Southern Hemisphere weather forecasts suffer from observation gaps over Antarctica, but Japanese researchers demonstrated on September 28, 2026, that continuously assimilating hourly radar winds improves predictions for hazardous atmospheric rivers across mid-latitude regions.
The Antarctic Observation Gap and Southern Hemisphere Forecast Errors
Weather forecasting in the Southern Hemisphere is generally less accurate than that of the Northern Hemisphere. The underlying cause is that Antarctica and the surrounding Southern Ocean have relatively few weather observation sites. With limited information about atmospheric conditions in one of the planet’s most remote regions, forecasting models often struggle to accurately predict the movement of weather systems. This shortfall directly impacts the prediction of atmospheric rivers—long, narrow bands of moisture capable of delivering heavy rain, snowfall, flooding, and strong winds across Australia, New Zealand, South America, and Antarctica. Improving forecasts of these weather systems is essential for protecting lives and property, supporting aviation and shipping, and helping communities prepare for hazardous weather.
Traditional weather balloons and radiosondes provide essential atmospheric soundings globally, but launching them at high frequencies in the extreme cold of Antarctica is constrained by operational and financial barriers. Consequently, operational numerical weather prediction systems do not currently incorporate observations from the Antarctic radar.
Assimilating Hourly Winds From Syowa Station’s PANSY Radar
To overcome the limits of traditional radiosondes, a team led by Assistant Professor Kazutoshi Sato of the National Institute of Polar Research tested a continuous data stream. Alongside Professor Jun Inoue and Associate Professor Yoshihiro Tomikawa of NIPR, Dr. Akira Yamazaki of the Japan Agency for Marine-Earth Science and Technology, and Professor Kaoru Sato of The University of Tokyo, the researchers utilized wind data collected by the Program of the Antarctic Syowa Mesosphere–Stratosphere–Troposphere/Incoherent Scatter (MST/IS) Radar (PANSY) located at Japan’s Syowa Station.
The research team tested the impact of this continuous radar feed by running experiments through ALEDAS, a data assimilation and forecasting system managed by JAMSTEC on the Earth Simulator supercomputer. In a study made available online on July 11, 2026, in the Scientific Reports journal, Assistant Professor Kazutoshi Sato of the National Institute of Polar Research (NIPR), Japan, together with Professor Jun Inoue and Associate Professor Yoshihiro Tomikawa of NIPR, detailed these findings.
“Observations from the Antarctic radar are not currently incorporated into operational numerical weather prediction systems.”
Asst. Prof. Kazutoshi Sato, National Institute of Polar Research
Measurable Gains in Moisture Transport and Model Performance
The inclusion of continuous radar winds produced improvements.
“Although radiosonde observations are routinely performed to improve weather forecasts, high-frequency launches are challenging in extremely cold environments.”
Prof. Jun Inoue, National Institute of Polar Research
Operational Integration Challenges for Polar Weather Models
Despite the demonstrated reduction in forecast errors, continuous polar radar streams remain absent from standard daily meteorological operations. The findings published in Scientific Reports highlight that while continuous radar measurements have value for weather forecasting, they are not currently incorporated into operational numerical weather prediction systems.