气候变化领域集成服务门户(CCPortal): Improving Wind Forecasts in the Lower Stratosphere by Distilling an Analog Ensemble Into a Deep Neural Network

CCPortal

DOI	10.1029/2020GL089098
	Improving Wind Forecasts in the Lower Stratosphere by Distilling an Analog Ensemble Into a Deep Neural Network
	Candido S.; Singh A.; Delle Monache L.
发表日期	2020
ISSN	0094-8276
卷号	47 期号:15
英文摘要	We discuss improving forecasts of winds in the lower stratosphere using machine learning to postprocess the output of the European Centre for Medium-Range Weather Forecasts (ECMWF) Integrated Forecast System. We postprocess global three-dimensional predictions and demonstrate distilling the analog ensemble (AnEn) method into a deep neural network, which reduces postprocessing latency to near zero maintaining increased forecast skill. This approach reduces the error with respect to ECMWF high-resolution deterministic prediction between 2–15% for wind speed and 15–25% for direction and is on par with ECMWF ensemble (ENS) forecast skill to hour 60. Verifying with Loon data from stratospheric balloons, AnEn has 20% lower error than ENS for wind speed and 15% for wind direction, despite significantly lower real-time computational cost to ENS. Similar performance patterns are reported for probabilistic predictions, with larger improvements of AnEn with respect to ENS. We also demonstrate that AnEn generates a calibrated probabilistic forecast. ©2020. The Authors.
英文关键词	Deep neural networks; Meteorological balloons; Neural networks; Wind; Computational costs; European centre for medium-range weather forecasts; Lower stratosphere; Performance patterns; Probabilistic forecasts; Probabilistic prediction; Stratospheric balloon; Three-dimensional predictions; Weather forecasting; artificial neural network; ensemble forecasting; machine learning; stratosphere; wind direction; wind velocity
语种	英语
来源期刊	Geophysical Research Letters
文献类型	期刊论文
条目标识符	http://gcip.llas.ac.cn/handle/2XKMVOVA/170009
作者单位	Loon, Mountain View, CA, United States; Center for Western Weather and Water Extremes, Scripps Institution of Oceanography, University of California, San Diego, La Jolla, CA, United States
推荐引用方式 GB/T 7714	Candido S.,Singh A.,Delle Monache L.. Improving Wind Forecasts in the Lower Stratosphere by Distilling an Analog Ensemble Into a Deep Neural Network[J],2020,47(15).
APA	Candido S.,Singh A.,&Delle Monache L..(2020).Improving Wind Forecasts in the Lower Stratosphere by Distilling an Analog Ensemble Into a Deep Neural Network.Geophysical Research Letters,47(15).
MLA	Candido S.,et al."Improving Wind Forecasts in the Lower Stratosphere by Distilling an Analog Ensemble Into a Deep Neural Network".Geophysical Research Letters 47.15(2020).