TCdiag—A Visualization Tool for Tropical Cyclone Intensity Forecasting

While tropical cyclone (TC) track prediction could be considered one the greatest successes in the history of Numerical Weather Prediction (NWP), TC intensity forecasts have been much slower to improve. The most skillful intensity forecast aids currently are the dynamical-statistical schemes (e.g., SHIPS – Statistical Hurricane Intensity Prediction System) that relate TC intensity change to large-scale environment parameters such as 850-200 hPa wind shear and SST anomalies. In fact, dynamical-statistical models (DSMs) often outperform the maximum surface wind speed (intensity) taken from the dynamical model itself. Thus, forecasters almost always consult both the DSM forecasts and the dynamical NWP model input to the DSM (the so-called ‘diagnostic file’) during the forecast process.

TCdiag (https://ruc.noaa.gov/hfip/tcdiag) is a web site that displays the diagnostic file for all TCs globally and is derived from both local ESRL-runs of Finite-Volume version3 (FV3) model as well as the big operational global models of NCEP, UKMO, ECMWF. The site also provides graphics of the field used in the diagnostic and how the diagnostic is calculated. A good example is vertical wind shear (850-200 hPa) – perhaps the most important variable for TC intensity forecasting. While the mean shear is related to intensity change, the convection driving the model TC is the local wind shear and should be interrogated to understand how this important forcing mechanism works in the dynamical model and the DSM.

Although TCdiag provides both graphics and numbers, the interface is plain and inefficient for real-time operations. So, in collaboration with the Joint Typhoon Warning Center (JTWC), TCdiag was simplified to both quickly present plots of the critical parameters and to optimize model intercomparison. A prototype was first developed at ESRL and then adjusted/modified during real-time JTWC typhoon forecast operations. The site also benefited from JTWC expertise in developing web interfaces for their operations thus making the new TCdiag a win-win for both research and operations.