Sensitivity analysis of drag coefficient and length scale of wind influence on tropical cyclone intensity change using net energy gain rate

Predicting tropical cyclones (TC) rapid intensification (RI) is one of the most significant challenges. This study refines the Net Energy Gain Rate (NGR) metric to improve TC intensity predictions, focusing on uncertainties in the drag coefficient ( C d ​) at extreme wind speeds and the effective length scale of TC-induced momentum transfer to the ocean ( R w ). Using data from the western North Pacific basin (2004–2021), we conducted sensitivity analyses with four C d parameterizations (increasing, decreasing, constant, and control) and varied R w ​ from 0.5 to 4 times the radius of maximum wind ( R max ​). Results indicate that R w ​=1 R max ​ consistently yields the highest correlation coefficient between NGR and intensity change in 24-hour among all combinations, especially for strong TCs (Category 3 or higher). Among the C d parameterizations, the scenario where C d decreases at wind speeds exceeding 50 m s -1 showed superior performance in capturing intensity changes. Multi-linear regression models incorporating NGR, prior 12-hour intensity changes, and vertical wind shear confirmed that decreasing C d at R w =1 R max ​ provides the most reliable predictions, achieving the highest prediction performance in the TC intensity change in 24-hour. These findings underscore the importance of accurately representing C d ​ behavior under extreme wind conditions and precisely defining R w ​ to enhance the predictive skill of NGR-based TC intensity forecasts.