Before the advent of modern weather forecasting using computer models, meteorologists often used analogs to find similar weather in previous years and use that to predict how the weather would change. A modern take on that is to use machine learning with big data to identify patterns in weather that could help predict severe weather up to five days ahead. This is the high-tech version of analog forecasting because it looks for similar patterns in weather to see how they change and uses those relationships to increase our ability to forecast severe weather before it happens. You can read more about how this works at https://www.meteorologicaltechnologyinternational.com/news/extreme-weather/scientists-use-deep-learning-to-predict-extreme-weather.html

Lt. John Meisenheimer, the Air Force weather liaison officer at Cape Canaveral Air Force Station, gave the weather forecast for the launch of Explorer 1 in late January 1958. Here he is pictured at the Cape in 1956. (Courtesy of John Meisenheimer)