Spangler Subaru Blog

This time last year I penned a blog detailing the amount of snowfall Johnstown receives yearly compared to other areas. In it, I gave a brief yet concise expository on the history of weather prediction. When I say brief, I mean about a Bachelor’s degree worth of information in two paragraphs. Nonetheless, lately I have heard numerous people comment on the efficacy of “the weatherman” and his predictions after temperatures ranged colder and snow accumulated higher than expected. Meteorology and atmospheric forecasting are in no ways an exact science. There is a degree of error associated with forecast models, and I think it important to revisit weather forecasting so we can understand why, exactly, weather patterns never quite match forecasted conditions.

Out of my deep respect for weather forecasters, I will be using the term meteorologist to describe their position moving forward. I find using the terms weatherman or weathergirl to be somewhat derogatory for the length of schooling and amount of daily pressure (no pun intended) they withstand.

I will not bore you with the history of weather forecasting (which dates back to the Upanishads in Ancient India), but instead we will explore the fundamental theories behind numerical weather prediction that aid meteorologists in selecting forecasted weather patterns.

Let’s back up a bit before we get into that and describe what an average day might look like for a meteorologist. If he/she is anything like me, the morning probably has a bright outlook of coffee with a side of coffee. Of course, some early morning meteorologists probably arrive at the office around 4 am, there is a sufficient enough reason for their early rising. Upon arrival, a meteorologist will likely shack up with the local computer terminal to study possible scenarios and view different maps. He/she is looking up data collected globally and spread over global forecasts, then compacting that data into a microstudy of the local region. From there, quite a few models are generated for what the forecast could be, and he/she needs to make a judicious decision on which forecast he/she thinks is most likely to occur. This means the meteorologist is mentally choosing one forecast out of x models generated.

From there, the forecast is compiled and we see the meteorologist’s inexplicable lively face (because hey, I have dark circles under my eyes and I wake up two hours after his/her shift has started) on the broadcast giving us the local forecast. This bright visage is a premier in every household that is often heavily scrutinized, sometimes eliciting strong reactions from viewers if the forecast is not perfect. I have worked in retail before and I know how difficult it is to maintain composure when under pressure from an unhappy customer, and I can say I have never seen a meteorologist who looks like he/she is having a bad day.

After the forecast has been given, the meteorologist will then retreat back to his/her computer to study more data and adjust the day’s forecast as necessary. He/she may also partake in other tasks such as engaging with the community online, appearing in other news stories for the network, making appearances within the community, preparing special weather statements and alerts, communicating with other meteorologists and scientists across the globe, or, if he/she is reporting in the south, inserting him/herself in the middle of hurricane because why not?

While that may outline the strenuous day a meteorologist may have, it still does not completely explain why the forecasts never tend to match up with the actual path of the storm. That is when we look into numerical weather prediction. This science is rather confusing (and that means I do not entirely understand it), but I will do my best to describe how this process works, so you may better understand why weather prediction is never perfect.

Weather prediction analyzes a bunch of complex and difficult to measure data. It starts out by collecting data from the atmosphere and then using that data to solve equations under fluid dynamics and thermodynamics. Within these equations are the primitive equations­ – three rudimentary equations used to help solve future predictions in the fluid state of the atmosphere. From there, these are mixed with the ideal gas law to identify what may happen to density, pressure, potential temperature scalar fields, and the air velocity vector field. The predictions this creates is known as a system called the atmospheric model. These equations are so complex, if they were done by hand, it would literally take a roomful of mathematicians to solve.

Fortunately we have modern computing to put this together so that one individual can remedy these equations, and quite efficiently. Unfortunately, these equations can only help solve the patterns of a macro system. When it comes to processing fluid dynamic equations to solve patterns locally, the equations do not allow such micro targeting of air velocity in say a 1-mile square area.  While that may seem a large area to the unaided eye, it is a spec when looking at a storm that might encompass a 100,000-mile square area at any given time.

Therefore meteorologists need to dive into statistical models that coalesce climate information, numerical weather predictions, and surface observations to basically guess the future of weather. For those of you who have never taken a statistics class, statistics is not an absolute math. Instead, it gives the probability that a certain event or action may occur. Sometimes statistics will give us a near absolute answer, such as the probability of a football team to win a game when they enter the 4th quarter winning by 40 points, or it may give us a total equivocation, such as the probability of a coin landing on heads. What makes this calculation even further stressed, is that the atmosphere is extremely chaotic at all times. Minor variations in wind can skew the data so much that it is the difference in 5 inches of snow and a light dusting. Statistical modeling and the ability of the meteorologist to select the correct forecast takes a truly skilled, undaunted human being.

But even statistic modeling is not the only area where a meteorologist can falter. Keep in mind that global forecasts are conducted and consolidated with the work of many people around the globe. Weather balloons and ships collect the data, special weather labs need to decipher and put the data together, and they need to disseminate that to every meteorologist who is developing a forecast.

That does not mean all meteorologist get a break, though. For example, if a meteorologist (or maybe he is a weatherman in this case) forecasts a sunny day with 60 degree weather, but instead there is a terrible snowstorm, I am pretty sure he can be booed off the set. Sometimes meteorologist obscure details in order to fear monger and gain more viewers. A recent example of this occurred in Tampa Bay, FL. Many areas to the far north in Florida were, in fact, under a winter weather watch (or similar). However, Tampa Bay was not, despite many local forecasters claiming that snow might happen. I am sure some far ranging statistical models may have shown trace amounts of snow, but in the end Tampa Bay never saw a single snowflake and its residents had nothing to worry about.

That is just showbiz. The majority of those meteorologists are all very competent individuals; they just obscured stories slightly in order to help their networks gain a few more viewers. As a general rule, meteorologists will not mis-predict 6 inches of snow on purpose to help ratings. Remember, they want your viewership long-term and have to be as accurate as possible to retain your loyalty.

This was a little lengthy and it more than likely confused you if you read through the entirety of the blog. If you do not have a good grasp on it or did not read it, then you can boil the whole thing down to a simple common sense point: no one can predict the future.

We may know a lot more about numbers, relativity, physics, and the like, but at the end of the day we still have no clue what will happen tomorrow. I am thankful that I can watch someone reliable put their best foot forward in predicting tomorrow’s weather. It may not be exact, but I am happy that I can put full confidence behind what I wear to work, and you should too. 

Categories: Green, Other
; ;