Why Should I Care?

For anyone with experience making plans around a weather forecast, there is no doubt that you have wondered how sure the weather forecaster was about that forecast that seemed to be nothing like the day actually turned out. Of course there are many days when the forecast seems spot on (although we never seem to remember those). Now that blown forecast may be due to some lack of skill, but often if can be attributed to knowingly high uncertainty in the forecast when it was issued. So why in the world did they forecast a specific temperature or condition?

Well, there are a variety of reasons this may be the case – limited time, what folks are use to, keeping things simple, etc. And let’s be realistic, most folks don’t want to have to spend 30+ minutes trying to comprehend whether it will rain tomorrow or how warm it will be, that is why we have meteorologist – to make a best estimate. Now, your logical next question might be – ‘why do they have to estimate at all?’ Ah ha! That is the key question, and if the answer were simple then maybe it would not be the key question.

In most areas of science, just like most of the other realms around us, things are seldom certain. Whether it is a weather forecast or a report about a new ‘scientific finding’, while often presented as certainty, it is most often a best projection based on our current understanding. Some of the reasons for this uncertainty include:

• Sensitivity of calculations used in prediction to small fluctuations
• Lack of sufficient data in time or space domains
• Imprecise equations to represent the behaviors
• Incomplete understanding of the criteria that drive a situation

Example of forecast ranges instead of a specific value

Remember last year when we were discussing tropical cyclone forecasts and how the forecasts where in ranges not a specific number (see the picture to the right as reminder). Well all those items I mentioned above influence why we can’t reasonably predict a single value, and why even the ranges are sometimes incorrect. So now that you have seen that there is indeed a man behind the curtain and the great Oz is not real, what does it mean for you?

Simple really, when considering a forecast or projection of weather, climate, or other science, treat it like you would any forecast that impacts your life (economic, financial, health, etc.). Particularly for weather and climate when evaluating how it will impact you and/or your organization you need to place it in the context of two fundamental criteria:

1. What is your level of sensitivity to what is being forecasted?
2. What is the degree of uncertainty in the forecast?

The good news on the first question is that usually determining the answer is in your hands. How will the ‘potential’ of rain impact your day tomorrow, or how a hot summer ‘might’ influence your business? Coming up with the answer can take some time if more complex, but hopefully the data to some degree is in your control.

The second question can be a bit more difficult for you to get an answer to, but that is why the answer to the first question can potentially narrow the cases down where getting an answer to the second question becomes necessary. For example if you will be working tomorrow inside, what does it really matter if it rains while you are at the office? But when the information is critical to you and could influence your decision making it becomes important to understand the level of confidence you can place in that information.

For a weather forecast, you might read a forecast discussion (sample) at the National Weather Service to get a bit more detail behind the forecast and the level of certainty in a prediction. For a report about a scientific finding that could impact your business, get a copy of the actual study and read it. The more influence the forecast could have, the more time you should spend making sure you really understand the degree of certainty or uncertainty underlying the forecast. It may not always be a trivial exercise to get the information or comprehend it, but usually with a little effort you can greatly increase your understanding of the uncertainty and accordingly make better decisions.

Interested in learning some more about this topic? There is a book that you can read for free online called ‘Completing the Forecast‘. It talks about effectively communicating the uncertainty in weather and climate forecasts and covers in more detail the topics introduced here today.

Until next time!