Extended Range Machine-Learning Severe Weather Guidance Based on the Operational GEFS

Research by Hill et al. (2023) demonstrated extremely promising results using a random forest machine-learning algorithm with input from the Global Ensemble Forecast System v12 (GEFSv12) to generate probabilistic severe weather forecasts out to days 4-8. In their work, the GEFS Reforecast Dataset (GEFS/R) was used to train and test their random forest model, which was used to generate forecasts using operational GEFS forecasts as input. One limitation of the Hill et al. (2023) work was that, due to computational limitations, the GEFS/R forecasts only include 5 members. This work aims to build on Hill et al. (2023) by using the operational GEFSv12 dataset for training and testing. At the time of this writing, GEFSv12 has been operational for more than two years, so it may be possible to take advantage of the 31 members in the operational GEFSv12 for training and testing to get an improved result. We test this hypothesis by conducted several different experiments where the number of ensemble members in the training and testing dataset is varied between 5 and 31. We also test using individual ensemble members vs. ensemble summary measures as predictors.