674 Evaluation of Flash Drought Identification with Machine Learning Techniques, Part 1: Standard Machine Learning Algorithms

Wednesday, 31 January 2024
Hall E (The Baltimore Convention Center)
Stuart Galen Edris, University of Oklahoma, Norman, OK; and J. B. Basara, J. I. Christian, J. C. Furtado, A. McGovern, and X. Xiao

Handout (10.6 MB)

Drought is a climatological dry extreme that can have a number of impacts on water resources and agriculture. To mitigate impacts, many studies have focused on identifying and predicting drought events. In recent decades, machine learning (ML) has emerged as a useful tool for drought identification and prediction, and ML applications using long-term metrics can successfully identify seasonal scale droughts and have shown some skill at drought prediction. However, recent investigations have demonstrated that long-term drought metrics, such as the Palmer drought severity index and standardized precipitation index, cannot adequately identify the rapidly developing conditions 1 month) that characterize flash drought. Thus, studies focused on flash drought have yielded a number of indices to represent rapidly evolving conditions using precipitation, soil moisture, evapotranspiration (ET), and potential ET (PET). Further, while much work has been conducted to investigate and identify flash drought events, limited work has been completed using ML techniques. Thus, this study aims to use several ML techniques to identify flash drought such as random forests, support vector machines, and ada boosted decision trees. The ML techniques were trained on a set of flash drought variables (soil moisture, ET, PET, precipitation, and temperature) with an additional analysis to determine how much each variable contributes toward rapid intensification and flash drought identification.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner