Meteorologists have an extremely
influential and important job. Many
people check the weather first thing in the morning and may continue to check
the weather throughout the day. It is
crucial that the forecast is correct, or else future forecasts may be ignored
or discarded by viewers. If the
meteorologist “lies” to you with one forecast, why should you believe him or
her during the next forecast? This eventual
lack of belief could lead to safety issues when warnings are ignored. For this reason, meteorologists are under a
great deal of pressure to present an accurate forecast. Unfortunately, creating an accurate forecast
is not as easy as it may sound. An
incredible amount of data and uncertainty are involved with these
forecasts. It may seem like a bit too
much to ask for someone to predict the future with 100% accuracy, but
nonetheless, this is what the public demands.
Many statistics and other facts will be
presented in this article to help discover the accuracy of weather
forecasts. There are limitless amounts
of statistical data available to show the accuracy of the forecasts. People’s perceptions and feelings regarding
these forecasts will also be analyzed to better determine why weather forecasts
are often considered to be inaccurate or invalid. Finally, the difficulties of creating weather
predictions will be discussed in order for the reader to gain a greater
understanding of the field. Hopefully
the reader will realize that predicting the weather is not just a coin flip or
simply looking at computer models.
There are several possible methods of
approaching this topic since there are so many different aspects involved. It was determined that investigating the
accuracy of forecasts should be the foremost goal. This way, it will be possible to discover
whether or not people’s anger and frustration with meteorologist is justified
in any way. Since forecasts are issued
by countless forecasters many times a day all across the globe, there is
essentially an infinite amount of statistics to show the accuracy of forecasts.
One of the most important aspects of a
weather forecast is the precipitation aspect.
For many people, it does not matter what the temperature is, they just
want to know if it will rain or snow or do nothing. In one specific study, weather forecasts for
a long period of time, from August through September, were analyzed. In this analysis, it was determined that the precipitation
forecasts were extremely accurate. Out
of the 134 days observed, it only rained once when a 0% chance of rain was
predicted. The other observations show
that the other forecasts (for 10%, 20%, and so on) were also quite accurate. For example, it only rained once out of the
ten times a 10% chance of rain was given.
This means it rained 10% of the time when there was a 10% chance of
rain, which is about perfect. Some other
numbers were further off, most notably the 50% chance, since it rained all four
times that the forecast called for a 50% chance of precipitation. Since there were only four scenarios when a
50% chance of rain was in the forecast, it is difficult to come to a solid
conclusion with that. It would be better
to look at the entire picture, which shows that the forecasts were relatively
accurate. Of note, the forecasts became
slightly less accurate when the forecast was further out. In other words, a forecast of 80% chance of
rain nine days in advance is going to be much less accurate than a forecast for
one or two days in advance. There are
several reasons for this lack of accuracy as time goes on, which will be
discuss in latter sections of this article.
Although precipitation is important to
people, temperature also plays a vital role in people’s lives. Similar to the precipitation observations,
there is a well-defined trend of decreased accuracy as time goes on. The forecasts for the first few days were
quite accurate and consistent. Figure 5
best shows this accuracy with several graphs indicating the amount of times the
forecasted temperature was a certain amount of degrees off from the observed
temperature. During the first for
forecast days, it is clear that the actual temperature was typically zero, one,
or two degrees off from the predicted temperature. This shows an incredible accuracy which many
people may not be aware of. Nonetheless,
the forecast accuracy does decline with later days in the forecast. For example, the consensus with day nine is
that the forecast has little to no accuracy at all. The predicted and actual temperatures varied
by as much twenty degrees. While these
may seem very costly and potentially dangerous, it is important to realize that
the forecast was ultimately accurate within a few days. This means that if people understand the
inaccurate nature of longer term forecast, then they will not be as
disappointed when the forecast is relatively far off. Instead, the focus should be on short term
forecasts (within about five days) in order to receive the most accurate
forecasts.
In addition to the important aspect of
accuracy, it is also vital to understand why people want to know the weather,
as well as what is most important to these people. In an expansive survey completed by the
American Meteorological Society, many of these various questions were
answered.