气候变化领域集成服务门户(CCPortal): Multi-model seasonal forecasts for the wind energy sector

CCPortal

DOI	10.1007/s00382-019-04654-y
	Multi-model seasonal forecasts for the wind energy sector
	Lee D.Y.; Doblas-Reyes F.J.; Torralba V.; Gonzalez-Reviriego N.
发表日期	2019
ISSN	0930-7575
起始页码	2715
结束页码	2729
卷号	53 期号:2020-05-06
英文摘要	An assessment of the forecast quality of 10 m wind speed by deterministic and probabilistic verification measures has been carried out using the original raw and two statistical bias-adjusted forecasts in global coupled seasonal climate prediction systems (ECMWF-S4, METFR-S3, METFR-S4 and METFR-S5) for boreal winter (December–February) season over a 22-year period 1991–2012. We follow the standard leave-one-out cross-validation method throughout the work while evaluating the hindcast skills. To minimize the systematic error and obtain more reliable and accurate predictions, the simple bias correction (SBC) which adjusts the systematic errors of model and calibration (Cal), known as the variance inflation technique, methods as the statistical post-processing techniques have been applied. We have also built a multi-model ensemble (MME) forecast assigning equal weights to datasets of each prediction system to further enhance the predictability of the seasonal forecasts. Two MME have been created, the MME4 with all the four prediction systems and MME2 with two better performing systems. Generally, the ECMWF-S4 shows better performance than other individual prediction systems and the MME predictions indicate consistently higher temporal correlation coefficient (TCC) and fair ranked probability skill score (FRPSS) than the individual models. The spatial distribution of significant skill in MME2 prediction is almost similar to that in MME4 prediction. In the aspect of reliability, it is found that the Cal method has more effective improvement than the SBC method. The MME4_Cal predictions are placed in close proximity to the perfect reliability line for both above and below normal categorical events over globe, as compared to the MME2_Cal predictions, due to the increase in ensemble size. To further compare the forecast performance for seasonal variation of wind speed, we have evaluated the skill of the only raw MME2 predictions for all seasons. As a result, we also find that winter season shows better performance than other seasons. © 2019, Springer-Verlag GmbH Germany, part of Springer Nature.
英文关键词	10 m wind speed; Forecast verification; Multi-model ensemble; Seasonal prediction systems; Statistical post-processing
语种	英语
scopus关键词	accuracy assessment; climate prediction; error analysis; hindcasting; seasonal variation; weather forecasting; wind power; wind velocity
来源期刊	Climate Dynamics
文献类型	期刊论文
条目标识符	http://gcip.llas.ac.cn/handle/2XKMVOVA/146069
作者单位	Earth Sciences Department, Barcelona Supercomputing Center (BSC), C/Jordi Girona, 29, Barcelona, 08034, Spain; Computational Physics and Methods (CCS-2), Computer, Computational, and Statistical Sciences (CCS) Division, Los Alamos National Laboratory (LANL), Los Alamos, United States; Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spain
推荐引用方式 GB/T 7714	Lee D.Y.,Doblas-Reyes F.J.,Torralba V.,et al. Multi-model seasonal forecasts for the wind energy sector[J],2019,53(2020-05-06).
APA	Lee D.Y.,Doblas-Reyes F.J.,Torralba V.,&Gonzalez-Reviriego N..(2019).Multi-model seasonal forecasts for the wind energy sector.Climate Dynamics,53(2020-05-06).
MLA	Lee D.Y.,et al."Multi-model seasonal forecasts for the wind energy sector".Climate Dynamics 53.2020-05-06(2019).