El Niño-Southern Oscillation(ENSO)can be currently predicted reasonably well six months and longer,but large biases and uncertainties remain in its real-time prediction.Various approaches have been taken to impro...El Niño-Southern Oscillation(ENSO)can be currently predicted reasonably well six months and longer,but large biases and uncertainties remain in its real-time prediction.Various approaches have been taken to improve understanding of ENSO processes,and different models for ENSO predictions have been developed,including linear statistical models based on principal oscillation pattern(POP)analyses,convolutional neural networks(CNNs),and so on.Here,we develop a novel hybrid model,named as POP-Net,by combining the POP analysis procedure with CNN-long short-term memory(LSTM)algorithm to predict the Niño-3.4 sea surface temperature(SST)index.ENSO predictions are compared with each other from the corresponding three models:POP model,CNN-LSTM model,and POP-Net,respectively.The POP-based pre-processing acts to enhance ENSO-related signals of interest while filtering unrelated noise.Consequently,an improved prediction is achieved in the POP-Net relative to others.The POP-Net shows a high-correlation skill for 17-month lead time prediction(correlation coefficients exceeding 0.5)during the 1994-2020 validation period.The POP-Net also alleviates the spring predictability barrier(SPB).It is concluded that value-added artificial neural networks for improved ENSO predictions are possible by including the process-oriented analyses to enhance signal representations.展开更多
Traditional principal component analysis （PCA） is a second-order method and lacks the ability to provide higherorder representations for data variables. Recently, a statistics pattern analysis （SPA） framework has ...Traditional principal component analysis （PCA） is a second-order method and lacks the ability to provide higherorder representations for data variables. Recently, a statistics pattern analysis （SPA） framework has been incorporated into PCA model to make full use of various statistics of data variables effectively. However, these methods omit the local information, which is also important for process monitoring and fault diagnosis. In this paper, a local and global statistics pattern analysis （LGSPA） method, which integrates SPA framework and locality pre- serving projections within the PCK is proposed to utilize various statistics and preserve both local and global in- formation in the observed data. For the purpose of fault detection, two monitoring indices are constructed based on the LGSPA model. In order to identify fault variables, an improved reconstruction based contribution （IRBC） plot based on LGSPA model is proposed to locate fault variables. The RBC of various statistics of original process variables to the monitoring indices is calculated with the proposed RBC method. Based on the calculated RBC of process variables＇ statistics, a new contribution of process variables is built to locate fault variables. The simula- tion results on a simple six-variable system and a continuous stirred tank reactor system demonstrate that the proposed fault diagnosis method can effectively detect fault and distinguish the fault variables from normal variables.展开更多
Based on the observational data in summer,the variations of intraseasonal oscillation(ISO)of the daily rainfall over the lower reaches of the Yangtze River valley(LYRV)were studied by using the non-integer spectrum an...Based on the observational data in summer,the variations of intraseasonal oscillation(ISO)of the daily rainfall over the lower reaches of the Yangtze River valley(LYRV)were studied by using the non-integer spectrum analysis.The NCEP/NCAR reanalysis data for the period of 1979―2005 were analyzed by principal oscillation pattern analysis(POP)to investigate the spatial and temporal characteristics of principal ISO patterns of the global circulation.The relationships of these ISO patterns to the rainfall ISO and the heavy precipitation process over LYRV were also discussed.It is found that the rainfall over LYRV in May―August is mainly of periodic oscillations of 10―20,20―30 and 60―70 days,and the interannual variation of the intensity of its 20―30-day oscillation has a strongly positive correlation with the number of the heavy precipitation process.Two modes(POP1,POP2)are revealed by POP for the 20―30-day oscillation of the global 850 hPa geopotential height.One is a circumglobal telecon-nection wave train in the middle latitude of the Southern Hemisphere(SCGT)with an eastward propagation,and the other is the southward propagation pattern in the tropical western Pacific(TWP).The POP modes explain 7.72%and 7.66%of the variance,respectively.These two principal ISO patterns are closely linked to the low frequency rainfall and heavy precipitation process over LYRV,in which the probability for the heavy precipitation process over LYRV is 54.9%and 60.4%for the positive phase of the imaginary part of POP1 and real part of POP2,respectively.Furthermore,the models of the global atmospheric circulation for the 20―30-day oscillation in association with or without the heavy pre-cipitation process over LYRV during the Northern Hemisphere summer are set up by means of the composite analysis method.Most of the heavy precipitation processes over LYRV appear in Phase 4 of SCGT or Phase 6 of TWP.When the positive phases of 20―30-day oscillations for the rainfall over LYRV are associated with(without)the heavy precipitation process,a strong westerly stream appears(disappears)from the Arabian Sea via India and Bay of Bengal(BOB)to southern China and LYRV for the global 850 hPa filtered wind field during Phase 4 of SCGT.This situation is favorable(unfavorable)for the forming of the heavy precipitation process over LYRV.Similarly,a strong(weak)western wind belt forms from India through BOB to southern China and LYRV and the subtropical northwestern Pacific and central and eastern equatorial Pacific during Phase 6 of TWP for the cases with(without)the heavy precipitation process.The evolutions of these ISO patterns related to the 20―30-day oscillation are excited by either the interaction of extratropical circulation in both hemispheres or the heat source forcing in Asia monsoon domain and internal interaction of circulation in East Asia.These two global circulation models might therefore provide valuable information for the extended-range forecast of the heavy precipitation process over LYRV during the 10―30 days.展开更多
基金supported by the Strategic Priority Research Program of the Chinese Academy of Sciences(Grant No.XDA19060102)the National Natural Science Foundation of China[NSFCGrant Nos.41690122(41690120),and 42030410].
文摘El Niño-Southern Oscillation(ENSO)can be currently predicted reasonably well six months and longer,but large biases and uncertainties remain in its real-time prediction.Various approaches have been taken to improve understanding of ENSO processes,and different models for ENSO predictions have been developed,including linear statistical models based on principal oscillation pattern(POP)analyses,convolutional neural networks(CNNs),and so on.Here,we develop a novel hybrid model,named as POP-Net,by combining the POP analysis procedure with CNN-long short-term memory(LSTM)algorithm to predict the Niño-3.4 sea surface temperature(SST)index.ENSO predictions are compared with each other from the corresponding three models:POP model,CNN-LSTM model,and POP-Net,respectively.The POP-based pre-processing acts to enhance ENSO-related signals of interest while filtering unrelated noise.Consequently,an improved prediction is achieved in the POP-Net relative to others.The POP-Net shows a high-correlation skill for 17-month lead time prediction(correlation coefficients exceeding 0.5)during the 1994-2020 validation period.The POP-Net also alleviates the spring predictability barrier(SPB).It is concluded that value-added artificial neural networks for improved ENSO predictions are possible by including the process-oriented analyses to enhance signal representations.
基金Supported by the National Natural Science Foundation of China(61273160,61403418)the Natural Science Foundation of Shandong Province(ZR2014FL016)the Fundamental Research Funds for the Central Universities(14CX06132A)
文摘Traditional principal component analysis （PCA） is a second-order method and lacks the ability to provide higherorder representations for data variables. Recently, a statistics pattern analysis （SPA） framework has been incorporated into PCA model to make full use of various statistics of data variables effectively. However, these methods omit the local information, which is also important for process monitoring and fault diagnosis. In this paper, a local and global statistics pattern analysis （LGSPA） method, which integrates SPA framework and locality pre- serving projections within the PCK is proposed to utilize various statistics and preserve both local and global in- formation in the observed data. For the purpose of fault detection, two monitoring indices are constructed based on the LGSPA model. In order to identify fault variables, an improved reconstruction based contribution （IRBC） plot based on LGSPA model is proposed to locate fault variables. The RBC of various statistics of original process variables to the monitoring indices is calculated with the proposed RBC method. Based on the calculated RBC of process variables＇ statistics, a new contribution of process variables is built to locate fault variables. The simula- tion results on a simple six-variable system and a continuous stirred tank reactor system demonstrate that the proposed fault diagnosis method can effectively detect fault and distinguish the fault variables from normal variables.
基金Supported by the Program for the Fundamental Research of China Meteorological Administration(Grant No.200726)
文摘Based on the observational data in summer,the variations of intraseasonal oscillation(ISO)of the daily rainfall over the lower reaches of the Yangtze River valley(LYRV)were studied by using the non-integer spectrum analysis.The NCEP/NCAR reanalysis data for the period of 1979―2005 were analyzed by principal oscillation pattern analysis(POP)to investigate the spatial and temporal characteristics of principal ISO patterns of the global circulation.The relationships of these ISO patterns to the rainfall ISO and the heavy precipitation process over LYRV were also discussed.It is found that the rainfall over LYRV in May―August is mainly of periodic oscillations of 10―20,20―30 and 60―70 days,and the interannual variation of the intensity of its 20―30-day oscillation has a strongly positive correlation with the number of the heavy precipitation process.Two modes(POP1,POP2)are revealed by POP for the 20―30-day oscillation of the global 850 hPa geopotential height.One is a circumglobal telecon-nection wave train in the middle latitude of the Southern Hemisphere(SCGT)with an eastward propagation,and the other is the southward propagation pattern in the tropical western Pacific(TWP).The POP modes explain 7.72%and 7.66%of the variance,respectively.These two principal ISO patterns are closely linked to the low frequency rainfall and heavy precipitation process over LYRV,in which the probability for the heavy precipitation process over LYRV is 54.9%and 60.4%for the positive phase of the imaginary part of POP1 and real part of POP2,respectively.Furthermore,the models of the global atmospheric circulation for the 20―30-day oscillation in association with or without the heavy pre-cipitation process over LYRV during the Northern Hemisphere summer are set up by means of the composite analysis method.Most of the heavy precipitation processes over LYRV appear in Phase 4 of SCGT or Phase 6 of TWP.When the positive phases of 20―30-day oscillations for the rainfall over LYRV are associated with(without)the heavy precipitation process,a strong westerly stream appears(disappears)from the Arabian Sea via India and Bay of Bengal(BOB)to southern China and LYRV for the global 850 hPa filtered wind field during Phase 4 of SCGT.This situation is favorable(unfavorable)for the forming of the heavy precipitation process over LYRV.Similarly,a strong(weak)western wind belt forms from India through BOB to southern China and LYRV and the subtropical northwestern Pacific and central and eastern equatorial Pacific during Phase 6 of TWP for the cases with(without)the heavy precipitation process.The evolutions of these ISO patterns related to the 20―30-day oscillation are excited by either the interaction of extratropical circulation in both hemispheres or the heat source forcing in Asia monsoon domain and internal interaction of circulation in East Asia.These two global circulation models might therefore provide valuable information for the extended-range forecast of the heavy precipitation process over LYRV during the 10―30 days.
