基于修正TV范数光流法的Himawari-8云导风算法研究
|
摘要
修正TV范数光流法是基于新一代静止轨道气象卫星Himwari-8探测的云图数据提出的反演云导风的方法.该方法首先通过ROF去噪模型提取云图细节部分,提出应用TV-L1范数局部惩罚项的光流法实现了高分辨率云导风反演,依靠质量标识对所求导风进行质量控制,去掉不合理风矢。实验结果表明,该方法适用于400 hPa高度以下的中低云云导风计算。
Based on the cloud image data of a new generation of stationary orbit meteorological satellite Himwari-8 AHI, the modified TV norm optical flow method is put forward to reverse the cloud derived wind.Firstly, through the ROF denoising model the image details are extracted. Using TV-L1 norm local penalty,the optical flow method is presented to achieve a high resolution cloud derived wind inversion. The quality of cloud derived wind is controlled based on quality identification. And unreasonable wind vector is removed.The experimental results show that this method is suitable for cloud derived wind of the low and medium cloud level under the height of 400 hPa.
Based on the cloud image data of a new generation of stationary orbit meteorological satellite Himwari-8 AHI, the modified TV norm optical flow method is put forward to reverse the cloud derived wind.Firstly, through the ROF denoising model the image details are extracted. Using TV-L1 norm local penalty,the optical flow method is presented to achieve a high resolution cloud derived wind inversion. The quality of cloud derived wind is controlled based on quality identification. And unreasonable wind vector is removed.The experimental results show that this method is suitable for cloud derived wind of the low and medium cloud level under the height of 400 hPa.
引文
[1] Peng Xiayun, Min Jinzhong, Zhou Zhenbo, et al.Quality control of wind field retrieved by single Doppler radar[J]. Journal of Nanjing Institute of Meteorology, 2007, 30(1):79-85(in Chinese).彭霞云,闵锦忠,周振波,等.单多普勒雷达反演风场的质量控制[J].南京气象学院学报,2007,30(1):79-85.
[2] Zhu Lin, Shou Shaowen, Peng Jiayi, et al. Numerical experiments on the assimilation of sounding data by Caiman wave[J]. Journal of Nanjing Institute of Meteorology, 2008, 31(2):264-271(in Chinese).朱琳,寿绍文,彭加毅,等.集合卡尔曼滤波同化探空资料的数值试验[J].南京气象学院学报,2008, 31(2):264-271.
[3] Zhang Shaobo, Lu Hancheng, Kang Jianwei. The lower level wind retrieval applied to a super cell storm using Doppler radar[J]. Scientia Meteorologica Sinica, 27(2):147-154(in Chinese).张少波,陆汉城,康建伟.一次超级单体风暴多普勒天气雷达资料的低层风场反演研究[J].气象科学,2007, 27(2):147-154.
[4] Xu Fen, Xia Wenmei, Hu Zhiqun, et al. Application of Doppler weather radar wind products in helicity computation[J]. Scientia Meteorologica Sinica, 2007, 27(5):495-501(in Chinese).徐芬,夏文梅,胡志群,等.多普勒天气雷达风场产品在螺旋度计算中的应用[J].气象科学,2007,27(5):495-501.
[5] Hou Qing, Xu Jianmin. The relationship between the upper troposphere and the main rain belts in summer revealed by the satellite guidance data[J]. Journal of Applied Meteorological Science,2006, 17(2):138-144(in Chinese).侯青,许健民.卫星导风资料所揭示的对流层上部环流形势与我国夏季主要雨带之间的关系[J].应用气象学报,2006, 17(2):138-144.
[6] Li Huahong, Wang Man, Xue Jishan, et al. FY-2C trace wind data in mesoscale numerical model[J]. Acta Meteorologica Sinica, 2008, 66(1):50-58(in Chinese).李华宏,王曼,薛纪善,等.FY-2C云迹风资料在中尺度数值模式中的应用研究[J].气象学报,2008, 66(1):50-58.
[7] Zhou Y S, Cao J, Gao S T. The method of decomposing wind field in a limited area and its application to typhoon SAOMEI[J]. Acta Physica Sinica, 2008, 57(10):6654-6665.
[8] Zhuang Zhaorong, Xue Jishan. Assimilation of coud-derived winds and its impact on typhoon forecast[J]. Journal of Tropical Meteorology,2004, 20(3):225-236(in Chinese).庄照荣,薛纪善.云迹风资料的三维变分同化及对台风预报的影响试验[J].热带气象学报,2004,20(3):225-236.
[9] Zhan Yizhe, Wang Zhenhui, Guan Li, et al. A summary of the application of cloud motion wind datato typhoon track prediction[J]. Advances in Earth Science,2011,26(4):386-393(in Chinese).詹奕哲,王振会,官莉,等.卫星导风在台风路径预报中的应用进展[J].地球科学进展,2011,26(4):386-393.
[10] Xu Jianmin, Zhang Qisang. Status review on atmospheric motion vectors-derivation and application[J]. Journal of Applied Meteorology, 2006,17(5):574-582(in Chinese).许健民,张其松.卫星风推导和应用综述[J].应用气象学报,2006, 17(5):574-582.
[11] Schmetz J, Holmlund K, Hoffman J, et al. Operational cloud-motion winds from meteosat infrared images[J]. Journal of Applied Meteorology, 1993,32(7):1206-1225.
[12] Leese J A, Novak C S, Clark B B. An automated technique for obtaining cloud motion from geosynchronous satellite data using cross correlar tion[J]. Journal of Applied Meteorology, 1971,10(1):118-132.
[13] Xu J, Holmlund K, Zhang Q, et al. Comparison of two schemes for derivation of atmospheric motion vectors[J]. Journal of Geophysical Research Atmospheres, 2002, 107(D14):ACL 4-1-ACL 4-15.
[14] Wang Changshuai. Research on Cloud Guided Wind Algorithm of Parallel Satellite[D].Changchun:Master's Thesis of Jilin University,2016(in Chinese).王昌帅.并行卫星云导风算法研究[D].长春:吉林大学硕士论文,2016.
[15] Purdom J F W. Detailed cloud motions from satellite imagery taken at thirty second, one and three minute intervals[C]. Proceeding to the 3rd international wind workshop in Ascona, Switzerland, 1996:10-12.
[16] Xu Jianming, Wang Zhenhui. The application condition of Fourier analysis technique and the analysis of phrase overlap[J]. Scientia Meteorologica Sinica, 2004, 24(3):309-313(in Chinese).许建明,王振会.傅立叶相位分析导风技术的适用范围和相位重叠分析[J].气象科学,2004, 24(3):309-313.
[17] Long Zhiyong, Shi Hanqing, Huang Sixun. A new idea of cloud motion wind derived from satellite images[J]. Acta Physica Sinica, 2011, 60(5):059202(in Chinese).龙智勇,石汉青,黄思训.利用卫星云图反演云导风的新思路[J].物理学报,2011, 60(5):059202.
[18] Ma Xialin, Luo Peng, Chen Zhibin, et al. A method for cloud motion wind vector prediction based on scale-invariant feature transform[J]. Meteorological Science and Technology, 2014, 42(3):391-396(in Chinese).马侠霖,罗鹏,陈志斌,等.基于尺度不变特征变换的云导风计算方法[J].气象科技,2014, 42(3):391-396.
[19] Xiao Wuping. Japanese first next-generation weather satellite Himawari-8 enters geostationary orbit[J]. Space International, 2015(2):87-91(in Chinese).肖武平.日本首颗新一代气象卫星向日葵-8入轨[J].国际太空,2015(2):87-91.
[20] Horn B K P, Schunck B G. Determining optical flow[J]. Artificial Intelligence, 1981,17(1-3):185-203.
[21] Wedel A, Pock T, Zach C, et al. An Improved Algorithm for TV-L1 Optical Flow[M]//Statistical and Geometrical Approaches to Visual Motion Analysis. Heidelberg:Springer, 2009:23-45.
[22] Shou Yixuan. Pattern Recognition Technology and Its Application in Meteorological Research[D]. Nanjing:Master's Thesis of Nanjing University of Information Science and Technology,2004(in Chinese).寿亦萱.模式识别技术及其在气象研究中的应用[D].南京:南京信息工程大学硕士论文,2004.
[23] Tan Jun. Research on Optical Flow Estimation Method Based on Variational Principle[D]. Beijing:Master's Thesis of Beijing University of Technology, 2015(in Chinese).谭君.基于变分原理的光流估计方法研究[D].北京:北京工业大学硕士论文,2015.
[24] Rudin L I, Osher S, Fatemi E. Nonlinear total variation based noise removal algorithms[J].Physica D:Nonlinear Phenomena, 1992, 60(1-4):259-268.
[25] Sun D, Roth S, Black M J. Secrets of optical flow estimation and their principles[J]. 2010IEEE Conference on Computer Vision and Pattern Recognition, 2010, 119(5):2432-2439.
[26] Dong W, Shi G, Hu X, et al. Nonlocal sparse and low-rank regularization for optical flow estimation[J]. IEEE Transactions on Image Processing, 2014,23(10):4527-4538.
[27] Liu Min.The Tracking and Prediction of Multiple Clouds in a Mesoscale Convective System[D].Nanjing:Master's Thesis of Nanjing University of Science and Technology, 2013(in Chinese).刘闵.中尺度对流系统中多个云团的跟踪和预测[D].南京:南京理工大学硕士论文,2013.
[28] Xue Chenbin, Gong Jiandong, He Caifu, et al.Quality control of cloud derived wind vectors from geostationary meteorological satellites with its application to data assimilation system[J]. Journal of Applied Meteorological Science, 2013,24(3):356-364(in Chinese).薛湛彬,龚建东,何财福,等.静止卫星云导风的质量控制及在同化中的应用[J].应用气象学报,2013, 24(3):356-364.
[29] Wei Qingyu. FY-2C. The Design and Development of a Half Time Cloud Profile Wind Service System for Meteorological Satellite[D]. Nanjing:Master's Thesis of Nanjing University of Information Science and Technology, 2006(in Chinese).魏清宇. FY-2C气象卫星半小时间隔云图导风业务系统的设计与开发[D].南京:南京信息工程大学硕士论文,2006.
[2] Zhu Lin, Shou Shaowen, Peng Jiayi, et al. Numerical experiments on the assimilation of sounding data by Caiman wave[J]. Journal of Nanjing Institute of Meteorology, 2008, 31(2):264-271(in Chinese).朱琳,寿绍文,彭加毅,等.集合卡尔曼滤波同化探空资料的数值试验[J].南京气象学院学报,2008, 31(2):264-271.
[3] Zhang Shaobo, Lu Hancheng, Kang Jianwei. The lower level wind retrieval applied to a super cell storm using Doppler radar[J]. Scientia Meteorologica Sinica, 27(2):147-154(in Chinese).张少波,陆汉城,康建伟.一次超级单体风暴多普勒天气雷达资料的低层风场反演研究[J].气象科学,2007, 27(2):147-154.
[4] Xu Fen, Xia Wenmei, Hu Zhiqun, et al. Application of Doppler weather radar wind products in helicity computation[J]. Scientia Meteorologica Sinica, 2007, 27(5):495-501(in Chinese).徐芬,夏文梅,胡志群,等.多普勒天气雷达风场产品在螺旋度计算中的应用[J].气象科学,2007,27(5):495-501.
[5] Hou Qing, Xu Jianmin. The relationship between the upper troposphere and the main rain belts in summer revealed by the satellite guidance data[J]. Journal of Applied Meteorological Science,2006, 17(2):138-144(in Chinese).侯青,许健民.卫星导风资料所揭示的对流层上部环流形势与我国夏季主要雨带之间的关系[J].应用气象学报,2006, 17(2):138-144.
[6] Li Huahong, Wang Man, Xue Jishan, et al. FY-2C trace wind data in mesoscale numerical model[J]. Acta Meteorologica Sinica, 2008, 66(1):50-58(in Chinese).李华宏,王曼,薛纪善,等.FY-2C云迹风资料在中尺度数值模式中的应用研究[J].气象学报,2008, 66(1):50-58.
[7] Zhou Y S, Cao J, Gao S T. The method of decomposing wind field in a limited area and its application to typhoon SAOMEI[J]. Acta Physica Sinica, 2008, 57(10):6654-6665.
[8] Zhuang Zhaorong, Xue Jishan. Assimilation of coud-derived winds and its impact on typhoon forecast[J]. Journal of Tropical Meteorology,2004, 20(3):225-236(in Chinese).庄照荣,薛纪善.云迹风资料的三维变分同化及对台风预报的影响试验[J].热带气象学报,2004,20(3):225-236.
[9] Zhan Yizhe, Wang Zhenhui, Guan Li, et al. A summary of the application of cloud motion wind datato typhoon track prediction[J]. Advances in Earth Science,2011,26(4):386-393(in Chinese).詹奕哲,王振会,官莉,等.卫星导风在台风路径预报中的应用进展[J].地球科学进展,2011,26(4):386-393.
[10] Xu Jianmin, Zhang Qisang. Status review on atmospheric motion vectors-derivation and application[J]. Journal of Applied Meteorology, 2006,17(5):574-582(in Chinese).许健民,张其松.卫星风推导和应用综述[J].应用气象学报,2006, 17(5):574-582.
[11] Schmetz J, Holmlund K, Hoffman J, et al. Operational cloud-motion winds from meteosat infrared images[J]. Journal of Applied Meteorology, 1993,32(7):1206-1225.
[12] Leese J A, Novak C S, Clark B B. An automated technique for obtaining cloud motion from geosynchronous satellite data using cross correlar tion[J]. Journal of Applied Meteorology, 1971,10(1):118-132.
[13] Xu J, Holmlund K, Zhang Q, et al. Comparison of two schemes for derivation of atmospheric motion vectors[J]. Journal of Geophysical Research Atmospheres, 2002, 107(D14):ACL 4-1-ACL 4-15.
[14] Wang Changshuai. Research on Cloud Guided Wind Algorithm of Parallel Satellite[D].Changchun:Master's Thesis of Jilin University,2016(in Chinese).王昌帅.并行卫星云导风算法研究[D].长春:吉林大学硕士论文,2016.
[15] Purdom J F W. Detailed cloud motions from satellite imagery taken at thirty second, one and three minute intervals[C]. Proceeding to the 3rd international wind workshop in Ascona, Switzerland, 1996:10-12.
[16] Xu Jianming, Wang Zhenhui. The application condition of Fourier analysis technique and the analysis of phrase overlap[J]. Scientia Meteorologica Sinica, 2004, 24(3):309-313(in Chinese).许建明,王振会.傅立叶相位分析导风技术的适用范围和相位重叠分析[J].气象科学,2004, 24(3):309-313.
[17] Long Zhiyong, Shi Hanqing, Huang Sixun. A new idea of cloud motion wind derived from satellite images[J]. Acta Physica Sinica, 2011, 60(5):059202(in Chinese).龙智勇,石汉青,黄思训.利用卫星云图反演云导风的新思路[J].物理学报,2011, 60(5):059202.
[18] Ma Xialin, Luo Peng, Chen Zhibin, et al. A method for cloud motion wind vector prediction based on scale-invariant feature transform[J]. Meteorological Science and Technology, 2014, 42(3):391-396(in Chinese).马侠霖,罗鹏,陈志斌,等.基于尺度不变特征变换的云导风计算方法[J].气象科技,2014, 42(3):391-396.
[19] Xiao Wuping. Japanese first next-generation weather satellite Himawari-8 enters geostationary orbit[J]. Space International, 2015(2):87-91(in Chinese).肖武平.日本首颗新一代气象卫星向日葵-8入轨[J].国际太空,2015(2):87-91.
[20] Horn B K P, Schunck B G. Determining optical flow[J]. Artificial Intelligence, 1981,17(1-3):185-203.
[21] Wedel A, Pock T, Zach C, et al. An Improved Algorithm for TV-L1 Optical Flow[M]//Statistical and Geometrical Approaches to Visual Motion Analysis. Heidelberg:Springer, 2009:23-45.
[22] Shou Yixuan. Pattern Recognition Technology and Its Application in Meteorological Research[D]. Nanjing:Master's Thesis of Nanjing University of Information Science and Technology,2004(in Chinese).寿亦萱.模式识别技术及其在气象研究中的应用[D].南京:南京信息工程大学硕士论文,2004.
[23] Tan Jun. Research on Optical Flow Estimation Method Based on Variational Principle[D]. Beijing:Master's Thesis of Beijing University of Technology, 2015(in Chinese).谭君.基于变分原理的光流估计方法研究[D].北京:北京工业大学硕士论文,2015.
[24] Rudin L I, Osher S, Fatemi E. Nonlinear total variation based noise removal algorithms[J].Physica D:Nonlinear Phenomena, 1992, 60(1-4):259-268.
[25] Sun D, Roth S, Black M J. Secrets of optical flow estimation and their principles[J]. 2010IEEE Conference on Computer Vision and Pattern Recognition, 2010, 119(5):2432-2439.
[26] Dong W, Shi G, Hu X, et al. Nonlocal sparse and low-rank regularization for optical flow estimation[J]. IEEE Transactions on Image Processing, 2014,23(10):4527-4538.
[27] Liu Min.The Tracking and Prediction of Multiple Clouds in a Mesoscale Convective System[D].Nanjing:Master's Thesis of Nanjing University of Science and Technology, 2013(in Chinese).刘闵.中尺度对流系统中多个云团的跟踪和预测[D].南京:南京理工大学硕士论文,2013.
[28] Xue Chenbin, Gong Jiandong, He Caifu, et al.Quality control of cloud derived wind vectors from geostationary meteorological satellites with its application to data assimilation system[J]. Journal of Applied Meteorological Science, 2013,24(3):356-364(in Chinese).薛湛彬,龚建东,何财福,等.静止卫星云导风的质量控制及在同化中的应用[J].应用气象学报,2013, 24(3):356-364.
[29] Wei Qingyu. FY-2C. The Design and Development of a Half Time Cloud Profile Wind Service System for Meteorological Satellite[D]. Nanjing:Master's Thesis of Nanjing University of Information Science and Technology, 2006(in Chinese).魏清宇. FY-2C气象卫星半小时间隔云图导风业务系统的设计与开发[D].南京:南京信息工程大学硕士论文,2006.