[News Space=Reporter seungwon lee] The reliability of the Korea Meteorological Administration's weather forecasts is in serious jeopardy. The frequency of false forecasts has led to nicknames like "Guracheong" (a false forecast) and "False Forecasting Office," leading to a surge in public discontent. In particular, the common saying, "It always rains on the day of the Korea Meteorological Administration's athletic meet," has gone beyond mere humor and become a symbol of the agency's declining reliability.

The accuracy rate of precipitation forecasts released by the Korea Meteorological Administration from 2012 to 2016 averaged only 46%, meaning more than half were off the mark. Since then, the figure has not improved significantly, remaining at 39% in 2017, 48.3% in 2018, and 46.3% in 2019.

The World Meteorological Organization (WMO) evaluated the performance of the Korean Numerical Forecasting Model (KIM) in 2019, ranking ninth among 11 major countries. Only China and Russia recorded lower performance than Korea. These figures are an indicator of the objective reliability of the Korea Meteorological Administration's (KMA) forecasts.

Real-world examples further illustrate this crisis of trust. In the summer of 2016, forecasts of 120 mm of rain per hour in the metropolitan area were repeatedly off the mark, and a forecast of up to 300 mm on August 15th was also incorrect.

In July 2020, contrary to the heat wave forecast for the metropolitan area, the actual temperature remained around 25 degrees Celsius, and on the 30th of the same month, instead of a heat wave of 30-33 degrees Celsius, the Daegu and Gyeongbuk regions experienced heavy rain and unusually low temperatures of 19-23 degrees Celsius.

Typhoon Soulik's 2018 forecast also caused widespread confusion in the community, with its path and intensity forecasts being significantly off. Furthermore, forecasts for rain throughout the first week of August in the Gwangju and South Jeolla Province often resulted in hasty revisions on the same day, as clear skies persisted despite forecasts predicting rain throughout the first week of August.

In particular, the myth of a "rainy day" related to the Korea Meteorological Administration's own sports competition has been confirmed to be true. According to a Kyunghyang Shinmun article dated May 5, 1994, despite the Korea Meteorological Administration's forecast advancing the competition by a day, it rained on the day of the event, prompting the media to remark, "Even the Korea Meteorological Administration can't cut its own hair."

Since then, repeated instances of rain during the autumn sports festival have become a widely known joke among the public. In a 2020 broadcast of "You Quiz on the Block," the weather forecaster testified that it actually rained on the day of the sports festival, forcing them to hastily take down the banner.

The Korea Meteorological Administration (KMA) reports an annual accuracy rate (ACC) of 89.1-90.9%, but this is a national average and differs significantly from actual prediction accuracy (POD). As of 2021-2022, the POD stands at a mere 60-65%, equivalent to a "correct 6-7 out of 10" prediction, placing Korea in the bottom tier of the OECD compared to competing countries like Japan.

The low level of trust reflects public sentiment on the ground, with approximately 40% of public complaints being about forecast dissatisfaction.

The Korea Meteorological Administration (KMA) attributes the errors in forecasts to a complex mix of factors, including the ever-changing climate (abnormal weather and localized, guerrilla-like downpours), limited observation and forecasting infrastructure, and the limitations of the KIM (Korea Meteorological Information Model). The government is investing hundreds of billions of won in supercomputers, AI forecasting, and high-resolution models to improve forecast accuracy, but no significant results have yet been achieved.

To restore public trust, we must simultaneously improve the transparency of weather data disclosure, strengthen feedback systems based on actual measurements in each region nationwide, and innovate forecast delivery methods. In other words, we must prioritize communication with the public and building trust beyond mere technological development.