福建师范大学学报（自然科学版）2024，Vol.40Issue(1)：87-95,9.DOI:10.12046/j.issn.1000-5277.2024.01.010
一种气温降尺度的半循环对抗生成网络
Half-Cycle Generative Adversarial Network for Temperature Downscaling
摘要
Abstract
This paper proposes a half-cycle generative adversarial downscaling model and ap-plies it to downscale ERA5 reanalysis global climate surface temperature data in China and its sur-rounding regions.This model introduces both a reconstruction generator network and a degradation generator network and optimizes the downscaling results using adversarial loss and half-cycle loss.The effectiveness of this method was validated through ablation experiments.The results show that compared with traditional interpolation methods and other deep learning models,this model exhibits improvements in objective evaluation metrics,resulting in more detailed surface temperature data generated.关键词
对抗生成网络/降尺度/深度学习Key words
generative adversarial networks/downscaling/deep learning分类
信息技术与安全科学引用本文复制引用
黎瑞泉,翁彬,陈家祯,黄添强,游立军..一种气温降尺度的半循环对抗生成网络[J].福建师范大学学报（自然科学版）,2024,40(1):87-95,9.基金项目
国家重点研发计划专项(2018YFC1505805) （2018YFC1505805）
福建省引导性项目(2021Y0057,2022Y0008) （2021Y0057,2022Y0008）
福建师大教改项目(I202201105) （I202201105）
