
Last week we learned about the science behind weather forecasting when Dr DeMaria (Mark) spoke to RCFC – What’s New in Weather Forecasting? Dr DeMaria is a Senior Research Scientist in the CSU Institute for Research in Atmosphere (CIRA).
First, Mark’s background, and why he chose a position in theoretical science. His first NOAA job was with the Hurricane Research Division located in his hometown (Miami, FL) from 1987-1995. He initially was eager to be part of a hurricane flight. His flight was part of a two aircraft mission into Hurricane Hugo (9/1989), which was initially underestimated (category 3 updated to 5). The other aircraft experienced severe turbulence (5G down and 3G up), way above tolerances of the NOAA-P3 aircraft. A slide of the inside of the aircraft showed extensive damage. Mark’s aircraft also experienced turbulence but not as bad as the other one.
So, why is public weather information thought to be wrong 50% of the time? For one, TV forecasts have at times become part of the “entertainment”. But present-day weather forecasting is based on science which is most accurate when it can be quantitated.
There are multiple influences on weather that vary from the small and local (eg, altitude and topography) to large (the circumference of the earth and the jet stream) and these all vary widely in their influence and by large orders of magnitude. As mentioned, today’s forecasting are quantitative (numbers that are assigned to temperature, pressure, wind speed and direction, as well as water content and chemical makeup of the atmosphere). These numbers are entered into equations that are fed into computers.
Next, Mark showed a historical timeline of atmospheric observations in the US from info obtained from kites in the 1700s to the latest upgrade to weather satellites (2016). The development of radar used in WWII was applied to weather observation - a major advance. Today the data network includes surface and ocean observations, upper air measurements (specialized aircraft and balloons) and a vast array of weather satellites.
Weather forecasting is quantitative because the atmosphere and oceans are composed of fluids that obey physical laws. The first computer (ENIAC) was applied to weather modelling and involved US scientists and the military in a Manhattan-like project. This produced the first numerical weather prediction model (NWP) in the Spring of 1950. We all are aware of the evolution of computer power and the miniaturization that has occurred since then. Using hurricane prediction as an example of the application we can see how improved pathway identification can save lives and direct evacuations.
But will predictions ever be perfect? At this point Mark introduced the concept of Chaos theory and how it applies to weather prediction. Chaos theory tells us how small initial changes/influences can grow and confuse over a short period of time. Nonetheless, regions where severe weather potential is likely to occur can be predicted accurately as far out as 5 days. This has led to the science of estimating forecast uncertainty (all those different lines we see on a prediction of hurricane path).
However, forecasts from the National Weather Service are not useful if misunderstood by the public. As always, good communication is essential. Remember President Trump’s inaccurate prediction of a hurricane pathway in 2019?
What is the future of weather prediction? Trends visible today include ever- improving data processing, AI which can review massive amounts of data faster than a room full of scientists, unmanned weather aircraft, lower cost private sector involvement and a better understanding of the influence of our rapidly warming climate.
Lots of great information led to a short Q & A.