The role of moist processes in the intrinsic predictability of indian ocean cyclones

The role of moist processes in short-range forecasts of Indian Ocean tropical cyclones (TCs) track and intensity and upscale error cascade from cloud-scale processes affecting the intrinsic predictability of TCs was investigated using the Weather Research and Forecasting model with parameterized and explicitly resolved convection. Comparing the results from simulations of four Indian Ocean TCs at 10 km resolution with parameterized convection and convection-permitting simulations at 1.1 km resolution, both reproduced the observed TC tracks and intensities significantly better than simulations at 30 km resolution with parameterized convection. “Identical twin” experiments were performed by introducing random perturbations to the simulations for each TC. Results show that moist convection plays a major role in intrinsic error growth that ultimately limits the intrinsic predictability of TCs, consistent with past studies of extratropical cyclones. More specifically, model intrinsic errors start to build up from the regions of convection and ultimately affect the larger scales. It is also found that the error at small scale grows faster compared to the larger scales. The gradual increase in error energy in the large scale is a manifestation of upscale cascade of error energy from convective to large scale. Rapid upscale error growth from convective scales limits the intrinsic predictability of the TCs up to 66 h. The intrinsic predictability limit estimated by the 10 km resolution runs is comparable to that estimated by the convection-permitting simulations, suggesting some usefulness of high-resolution (~10 km) models with parameterized convection for TC forecasting and predictability study.