Early lead: A WRF ensemble demonstrating a data mining capability

EarlyLEAD should not be viewed as a surrogate for the meteorology and IT research efforts that are central to the LEAD mission. Instead, EarlyLEAD focuses on the testing of a simple workflow that uses a data mining tool to select regions of interest in a prescribed way to establish a WRF domain at three institutions, concluding with the creation of an ensemble product that can be cataloged, accessed, and visualized. By updating the WRF model with new NAM or WRF output, the WRF domain can be steered to follow an interesting weather phenomenon. However, in contrast to LEAD, which will demonstrate the potential benefits of dynamic adaptability and a cyberinfrastructure that supports complex forecast systems that change configuration on demand in response to the weather, EarlyLEAD is static during the simulation. In this regard, EarlyLEAD resembles an autonomous floater, where the PEA takes the place of the observer in locating regions of interest. Nonetheless, EarlyLEAD has already proved useful as an effort where LEAD components such as PEA can be demonstrated, evaluated and refined.