深度学习网络在降水相态判识和预报中的应用     

黄骄文  蔡荣辉  姚蓉  王胜春  滕志伟 《气象》2021,(3):317-326

利用1996—2015年中国的高空探测资料和地面观测数据,挑选发生降水的数十万个样本将其分为降雨和降雪两类事件,抽象为二分类问题,采用深度学习网络技术构建降水相态判识模型,并用2016—2017年的数据进行测试检验,针对2018年1月下旬中国一次大范围雨雪天气过程进行个例检验,在此基础上探讨了深度学习网络在降水相态判识和预报中的应用。主要结论如下:基于深度学习网络判识模型的判识准确率为98.2%,雨、雪的TS评分分别为97.4%和94.4%,相应空报率为1.7%和2.0%,漏报率为1.0%和3.7%,较传统指标阈值法的判识准确率有较大提高;个例检验显示,基于实况探空数据的模型判识结果与降水相态实况在全国基本保持一致,欧洲中期数值预报中心(ECMWF)的降水相态预报产品和模型的预报结果对全国的降水相态都表现出较好的预报能力,而对雨雪分界线的预报,模型的预报结果较ECMWF总体上更接近实况。测试结果表明,模型较好地提取了雨、雪降水相态的结构特征,深度学习网络在降水相态判识和预报中的应用具有可行性和一定的优势,可为降水相态的客观判识和预报提供重要技术支撑。  相似文献   

基于卷积神经网络的京津冀地区高分辨率格点预报偏差订正试验          下载免费PDF全文

张延彪  宋林烨  陈明轩  韩雷  杨璐 《大气科学学报》2022,45(6):850-862

为了进一步提高RISE系统高分辨率网格化预报产品的准确率,同时考虑到深度学习近年来在地学领域的有效应用,采用2019—2021年高分辨率RISE系统数据,设计出卷积神经网络模型Rise-Unet,实现了未来4～12 h地面2 m温度、2 m相对湿度、10 m-U风速以及10 m-V风速预报结果的订正。订正试验结果表明,采用均方根误差和平均绝对误差作为评分标准,与RISE原始预报结果相比,基于Rise-Unet模型可以有效提高温湿风预报结果的准确率。该基于深度学习的Rise-Unet偏差订正技术可应用于RISE系统的后处理模块,对提升RISE系统百米级分辨率或其他高分辨率模式系统格点预报水平具有重要的科学意义和应用价值。  相似文献   

A Deep Learning Method for Bias Correction of ECMWF 24–240 h Forecasts   总被引：1，自引：0，他引：1  

Lei HAN  Mingxuan CHEN  Kangkai CHEN  Haonan CHEN  Yanbiao ZHANG  Bing LU  Linye SONG  Rui QIN 《大气科学进展》2021,38(9):1444-1459

Correcting the forecast bias of numerical weather prediction models is important for severe weather warnings. The refined grid forecast requires direct correction on gridded forecast products, as opposed to correcting forecast data only at individual weather stations. In this study, a deep learning method called CU-net is proposed to correct the gridded forecasts of four weather variables from the European Centre for Medium-Range Weather Forecast Integrated Forecasting System global model(ECMWF-IFS): 2-m temperature, 2-m relative humidity, 10-m wind speed, and 10-m wind direction, with a forecast lead time of 24 h to 240 h in North China. First, the forecast correction problem is transformed into an image-toimage translation problem in deep learning under the CU-net architecture, which is based on convolutional neural networks.Second, the ECMWF-IFS forecasts and ECMWF reanalysis data(ERA5) from 2005 to 2018 are used as training,validation, and testing datasets. The predictors and labels(ground truth) of the model are created using the ECMWF-IFS and ERA5, respectively. Finally, the correction performance of CU-net is compared with a conventional method, anomaly numerical correction with observations(ANO). Results show that forecasts from CU-net have lower root mean square error, bias, mean absolute error, and higher correlation coefficient than those from ANO for all forecast lead times from 24 h to 240 h. CU-net improves upon the ECMWF-IFS forecast for all four weather variables in terms of the above evaluation metrics, whereas ANO improves upon ECMWF-IFS performance only for 2-m temperature and relative humidity. For the correction of the 10-m wind direction forecast, which is often difficult to achieve, CU-net also improves the correction performance.  相似文献   

Quantitative Precipitation Forecast Experiment Based on Basic NWP Variables Using Deep Learning     

Kanghui ZHOU  Jisong SUN  Yongguang ZHENG  Yutao ZHANG 《大气科学进展》2022,39(9):1472-1486

The quantitative precipitation forecast (QPF) performance by numerical weather prediction (NWP) methods depends fundamentally on the adopted physical parameterization schemes (PS). However, due to the complexity of the physical mechanisms of precipitation processes, the uncertainties of PSs result in a lower QPF performance than their prediction of the basic meteorological variables such as air temperature, wind, geopotential height, and humidity. This study proposes a deep learning model named QPFNet, which uses basic meteorological variables in the ERA5 dataset by fitting a non-linear mapping relationship between the basic variables and precipitation. Basic variables forecasted by the highest-resolution model (HRES) of the European Centre for Medium-Range Weather Forecasts (ECMWF) were fed into QPFNet to forecast precipitation. Evaluation results show that QPFNet achieved better QPF performance than ECMWF HRES itself. The threat score for 3-h accumulated precipitation with depths of 0.1, 3, 10, and 20 mm increased by 19.7%, 15.2%, 43.2%, and 87.1%, respectively, indicating the proposed performance QPFNet improved with increasing levels of precipitation. The sensitivities of these meteorological variables for QPF in different pressure layers were analyzed based on the output of the QPFNet, and its performance limitations are also discussed. Using DL to extract features from basic meteorological variables can provide an important reference for QPF, and avoid some uncertainties of PSs.  相似文献   

支持向量机及其在地震预报中的应用前景   总被引：2，自引：0，他引：2       下载免费PDF全文

王炜  林命週  马钦忠  赵利飞 《地震工程学报》2006,28(1):78-84

统计学习理论(SLT)是研究小样本情况下机器学习规律的理论。支持向量机(SVM)基于统计学习理论,可以处理高度非线性分类和回归等问题,不但较好地解决了小样本、过学习、高维数、局部最小等实际难题,而且具有很强的泛化(预测)能力。本文介绍了支持向量机的分类、回归方法,分析了这一方法的特点,讨论了该方法在地震预报中的应用前景。  相似文献   

Youth transitions and going to university: the perceptions of students attending a geography summer school access programme     

Peter E Hopkins 《Area》2006,38(3):240-247

This paper contributes to understandings of youth transitions in childhood and youth geographies through the use of a participatory diagramming exercise with students studying geography in a summer school access programme at a Scottish university. In particular, the paper explores young people's perceptions about adult/child binaries and their hopes and fears in applying to university. The discussion highlights the extended nature of youth transitions: the usefulness of participatory diagramming as a research method and teaching technique for this group; and the consequences that the increasing interest in childhood and youth geographies might have on teaching and learning in human geography.  相似文献   

Learning Conservation: the role of conservation covenants in landscape redesign at Project Hindmarsh,Victoria     

Chris Harrington  Ruth Lane 《The Australian geographer》2006,37(2):187-209

Biodiversity decline continues apace across the Australian landscape with a pressing need to redesign land use to address this situation. The significance of private land increasingly is recognised for the protection and enhancement of biodiversity as landholders inevitably make decisions that affect environmental quality. Biodiversity conservation is as much a social process as a physical one. Conservation covenants are perpetual agreements under which landholders choose to conserve land voluntarily, primarily for conservation purposes. The role covenants might play in landscape-scale conservation was investigated in north-western Victoria. In-depth interviews with a range of participants were undertaken, with an emphasis on the role covenantors might play as social learning and cultural change agents. Analysis of these interviews offers useful perspectives for understanding socio-cultural dimensions of landscape change and exploring the differing values of production farmers and nature conservation landholders. Consideration is then given to approaches to engaging local production farmers in nature covenants and promoting communication between this group and the largely non-production conservationists who currently form the mainstay of conservation covenants.  相似文献   

泌阳凹陷南部陡坡带地震资料的炮域波动方程叠前深度偏移     

李锋  顾汉明 《地质科技情报》2007,26(3):104-108

泌阳凹陷南部陡坡带是河南油田的重点勘探老区,其构造产状不仅陡,断层也多,叠后时间偏移难以得到较好的地震成像效果,基于Kirchhoff积分偏移方法具有运算速度快、效率高的特点,以及炮域波动方程方法能保持地震波的动力学特征,应用Kirchhoff积分偏移方法进行了偏移速度分析,建立了精细的速度模型.然后基于该模型在深度域进行了剩余速度分析以及层析成像,进一步提高了速度模型的精度. 通过分析偏移孔径、去假频参数以及延拓步长等成像处理参数对成像效果的影响,确定出最佳偏移处理参数,最后利用炮域波动方程方法对工区的三维地震数据进行了叠前偏移成像.成像结果表明,该方法能够使该区陡坡带反射层位得到较好的偏移,信噪比和横向分辨率都得到提高.  相似文献   

支持向量数据描述在西北暴雨预报中的应用试验   总被引：1，自引：0，他引：1       下载免费PDF全文

燕东渭  孙田文杨艳方建刚  刘志镜 《应用气象学报》2007,18(5):676-681

传统机器学习中通常隐含假设所研究问题是类别平衡的, 气象预报中预测灾害天气时就不满足这个假设, 这时往往需要预测重要而稀少的正类 (少数类)。传统机器学习以精度最大化为目标, 在遇到不平衡类别问题时, 容易训练出把所有实例都分为反类 (多数类) 的平庸的分类器。支持向量数据描述是从支持向量机 (SVM) 发展而来的基于核的机器学习方法, 只使用一类样本就可以工作, 适合于不平衡类别。以铜川暴雨预测作为试验对象, 对SVM和支持向量数据描述 (SVDD) 进行了对比试验。试验结果表明对于这个不平衡类别问题SVDD具有优势。  相似文献   

Inter-organizational coordination in extreme events: The World Trade Center attacks, September 11, 2001   总被引：1，自引：2，他引：1  

Louise K. Comfort  Naim Kapucu 《Natural Hazards》2006,39(2):309-327

This paper addresses the problem of inter-organizational coordination in response to extreme events. Extreme events require coordinated action among multiple actors across many jurisdictions under conditions of urgent stress, heavy demand and tight time constraints. The problem is socio-technical in that the capacity for inter-organizational coordination depends upon the technical structure and performance of the information systems that support decision making among the participating organizations. Interactions among human managers, computers and organizations under suddenly altered conditions of operation are complex and not well understood. Yet, coordinating response operations to extreme events is an extraordinarily complex task for public and nonprofit managers. This paper will analyze the interactions among public, private and nonprofit organizations that evolved in response to the 11 September 2001 attacks, examining the relationships among organizations in terms of timely access to information and types of supporting infrastructure. The performance of the inter-organizational system is examined in the context of the events of 11 September 2001 from the theoretical perspective of complex adaptive systems. A model of auto-adaptation is proposed for implementation to improve inter- organizational performance in extreme events. This model is based on the concept of individual, organizational and collective learning in environments exposed to recurring risk, guided by a shared goal. Such a model requires public investment in the development of an information infrastructure that can support the intense demand for communication, information search, exchange and feedback that characterizes an auto-adaptive system.  相似文献