The predictability of tropical cyclone intensity: Results from a simple dynamical model

It is possible that the time-evolution of tropical cyclone (TC) intensity is highly sensitive to differences in the initial vortex state and environment that are consistent with the typical magnitudes of analysis error. If this is the case, then the expected accuracy of deterministic TC intensity predictions would consequently be limited. To determine the degree to which deterministic TC intensity prediction accuracy is constrained by predictability, it is first necessary to quantify the sensitivity of TC intensity to the initial state of the TC vortex and its environment. Here, such a sensitivity study is carried out in the context of the Coupled Hurricane Intensity Prediction System (CHIPS), a simple dynamical model that simulates a TC vortex and its interaction with the underlying ocean. While CHIPS TC simulations are not as realistic as those of contemporary mesoscale models, CHIPS does simulate the basic potential intensity-limited intensification dynamics and the effects of environmental vertical wind shear and TC-induced upper oceanic mixing. This is done by CHIPS at a computational cost that is vastly lower than a mesoscale model, enabling a systematic exploration of the sensitivity of CHIPS-simulated intensity to select initial state and environmental variables.

The sensitivity of CHIPS-simulated intensity is evaluated with respect to four primary variables: (1) Initial intensity, (2) Initial intensification rate, (3) Magnitude of the 850-200 hPa environmental vertical wind shear, and (4) Oceanic mixed layer depth. In an overall sense, CHIPS-simulated intensity is considerably more sensitive to shear magnitude than oceanic mixed layer depth, and considerably more sensitive to initial intensity than initial intensification rate. Furthermore, the sensitivity of CHIPS-simulated intensity to shear magnitude and initial intensity is highly state dependent. This state dependence of the sensitivity is such that CHIPS-simulated intensity is most sensitive to the initial intensities and shear magnitudes most likely to be observed for Atlantic basin TCs. Intriguingly, this result suggests that the most common TC intensity forecast situations are those in which the expected deterministic forecast accuracy is lowest.