Earlier this month, Typhoon Hinnamno struck the Yeongnam region.



Heavy rain fell in Pohang, and the shopping district off the coast of Busan was devastated.



Typhoons Muipa and Nanmadol, which occurred later, evaded the Korean Peninsula, but caused damage one after another in China and Japan.



Accurate forecasting is important to reduce typhoon damage, but how accurate is the current typhoon forecast?



There are apps that people see a lot these days.



It is a numerical forecasting model that predicts the weather numerically, and it shows the results calculated by a supercomputer without human judgment.



Typhoons travel north along a huge mass of air called the North Pacific High Pressure.



Therefore, it is important to get the location of the North Pacific high pressure for accurate typhoon forecasting. When I looked at the errors of the numerical forecasting models of each country, Europe has the smallest error, and Korea is a little more accurate than the United States and Japan.



You'll probably feel that the forecast is right, as many apps use mainly the European model.



However, this is not always true either.



At the beginning of the typhoon, Nan Madol, which hit Japan this week, Korea predicted that the typhoon would head northeast, but the European model gave a different forecast that it would head to China.



The reason why world-class numerical forecasting models have such large errors is that typhoons are so complex.



Many people think of a typhoon as a round tornado, but the actual structure of the section in the typhoon looks like this.



Below the tornado, rain cloud bands of this shape are distributed, and the area below the band is raining heavily.



Although we are in the same area as typhoon, it rains in some places, and it doesn't rain in the vicinity, because of this very structure of the typhoon.



The reason why typhoon forecasting is difficult is because there are so many variables.



Typhoons Hinnamno, Muipa, and Nanmadol all had different routes, and Hinnamno came to Korea along the edge of a huge mass of air called the North Pacific high pressure.

Because of the weak west wind, Muyipa went up to China without turning to Korea.



Nan Madol was pushed to Japan by the cold, dry air from the north.



If you get all of these variables right, you can know the path of the typhoon.



What was it like 10 years ago?



In 2010, even if the course of the typhoon was predicted three days before the arrival, the error was 349 km.



Since the width of our country from east to west is about 300 km, to put it simply, it was difficult to determine whether a typhoon would go to the East Sea or the West Sea three days ago even at this time.



However, 10 years later, in 2020, this error has been halved, 173 km.



Now, if it's three days before the typhoon comes, there's little chance that the direction will be wrong.

Since typhoons usually have a radius of more than 200 km, if they were already in the impact zone in the forecast three days ago, it is very likely that they will actually be affected.



The typhoon observation performance was improved with the Chollian 2A satellite, and the calculation power became faster with the introduction of supercomputer No. 5, so this error could be reduced.



Nevertheless, scientific limitations still remain.



Even if you predict the central location of a typhoon one day in advance, the error is an average of 93 km.



It is not known exactly whether they will go to Busan or Sacheon, Gyeongsangnam-do, until the day before the typhoon landed.



This is the reason why you need to see the minute-by-minute weather information until the moment a typhoon approaches.



(Planning: Jung Yu-mi, Composition: Shin Hee-suk, Video coverage: Jeon Gyeong-bae, Park Hyun-cheol, Video editing: Kim Jun-hee, CG: Ban So-hee)