News Release

Scientists develop a new way to verify forecasts of where typhoons hit land

Peer-Reviewed Publication

Institute of Atmospheric Physics, Chinese Academy of Sciences

Typhoon’s landfall position

image: The new verification approach can help identify “near miss” cases in forecasts of a typhoon’s landfall position. view more 

Credit: Daosheng Xu

Landfalling typhoons can cause widespread damage, so accurately forecasting them is crucial. More specifically, the errors associated with a typhoon's landfall position are often more important than those related to the timing of landfall. However, the most widely used method to verify typhoon tracks (the traditional “point-to-point matching” approach) can be easily affected by the predicted moving speed of the typhoon. Consequently, a large track error can sometimes be obtained even if the predicted landfall position is close to the observed one.

To address this issue, Dr. Banglin Zhang and colleagues from the Guangzhou Institute of Tropical and Marine Meteorology have proposed a simple method to evaluate landfalling typhoon track forecasts. Their findings, recently published in Advances in Atmospheric Sciences, reveal that this new approach can lessen the timing errors when verifying a predicted landfalling typhoon track.

"With the traditional method for verifying typhoon track forecasts, the error in the typhoon track may derive from the error in the typhoon's moving speed error rather than its position error," explains Dr. Daosheng Xu, first author of the study. "However, the neighborhood strategy upon which our method is based verifies the forecast within a time window around each grid point, hence reduces influences of the timing error."

As mentioned above, the traditional point-to-point method for evaluating typhoon tracks is sometimes sensitive to timing errors and/or typhoon moving-speed errors. Hence, Dr. Xu and his colleagues proposed this "time neighborhood method" to solve the problem. By comparing the properties of this new method to those of the traditional approach by forecasting 12 landfalling typhoon cases, they found that the new method is not sensitive to the timing errors and that the difference between the new and traditional method is more apparent when the typhoon has a moderate movement speed (between 15 and 30 km h−1). There is hope, therefore, that this method might be able to provide meteorologists with more intuitive evaluations of typhoon track forecasts, which will be of broader benefit to the government and the general public.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.