Image Up Advertising & Design

14 FOUR SEASONS BREEZE | DECEMBER 2021 "Understanding the Weather" – A Series by Mel Zeldin, Retired Meteorologist We often take for granted weather forecasts that are readily available via television, radio, internet, and a variety of other sources. So, let's go behind the scenes to better understand how these forecasts are made. To do that, we need to go back in time to get a sense of how forecasting skills have evolved. Believe it or not, weather forecasting dates back several centuries. In the early days of forecasting, weather forecasters, as such, relied primarily on local observations and a skilled knowledge of local climate history. In general, such forecasts could probably be somewhat accurate for a day. By the middle of the 19th century, with the development of the telegraph, weather forecasters of the day could then get information from more distant locations, and from that, they could chart a "weather map" of sorts, plotting the data they received. And from these maps, they could get a broader picture of overall weather conditions in the present, and the accuracy of one-day forecasts improved. Another major improvement occurred in the 1920s when a group of Norwegian scientists developed new understandings of cold fronts and movement of air masses. From that, forecasters could better understand the existence of high and low pressure systems, their movements, and the causes of precipitation. These innovative findings allowed for forecasts out to about two days. Then, in the 1930s, the development of radio technology allowed for the innovation of radiosondes. These are balloon-instrumented packages sent upward into the atmosphere, sending data back to the host station from many levels of the atmosphere. These data provided more than just surface observations, but conditions about temperature, humidity, and other parameters. Also, the change in location due to winds, as the balloon-package rose higher, allowed for calculations of the direction and strength of the winds aloft. In the 1950s, with the development of computer technology, scientists were able to develop equations of the physics of atmospheric behavior in which surface observations and radiosonde data could be incorporated as inputs. These initial efforts were so successful, they became the basis upon which today's numerical atmospheric modeling now exists. When I started doing weather forecasts in the 1960s, computer models extended out to three days. One- to two-day model outputs were reasonably good, but somewhat iffy for the third day. Today, models are more complex and require supercomputers to solve the necessary equations to predict the future. The input data to these models are much more robust as well, with progress in satellite imagery and other data which can readily be incorporated into the models. Models serve a variety of purposes, from smaller, more regionalized outputs, to full global projections. Today's weather forecaster has a variety of model outputs from which to help with a local forecast. But the models are just one tool in the toolbox. Knowledge of local climatology is still important, as it was centuries ago, along with the availability of satellite imagery. The forecasting today is very good out to about three days, and reasonably good out to about seven days. Extended range forecasts beyond seven days are not very good because the equations used in the models do not fully encapsulate all the atmospheric dynamics and the fact that the atmosphere is extremely chaotic in nature and may never be fully captured by equations. Skill in interpreting the models, applying the results to local climatology, and sometimes just looking at the sky are the foundation for a forecast. So the next time you watch the weather segment of a news broadcast, you'll have a better understanding of what goes on behind the scenes of a forecast. HOW FORECASTERS FORECAST Launch of a radiosonde National Weather Service supercomputer (which can do in one second what it takes 2.7 hours on a laptop)

• Cover