Parallel Analog Ensemble -- The Power of Weather Analogs

CG Auditorium
Weiming Hu

The Analog Ensemble (AnEn) is a technique used to generate ensemble forecasts from a single simulation of a deterministic numerical weather prediction (NWP) model. In contrast to relying on multi-model or multi-realization approaches, AnEn generates ensemble forecasts by searching through a historical repository for the most similar past weather forecasts and selecting ensemble members from observations that are associated with the most similar historical forecasts.

AnEn generates accurate and calibrated weather forecasts. It also has several advantages over the other ensemble methods: (1) it can be easily coupled with high-resolution forecast models; (2) single deterministic simulation is sufficient for ensemble generation which saves on computational resources; (3) it can generate additional predictand variables that are, in some cases, not produced by the underlying weather model (e.g. wind and solar photovoltaic energy production). AnEn has been applied to various fields including weather forecasts, renewable energy assessment, extreme events early warning, and air quality control.

This technical note introduces a fast and extensible implementation of the AnEn algorithm, Parallel Analog Ensemble, aiming to close the gaps between ongoing research and the computational algorithm. The package is implemented in C++ for performance and shipped with an R API for broader usability. It also includes visualization and file I/O functionality to serve the needs of multiple research projects. This technical note focuses on discussing the implementation details of the package which serves as a developers' documentation. It contains information on class designs, interface designs, and how to extend the package and contribute. Usage information and tutorials are also included and available online from the official website.

Speaker Description: 

Weiming Hu is a Ph.D. candidate at Penn State University in the Dept. of Geography focusing on computational algorithms and geographic information science.

Event Category: