Go (pronounced go) is a board game that combines strategy and tactics with intuition and imagination. It is played in East Asia, but its popularity is spreading to the West. In the 18th and 19th centuries, Go was introduced to Europe via China, Japan, Korea, and the German-speaking countries. Its popularity grew rapidly after the 1905 publication of a book on the game by the Japanese-born German scientist Oskar Korschelt. The game has since become a major international pastime, with tens of thousands of tournaments worldwide and several professional Go players in the United States, China, and Korea.

Unlike chess, which has 64 squares, the number of spaces in go is much greater and there are more possible moves on most turns. This creates many more complex strategic decisions, making it difficult for computers to evaluate and compare all the options in a given situation. However, advances in computer technology are increasing their ability to solve complicated strategic positions. For example, in the fall of 2017 an artificial intelligence (AI) system developed at IBM defeated the world champion of human-computer Go in a four-match match.

A team of scientists led by Dr. David Phillips, a NASA Goddard Space Flight Center (GSFC) chemist, has used AI to analyze and interpret images produced by GOES-18’s primary instruments—the Imager and Sounder. The GOES series of satellites are geostationary, meaning they rotate in the same spot in Earth’s orbit and can be observed at all times from locations around the globe.

The GOES-18 Imager and Sounder instruments both use a 31 cm Cassegrain telescope with two-axis servo-driven, gimballed mirror systems. These enable the sensors to scan both global scenes—Earth’s full disk—and close-up observations of severe storms. The Imager captures radiation that is reflected off of clouds and the surface of the Earth, while the Sounder measures the amount of energy and direction of the Sun’s electromagnetic (EM) field as it passes through the atmosphere.

This information is used by meteorologists to make weather forecasts and detect conditions such as tornadoes, hail storms, snowstorms, and hurricanes. It also aids in detecting rainfall during thunderstorms, measuring snow accumulations, and monitoring the movements of glaciers and sea and lake ice.

By analyzing GOES-18 data, Hashimoto and his colleagues hope to improve climate models of the Amazon rainforest’s seasonality. While previous studies have relied on polar-orbiting satellites that pass over the area only once or twice per day, GOES-18’s geostationary orbit allows the sensor to keep its eyes trained on the region. It can collect new data every 10-15 minutes, allowing for more nuanced views of the rainforest’s seasonality than were previously available. This can help researchers better understand how the forest may impact the carbon cycle, and inform climate policy for reducing greenhouse gas emissions.