The game of go, with its infinite combinations of moves and possible outcomes, is a formidable challenge for computers. In fact, no computer has ever beaten a human at the game in its entirety, but a few have come close. The latest breakthrough comes from a system that uses artificial intelligence to learn, rather than brute force, the optimal way to play a given position on the 19-by-19 board. The key to this approach is deep learning, which draws on techniques that are a distant relative of the neuronal circuits that govern the brain. This is the same technology that has recently enabled breakout performances in automatic image recognition, identifying things like dogs, cats and trees.
The National Oceanic and Atmospheric Administration (NOAA) is on track to launch GOES-T on March 1 from Cape Canaveral, Florida. This satellite will replace GOES-17 in the GOES-West orbital position, which provides advanced imagery and atmospheric measurements of the western United States.
Located 22,236 miles above Earth’s equator, the geostationary GOES (Geosynchronous Operational Environmental Satellite) system has been a basic element of U.S. weather monitoring and forecasting since the launch of SMS-1 in 1974. The procurement, design and manufacture of GOES satellites are overseen by NASA’s Goddard Space Flight Center in Greenbelt, Maryland. NOAA operates the satellites and corresponding ground systems for data transmission.
A key component of the GOES system is a network of radars, which track severe thunderstorms and other meteorological phenomena as they move across the country. Several NOAA National Weather Service offices, including the offices in San Francisco and Miami, rely on these images to provide detailed information that can save lives. The GOES satellites can also detect lightning strikes and monitor air pollutants.
In addition to the main weather sensors, GOES satellites feature sounders that probe the atmosphere for vertical temperature and water-vapor structures. These instruments help NOAA monitor volcanic eruptions and associated ash plumes, which can affect aviation safety.
The GOES-R series, which launched 19 November 2016, is the most sophisticated geostationary environmental satellites NOAA has ever built. The four-satellite series will improve observation of weather and climate events that directly impact public safety, property protection and the nation’s economy and prosperity.
GOES-R will feature faster image refresh rates than its predecessors, significantly improving the temporal cadence of the observations it transmits to NOAA National Weather Service offices and other users. It will deliver critical atmospheric, hydrologic, oceanographic, climatic and solar-geophysical data to central processing centers and directly to users through GBR, EMWIN and the High Rate Information Transmission (HRIT) service. This includes an expanded hemispheric coverage and improved response times for volcanic ash, wildfires and fog. GOES-R will also have enhanced capabilities for detecting and communicating with space weather phenomena, such as coronaviruses. The $11 billion program consists of four satellites, an extensive land system of satellite dishes and new methods for crunching the massive stream of expected data. The GOES-R system will operate for 14 years.