerable to those changes in initial conditions than bigger systems such as mid-latitude depressions. This makes them much more difficult to predict. With hurricanes, it is not only more difficult to get input data that is accurate enough, but the small differences in the output can have big consequences:

The inputs to the model – the surrounding oceanic and atmospheric conditions that drive the storm – need to be accurately measured at regular intervals, otherwise the forecast will have no hope of being accurate. Observations come from buoy and ship reports, satellite data and planes that actually fly into the tropical storm. But these are difficult to obtain and are sometimes not enough.

The output of the model – a realistic-as-possible vision of the storm at some time in the future – needs to be more accurate than it does for bigger systems. This is because any slight change in the position or the intensity of the storm might make a massive difference in the consequences for people living on a nearby coastline. Whether a particular coastal town will be utterly devastated or miraculously saved, could depend on whether the system tracks a hundred kilometres north or south. If the system becomes really slow-moving, like Dorian did, then things become even more uncertain.