The AnEn is developed here using HWRF reforecast data and corresponding Best Track intensity data from 2011-2015. There are 95 Atlantic TCs and 124 Eastern Pacific TCs in this dataset. Over 60 predictors (based on environmental and storm inner-core characteristics) were derived from all of the 1339 Atlantic and 1729 Eastern Pacific Ocean HWRF runs available (with 0-126 h lead-times, at 3 h increments). In this application, the predictors are averaged over the three time intervals (0-24 hours, 0-48 hours, 0-72 hours) used to compute DeltaVmax. The analogs are searched in a training dataset made by HWRF reforecast data built by randomly selecting half of the TCs available in the 2011-2015 period. This dataset is also used to select a subset of the best predictors and define their weights optimizing the AnEn's performance in term of Continuous Ranked Probability Score (CRPS). The predictor weights are kept unchanged during the testing dataset (made by the remaining TCs) used to evaluate the AnEn performance.
The AnEn TC DeltaVmax deterministic forecasts have been obtained as the mean of the 20 ensemble members at each time increment interval. In this presentation, the improvements compared to the raw TC DeltaVmax forecast from HWRF will be shown in terms of Mean Absolute Error (MAE) and other commonly used metrics. An in-depth analysis of critical attributes of probabilistic predictions of TC DeltaVmax generated with the AnEn will also be shown. These attributes include statistical consistency, reliability, resolution, sharpness, and the spread-skill relationship. In particular, we will assess the AnEn skills of forecasting the probability of exceeding three distinct thresholds of DeltaVmax: 30 kt, 55 kt and 65 kt respectively for the 0-24 hours, 0-48 hours and 0-72 hours time intervals. The improvements to TC rapid intensification forecasting produced by this computationally inexpensive technique with respect to HWRF are statistical significant between 20% to 50%, and contribute towards NOAA's longstanding goal to improve intensity change forecasting.