Uncertainty in Tropical Cyclone Intensity Predictions due to Uncertainty in Initial Conditions

• Published in: Advances in atmospheric sciences 2020-03, Vol.37 (3), p.278-290
• Main Authors: Wang, Chenxi, Zeng, Zhihua, Ying, Ming
• Format: Article
• Language: English
• Subjects: Atmospheric Sciences; Convective instability; Cyclones; Earth and Environmental Science; Earth Sciences; Ensemble forecasting; Fluctuations; Geophysics/Geodesy; Heat flux; Heat transfer; Humidity; Hurricanes; Initial conditions; Meteorology; Original Paper; Perturbation; Surface stability; Temperature differences; Temperature gradients; Tropical climate; Tropical cyclone intensities; Tropical cyclones; Uncertainty; Water vapor; Water vapor flux; Water vapour
• ISSN: 0256-1530; 1861-9533
• DOI: 10.1007/s00376-019-9126-6

Focusing on the role of initial condition uncertainty, we use WRF initial perturbation ensemble forecasts to investigate the uncertainty in intensity forecasts of Tropical Cyclone (TC) Rammasun (1409), which is the strongest TC to have made landfall in China during the past 50 years. Forecast results indicate that initial condition uncertainty leads to TC forecast uncertainty, particularly for TC intensity. This uncertainty increases with forecast time, with a more rapid and significant increase after 24 h. The predicted TC develops slowly before 24 h, and at this stage the TC in the member forecasting the strongest final TC is not the strongest among all members. However, after 24 h, the TC in this member strengthens much more than that the TC in other members. The variations in convective instability, precipitation, surface upward heat flux, and surface upward water vapor flux show similar characteristics to the variation in TC intensity, and there is a strong correlation between TC intensity and both the surface upward heat flux and the surface upward water vapor flux. The initial condition differences that result in the maximum intensity difference are smaller than the errors in the analysis system. Differences in initial humidity, and to a lesser extent initial temperature differences, at the surface and at lower heights are the key factors leading to differences in the forecasted TC intensity. These differences in initial humidity and temperature relate to both the overall values and distribution of these parameters.