TY - JOUR

T1 - Dynamically Optimized Unstructured Grid (DOUG) for Analog Ensemble of numerical weather predictions using evolutionary algorithms

AU - Hu, Weiming

AU - Cervone, G.

PY - 2019/12

Y1 - 2019/12

N2 - The Analog Ensemble is a statistical technique that generates probabilistic forecasts using a current deterministic prediction, a set of historical predictions, and the associated observations. It generates ensemble forecasts by first identifying the most similar past predictions to the current one, and then summarizing the corresponding observations. This is a computationally efficient solution for ensemble modeling because it does not require multiple numerical weather prediction simulations, but a single model realization. Despite this intrinsic computational efficiency, the required computation can grow very large because atmospheric models are routinely run with increasing resolutions. For example, the North American Mesoscale forecast system contains over 262 792 grid points to generate a 12 km prediction. The North American Mesoscale model generally uses a structured grid to represent the domain, despite the fact that certain physical changes occur non-uniformly across space and time. For example, temperature changes tend to occur more rapidly in mountains than plains. An evolutionary algorithm is proposed to dynamically and automatically learn the optimal unstructured grid pattern. This iterative evolutionary algorithm is guided by Darwinian evolutionary rule generation and instantiation to identify grid vertices. Analog computations are performed only at vertices. Therefore, minimizing the number of vertices and identifying their locations are paramount to optimizing the available computational resources, minimizing queue time, and ultimately achieving better results. The optimal unstructured grid is then reused to guide the predictions for a variety of applications like temperature and wind speed.

AB - The Analog Ensemble is a statistical technique that generates probabilistic forecasts using a current deterministic prediction, a set of historical predictions, and the associated observations. It generates ensemble forecasts by first identifying the most similar past predictions to the current one, and then summarizing the corresponding observations. This is a computationally efficient solution for ensemble modeling because it does not require multiple numerical weather prediction simulations, but a single model realization. Despite this intrinsic computational efficiency, the required computation can grow very large because atmospheric models are routinely run with increasing resolutions. For example, the North American Mesoscale forecast system contains over 262 792 grid points to generate a 12 km prediction. The North American Mesoscale model generally uses a structured grid to represent the domain, despite the fact that certain physical changes occur non-uniformly across space and time. For example, temperature changes tend to occur more rapidly in mountains than plains. An evolutionary algorithm is proposed to dynamically and automatically learn the optimal unstructured grid pattern. This iterative evolutionary algorithm is guided by Darwinian evolutionary rule generation and instantiation to identify grid vertices. Analog computations are performed only at vertices. Therefore, minimizing the number of vertices and identifying their locations are paramount to optimizing the available computational resources, minimizing queue time, and ultimately achieving better results. The optimal unstructured grid is then reused to guide the predictions for a variety of applications like temperature and wind speed.

UR - http://www.scopus.com/inward/record.url?scp=85070653603&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85070653603&partnerID=8YFLogxK

U2 - 10.1016/j.cageo.2019.07.003

DO - 10.1016/j.cageo.2019.07.003

M3 - Article

AN - SCOPUS:85070653603

VL - 133

JO - Computers and Geosciences

JF - Computers and Geosciences

SN - 0098-3004

M1 - 104299

ER -