PROGRESS OF LARGE-SCALE AIR-SEA INTERACTION STUDIES IN CHINA
PU Shuzhen, ZHAO Jinping, YU Weidong,
ZHAO Yongping * and YAONG Bo
First Institute of Oceanography, State Oceanic Administration, Qingdao 266061, China
*Institute of Oceanology, Chinese Academy of Sciences, Qingdao 266071, China
It is summarized in this paper that progress of large-scale air-sea interaction studies has been achieved in China in the four year period from July 1998 to July 2002, including the seven aspects in the area of the air-sea interaction i. e. the field experiments and oceanic cruise surveys, air-sea interaction related to the tropical Pacific, monsoon-related air-sea interaction, air-sea interaction in the north Pacific Ocean, air-sea interaction in the Indian Ocean, and the air-sea interactions in global oceans. However more attention has been paid to the second and the third aspects because a large number of papers in the reference literature for preparing and organizing this paper are concentrated in the tropical Pacific Ocean, such as ENSO process with its climatic effects and dynamics, and the monsoon-related air-sea interaction. The literature also involves various phenomena with their difference time and spatial scales from each other such as intraseasonal, annual, interannual, interdecadal variabilities in the atmosphere/ ocean interaction system, reflecting the contemporary themes in the four year period at the beginning of an era from the post TOGA to CLIVAR. Apparently it is a difficult task to summarize a great progress of this area, as it is extracted from a large quantity of the literature into such a short article, although authors tried hard. Anyway the progress is much greater and better than this article is.Key words: review of air-sea interaction, various time and spatial scales, atmosphere/ocean variability, climatic abnormality
The air-sea interaction study is an important scientific area in China. The research activities of Chinese scientists are vigorous and productive in this area. A large number of the scientists have participated in the international programs related to large scale air-sea interaction studies such as TOGO, COARE, WOCE, CLIVAR, and the others for many years, and organized domestic projects funded by National Natural Science Foundation of China, Chinese Academy of Sciences, China Meteorological Administration, State Oceanic Administration of China, and other agencies. This report is intended to briefly introduce the progress in the large-scale air-sea interaction studies made in China according to the literature published by Chinese scientists in the 4-year period from July 1998 to July 2002.
I. THE FIELD EXPERIMENTS AND CRUISE SURREYS
The South China Sea Monsoon Experiment (SCSMEX) is an international cooperative project in which China plays a major role to study the air-sea interaction processes related to the South China Sea monsoon. Three research vessels, “Xiangyanghong 14”, “Science 1”, and “Experiment 3” have been involved in the field experiment of the project on the ocean for 2 months in 1998. Measurements of physical oceanography, meteorology, and marine chemistry were conducted, including CTD casts, ADCP profilings, sea surface meteorological observations, high altitude soundings and marine chemistry water samplings. The multi-ship and multi-discipline observations lay the foundation of the basic researches for understanding the processes of the monsoon transition and its northward progression in the South China Sea in late spring and early summer, and the oceanic response to the monsoon atmospheric circulation. The research vessels for the field experiment of SCSMEX are supported by Chinese Academy of Sciences and State Oceanic Administration of China.
In addition to SCSMEX field observations a program, “Pacific-Indian Warm Pool. Its Current System and Air-sea Interaction”, is funded by Chinese Ministry of Sci. & Tech. and National Natural Science Foundation. A pilot study of this program started in 2001-2002, 3 Argo-floats have been deployed in the warm pool region on the board of R/V Snow Dragon when the ship navigates toward the Antarctica, and some XBT drops deployed in the tropics as well. The program will help further understand the ENSO dynamics and the variability in the tropical waters.
The Southern Ocean near Prydz Bay is another sea area where Chinese scientists frequently conduct air-sea field experiments on the board of Chinese R/V Snow-Dragon. A program, Physical Process Study of Air-sea-ice Interaction in the Antarctic Ocean (1996-2000), supported by State Oceanic Administration of China, and 5 summers cruises of the southern ocean expeditions were completed in the 5 years, obtaining full depth CTD data and water samples for chemistry.
II. THE AIR-SEA INTERACTION STUDIES RELATED TO THE TROPICAL PACIFIC OCEAN
The tropical Pacific Ocean is the tropical sea area nearest to China, where the oceanic abnormality is closely associated with the climatic variability in East Asia including China The Ocean has drawn much attention of Chinese oceanographers for a long time. The study activities can be divided into the following aspects.
1. Intraseasonal Variability in the Tropical Pacific Ocean
TOPEX/POSEIDON altimetric data have been used to analyze for the spatial distribution of the intraseasonal oscillation in the tropical Pacific Ocean (Liu and Wang 1999, Hu and Liu 2002), Liu and Wang (1999) also used the model simulation output from POM with eddy-resolution for this purpose. As their results show, the quasi-30 day oscillation can be found in the zonal belts about 5°N and 5°S to the east of 160°W by means of the data analysis. The quasi-90 day oscillation is evidenced in the zonal belts about 20°N and 20°S, and it is more obvious in the western Pacific Ocean of the 20°N belt. The signal for the quasi-60 day oscillation distributed in the 10°N zonal belt and the 20°S belt is weaker than the quasi-90 day signal although it is distinct in the data. The signal in the 10°N belt is more obvious than in the 10°S belt. In addition, quasi-60 day oscillation can be also found in the eastern equatorial Pacific Ocean (5°N-5°S, 170°-120°W). Hu and Liu (2002) report that the 90 day oscillation, with its annual variability, is closely related to ENSO occurrence. In addition to altimetry data, satellite OLR data (1979-1993) are also used to find the source for the intraseasonal oscillation (ISO) in the eastern tropical Indian Ocean and the western tropical Pacific Ocean. The water area where the intraseasonal oscillation with a period of 6.5—12.5 pentads is the most vigorous is the eastern tropical Indian Ocean, ISO with their respective periods can also be obviously evidenced in the tropical ocean to the northwest of Australia, in the tropical Pacific Ocean to the northeast of Australia, in the south part and the north part of the South China Sea, in the Luzon Strait, and the area to the south of Japan (Wang, Liu, and Xu 2000).
ISO and semiannual oscillation on the 200 hPa in the tropical atmosphere, and their dependence upon SSTA of the eastern equatorial Pacific Ocean are analyzed by means of the estimation for the oscillation kinetic energy and the calculation for the correlation between the energy and the SSTA. It is revealed that the positive SSTA corresponds to weaker atmospheric oscillations and the negative SSTA to stronger oscillations (Li, C. and G. Li 1999).
2. Interannual Variabilities of Wind, Sea Temperature etc. and ENSO Process Study
Of wind, sea temperature etc. and ENSO interannual variability of the tropical Pacific Ocean continue to be one of the most interesting subjects in the air-sea studies. Long time series of SST, precipitation, wind, oceanic heat, geopotential height, and other metrorological or oceanographic observational elements are widely used to determine their interannual variabilities and the ENSO-evolution process. Yin and Ni (2001) use the NCEP/NCAR monthly mean SST and 1000 hPa wind data (1979-1998) to study the interannual variability in the tropical Pacific Ocean, Indian Ocean and Atlantic Ocean, to describe the characteristics related to ENSO events, and to estimate the corre lation among the three oceans. It is reported that the contemporary correlation between the regional SSTA representative for the Indian Ocean and the regional SSTA representative for the eastern equatorial Pacific Ocean is positive and weakly negative between the regional SSTA representative for the eastern equatorial Atlantic and the regional SSTA representative for the eastern equatorial Pacific. The positive correlation reaches the greatest when the regional SSTA of the equatorial Indian Ocean 3 months lags behind the regional SSTA for the eastern equatorial Pacific Ocean and the negative correlation reaches the maximum when the regional SSTA for the equatorial Atlantic 6 months leads the regional SSTA for the equatorial Pacific Ocean.
It is reported (Li et al. 1999) that monthly mean SSTA (from COADS) and 500 hPa height (H500) field (from National Meteorological Centre of China) are analyzed by means of CEOF method to study the low-frequency variabilities of SSTA in the global tropical waters and H500 field in the Northern Hemisphere (Li et al. 1999). The both quasi-biennial and 3-7 year oscillations with phase-lock characteristics can be found in the SSTA and H500 field. The H500 variability about 2 months lags behind the SSTA. The component for the 3-7 year oscillation of H500 has a stationary wave pattern like the PNA pattern, while the component for the quasi-biennial oscillation of H500 has a progressive wave feature.
Since the heat content in the tropical Pacific Ocean is an important causative element related to heat exchange between the upper ocean and the above atmosphere, its variability attracts great attention of Chinese oceanographers. Scientific evidence shows that the heat content in the upper layer above the thermocline of the equatorial Pacific Ocean is accumulated before an El Nino event, and its variability a few months leads the SSTA of the eastern equatorial Pacific Ocean (Pu et al. 1999). The location for the maximum lag correlation between the heat anomaly in a water column of the off equator area and that of the eastern equatorial area tends to move anti-clockwise in the northern tropical Pacific Ocean, and clockwise in the southern tropical Pacific Ocean, completing circling in about 4 years. It is also reported (in the submitted paper by Yu and Qiao) a similar cycling related to ENSO can be found for the heat content in the upper layer above 400 m depth of the tropical Pacific Ocean.
Li, Yu and Zhang (2001) study the interannual variability of MJO using the daily mean oceanic rainfall data (1979-1991) of Microware Sounding Unite and the pentad SST data from Climate Prediction Center Merged Analysis of Global Precipitation (1982-1992). It is found that MJO is limited to the west of the dateline in a normal year while it can extend further eastward and reach the central-eastern tropical Pacific Ocean in an ENSO year. The most significant correlation between the interannual variability of MJO and SST in the central-eastern Pacific Ocean is located in the El Nino 3 region. Numerical simulation result confirms the interannual variability of MJO in the tropical Pacific rainfall by use of CCM3 model forced by the observational SST. It is also reported (Wu et al. 2002) that a global AGCM (L9R15 version) forced by COADS SST is integrated for the period from 1945 to 1993. It is found that there are the interannual and interdecadal variabilities of the simulated surface wind over the tropical Pacific Ocean. The model output has been compared with the observations, showing a good agreement between the two.
ENSO or ENSO-related studies remain to be a focus of the air-sea interaction research activities in China, resulting in improvement of the knowledge about ENSO processes and dynamics. Particularly papers are concentrated in the oceanic subsurface temperature variability and its close relation with ENSO occurrence. Pu et al. (1998) discusses the evolution of the 1997/1998 ENSO events using the observational data. It is concluded in the case study that the variability of the sea temperature anomaly in the tropical oceanic thermocline and the Kelvin wave propagation play an important and significant role in ENSO evolution, and the subsurface warming, which follows the westerly burst, leads the surface warming in the tropical ocean. Li (2002), Li and Mu (1999) point out that subsurface oceanic temperature anomaly in the equatorial Pacific Ocean is closely related to the occurrence of ENSO. It is the positive (negative) subsurface anomaly that results in El Nino (La Nina) occurrence. They attributes the interannual variability of the subsurface temperature in the tropical Pacific Ocean to the East Asia monsoon abnormality. Chao Q. and J. Chao (2001) analyse the 1955-1998 SST in the western tropical Pacific Ocean, concluding that the positive subsurface temperature anomalies have appeared at about 120 m depth in the western tropical Pacific Ocean before SST becomes warmer in an ENSO event. Chen et al. (2002), using the XBT data of the upper tropical Pacific Ocean, reveal that the sea temperature anomaly of the north equatorial current (10°N) is closely associated with the subsurface temperature anomaly of the western Pacific warm pool. In the early stages of an ENSO cycle, the warmer water located in the thermocline of the certral and eastern Pacific Ocean can be transported westward to the western Pacific warm pool region by the north equatorial current, accumulating in the region, and extending toward the equator. Therefore the warmer thermocline water transport by the north equatorial current is one of the mechanisms for ENSO occurrence.
Sea surface wind anomalies, which play an important role in the ENSO occurrence and evolution, also attract Chinese scientists attention. Zhang and Huang (1998), Pu et al. (1999) report that the westerly wind anormalies in the western tropical Pacific Ocean initially appear in the western and central Pacific Ocean during an ENSO event, and displace eastward. The SST in the eastern equatorial Pacific Ocean increases after the wind anormaly has become stronger and gradually propagated eastward. The SST reaches its highest when the wind anormaly becomes located in the central Pacific Ocean. Then SST becomes decreased in the eastern and central Pacific Ocean, and the easterly anormaly appears in the western tropical Pacific Ocean. Finally normal conditions are recovered. Zhang and Huang (1998); Huang et al. (1998); Yan, Huang and Zhang (2001); Huang, Zhang and Yan (2001) using the model simulation describe the dynamic effects of the wind anormaly on an ENSO cycle, and discuss the Kelvin wave reflection and the Rossby wave generation, and their effects on the ENSO duration in the tropics. Yan and Zhang (2002a) hindcast the 1997/1998 ENSO event using an intermediate coupled model and discuss the wind anormaly importance for the ENSO cycle. When the model is forced by the observational wind anormaly for the period (1971-2000). The model is able to simulate all the ENSO events with a correlation about 0.63 between the observational El Nino 3 SSTA and the model output. Yan and Zhang (2002b) also analyse the seasonal variability of the westerly anormaly during a composite ENSO event in the western tropical Pacific Ocean and find the similarity of the seasonal variability between the westerly anormaly and the monthly mean divergence. The characteristics of the seasonal variability can be successfully simulated with a simple equatorial ocean-atmosphere coupled model. Their model results show that the model heating terms dependant upon the basic climatic state is necessary in simulating the seasonal variability of the westerly anormaly, playing an important role for forming the westerly anormaly in the western equatorial Pacific Ocean at the beginning of an ENSO event. Then in the developing or mature phase of an ENSO cycle. SSTA in the eastern equatorial Pacific Ocean is more important for forming the westerly anormaly In addition to the zonal wind anormaly the meridional wind anormaly in the tropical ocean is also analysed. Zhang, Chao, and Tan (2001) using the singular value decomposition method analyze the meridional wind anormaly in the tropical Pacific Ocean and the El Nino 3 index, and indicate that the variability of the meridional wind anormaly about 6 months leads the SST variability. Zhang and Chao (2001) also utilize an analytic model forced by equatorial convergence of ideal meridional wind field, and make the diagnostic analysis of the model results for the effects of Rossby waves excited by the meridional wind anormaly on the SST variability. Zhang, Ding, Zhao and Huang (2002) analyze the SST and the meridional wind of the tropical Pacific Ocean to obtain their periodic variability by means of Singular Spectrum Analysis (SSTA) method. The analysis results show that the variabilities respectively with the quasi-4 year period, the quasi-2 year period and the interdecadal time scale are significant. The variability of the quasi-4 year period has the greatest variance contribution. They also compare the SSTA results from the regional meridional winds respectively corresponding to the western tropical Pacific Ocean, the northeast tropical Pacific Ocean and the southeast tropical Pacific Ocean to find the spatial variability of SSTA and the importance of the regional wind characteristics for the SSTA distribution.
Zhao and Chen (1998), using COADS data and SOI of National Climate Center of China, divide either the ENSO events or the La Nina events occurred since 1956 into two types, i. e. the eastern type and the central area type. They find that La Nina with unusually cold SSTA in the central and eastern tropical Pacific Ocean has occurred one year before the ENSO onset of the eastern area type, and ENSO with unusually warm SSTA in the central and eastern tropical Pacific Ocean has occurred one year before La Nina onset of the eastern area type. In the contrast with the eastern area type, the central and western tropical Pacific SSTA becomes essential for an event of the central area type, with positive SSTA in the central and western Pacific before ENSO of the central area type, and negative SSTA there before La Nina of the central area type.
Zhu et al. (1998) use that the objective analysis data from National Meteorological Center to describe the respective characteristics of the 1976/1977 ENSO event and the 1982/1982 ENSO event and to find the difference in the background of the interdecadal variability, in which the ENSO events respectively occurred. The 1976/1977 ENSO occurred in a background where SST field was relatively cold, the westerly anormaly can be found on the level of 850 hPa over the equatorial Pacific Ocean (0°-80°W), and the SSTA oscillation with a 4 year period was being propagated eastward. The 1982/1982 ENSO occurred in a background where SST field was relatively warm, the easterly anormaly can be found on the level of 850 hPa over the equatorial Pacific Ocean (0°-80°W), and the SSTA oscillation was being propagated westward.
In addition to these observational studies and the numerical model simulations for ENSO, Wang et al. (1991) presented an analytic model to explain the ENSO-like oscillation between the South China Sea and the western tropical Pacific Ocean. The analytic solution of the model coincides with the data analysis. Wang et al. (2000) discuss the parameters in the model and propose a method for determining the parameters suitable for the inteannual time scale oscillation.
3. Climatic Effects of Both ENSO and Oceanic Variability in the Tropical Pacific Ocean
A large number of papers are concentrated in the climatic effects of ENSO and oceanic variability in the tropical Pacific Ocean. It is convenient for reviewing the papers to classify them into 3 categories: oceanic background of the 1998 heavy flood in the Changjiang River Basin, ENSO effects on the Climate, and others.(1) Oceanic background of the 1998 heavy flood in the Changjiang River Basin
The 1998 flood is the heaviest flood of the river basin in the 20 century but the 1954 flood. It is found that when the flood occurred the 1997/1998 ENSO event was coming to an end. The background of the air-sea interaction, in which the flood was caused, attracts Chinese scientists attention. Huang et al. (1998) analyse the observational data in order to find the climatic and hydrological features and the cause of the catastrophic flood occurring in the 1998 summer. They find that the flood occurred during the transition of the 1997/1998 ENSO event from its mature phase to its decaying phase, when SST in the tropical Pacific Ocean, especially the subsurface temperature in the western tropical Pacific Ocean became cooler, the convective activities around Philippines weakened, thus the western subtropical high was situated further southward. As a result from the abnormal climatic situation, abundant moisture which was carried by Asia summer monsoon from the Bay of Bengal and the South China Sea converged with moisture which was carried from the western tropical Pacific Ocean, initially converging and flowing into the middle and lower reaches of the river and then into the upper and middle reaches, and caused the continuously severe rainfall in the river basin and catastrophic flood. Guo, Zhao and Wang (2002) using IAP/LASG GOALS model forced by global observational SST reproduce the heavy rainfall over the Yangtze River Valley in the 1998 summer, and the subtropical high abnormality over the western Pacific Ocean. A series of numerical experiments are conducted with the observed SST in the selected oceanic regions and the climatic SST in the rest in order to test the regional effects on the rainfall. In addition to the regional effects, two more numerical experiments are also designed with the observed SST in the selected time periods and the climatic SST in the rest in order to test the temporal effects on the rainfall. The results show that SSTA in the Indian Ocean plays a major role than the other oceanic area for the rainfall during the flood, the SSTA in the Indian Ocean and the western Pacific Ocean is more closely related to the subtropical high abnormalities over the western Pacific Ocean than SSTA in the other area. It is also found that SSTA in the 1998 summer plays a more important role for the rainfall causing the flood rather than SSTA in the 1997 winter and the 1998 spring.
(2) ENSO effects on climate
Huang R., Zhang and Q. Zhang (2000) analyse the observed SSTA and the subsurface temperature data, the NCEP/NCAR reanalysis data, and the data of the daily precipitation in China to describe their characteristics during the 1997/1998 ENSO cycle and to discuss the impact on the East Asia climate in the 1998 summer. Their analysis shows that the vapor transport by the Asia summer monsoon was very weak in both the North China and the southern of the Korea Peninsula in the 1997 summer when the strongest ENSO event in the 20 century was rapidly developing, causing the drought and the hot weather in the area, and a large amount of water vapor is continuously transported from the Bay of Bengal, the South China Sea, and the western tropical Pacific Ocean into the Yangtze River Valley of China, South Koriea, and Japan in the 1997/1998 winter when the strongest ENSO event in the 20th century was rapidly decaying. The situation was associated to the southward displacement of the western Pacific subtropical high, resulting in the severe floods of those regions. Zhang et al. (1999) study the seasonal precipitation occurring in China when an ENSO process reaches its mature phase respectively in spring, summer, autumn, or winter. They find that positive precipitation anomalies in the composite precipitation patterns are all located in the South of China no matter an ENSO process reaches its mature phase in spring, autumn, or winter, and the anticyclonic anomaly of the lower troposphere is all situated to the north of the maritime continent. It is the anticyclonic anomaly that intensifies the western Pacific subtropical high, makes it further westward extend, and results in the positive precipitation anomaly in the South China. The composite precipitation pattern in China, when an ENSO process reaches its mature phase in summer, is different from those for the other three seasons. The negative precipitation anomalies in the pattern are located respectively in the further southern part and the northern part of China, while the positive precipitation anomalies tend to be located in between i. e. in the lower reaches of the Yangtze River and the Huaihe River Valley. Although the western Pacific subtropical high becomes intensified it covers the southeastern periphery of China where the precipitation decreases. Meanwhile the Indian summer monsoon can not reach the northern part of China providing less moisture inflow.
Chen, Le, Jia and Peng (2002) analyze the frequency of cyclone occurrence in the Southern Hemisphere (70°-170°E, 0°-80°S) and its correlations respectively with the SST in the eastern equatorial Pacific Ocean and with the southern oscillation index. They find that the correlations depend on the latitudinal zones where the cyclones occur and the correlation coefficient between SST and cyclone frequency reaches the maximum when SST 24 months lags behind the frequency of the cyclone occurrence.
(3) Variabilities in the western Pacific warm pool and its climatic effects
Li, Mu and Zhou (1999) study the seasonal and interannual variabilities of oceanic temperature in the western Pacific warm pool and their impacts on climate including the influence to East Asia monsoon, the teleconnection with the atmospheric circulation, and the effect on ENSO occurrence. They find that the subsurface temperature in the western Pacific warm pool is closely related to the SST in the eastern equatorial Pacific Ocean rather than the local SST, the positive (negative) SSTA in the warm pool is related to the stronger (weaker) summer monsoon with the EPA wave-train pattern stretching further (less) northward to higher (lower) latitudes, and the occurrence of ENSO is closely related to the subsurface temperature anomaly in the warm pool and the quick developing of ENSO to the eastward propagation of the subsurface temperature anomaly.
Zhao, Chen and Bai (2000) study the long term variabilities in the western tropical Pacific warm pool, and find that its temperature decreases in the first decade of the 20th century and (about 0.5℃) increases in the 8th decade of the 20th century. When the warm pool temperature increases, the subtropical high becomes further westward extended in the northwest Pacific Ocean and its ridge is further southward displaced, resulting in a change of the precipitation pattern in China i. e. less rain in the north and more rain in the south. They also find that El Nino events become more frequent and stronger in the warm period and La Nina events become more frequent and stronger in the cold period.
Wong, Zhang and Yan (1998) study not only the variability of the subtropical high, but also the variability of the area of the western tropical Pacific warm pool. The time series of the both are analyzed showing the same spectral peak at a period of 46.7 months. The subtropical high becomes stronger (weaker) when the warm pool area is larger (less). Zhang et al. (1999) identify the 2 climatic precipitation regions of North China by use of the comparative analysis method, and find that one of the two regions is closely related to the heat content of the western tropical Pacific warm pool with the best negative correlation between the wet season precipitation in this region and the warm pool heat content in October of the previous year. The wet season precipitation of the another region is closely related to the heat transport of Kuroshio in winter of the same year.
Liu and Wang (2000) analyse the seasonal variation of SST around the Indonesia Archipelago. They find that the maximum longitudinal SST difference between the north and the south is located in 110°E in boreal winter while it is in 130—145°E and 110°E in boreal summer. The seasonal variation is advantageous to the formation and maintainace of the cross equator air-flow toward the north at 0°, 105°E.
(4) SST variability in the tropical oceans with the variability of the subtropical high over the northwest Pacific Ocean and their climatic effects.
Ying and Sun (2000) analyze the observational data to study the response of the subtropical high over the northwest Pacific Ocean to the SST of the tropical oceans. They find that the strength of the subtropical high is closely related to SSTA of the central and eastern equatorial Pacific Ocean, the Indian Ocean and the western Atlantic Ocean. The subtropical high becomes stronger for the SST pattern of El Nino type, and weaker for the SST pattern of La Nina type. Li Y. and C. Li (1999) analyze the air temperature variability of Sichuan Province and SST variability in the western tropical Pacific Ocean. They find that both of them have the same tendency within the 50 years toward their decreases from the warmest decade of 1950s to the coldest decade of 1980s, with the relatively abrupt change in 1970s. When SST in the western tropical Pacific Ocean becomes increased (decreased), the subtropical high over the western Pacific Ocean is extended further westward (eastward) and the air temperature of Sichuan Province becomes warmer (cooler). Zhao and Wu (2002) use the wavelet analysis and EOF methods to analyze the SSTA variability of the tropical Pacific Ocean. They also analyse the variabilities of the subtropical high index and precipitation of Changjiang River Basin and the North China. They conclude that their variabilities with a 60 year periodicity play an important role to the abundance of the rainfall in the river basin and droughts of the North China in the recent years.
III. MONSOON-RELATED AIR-SEA INTERACTION
Both South China Sea monsoon and East Asia monsoon are important climatic systems with annual cycle, which influence the precipitation and the agricultural economics in China. As the monsoon cycle is affected by the thermal condition of the upper oceans adjacent to China, the monsoon-related air-sea interaction is one of the most important subjects in oceanography and meteorology in China.
1. East Asia Monsoon
Peng, Sun and Ni (2000) propose a new definition for East Asia monsoon index which can better reflect the annual and interannual variations of the subtropical high over the northwest Pacific Ocean, and discuss the correlation among the SSTA in the eastern equatorial Pacific Ocean, the subtropical high in summer, and the East Asia summer monsoon. Their results show that East Asia summer monsoon becomes weaker, and the subtropical high is stronger and extends further southward and westward in summer if SSTA increases in the eastern equatorial Pacific Ocean in spring the same year. Following the SSTA increases in spring, the rainfall becomes increased in the Changjiang River Basin in summer and decreased in South China, the Hetao area and the area to the east of Hetao. The reverse is also true if the SSTA decreases in spring. Numerical diagnostic study for the effects of the duration of the positive SSTA in the eastern equatorial Pacific Ocean in winter and spring on the East Asia summer climate is conducted by Long and Li (1999). Their simulating results from LAP 2 layer AGCM and 9 layer spectrum AGCM show that the atmospheric response and climate abnormality in East Asia affected respectively by the positive SSTA durations (January only, January-February, January-April) are quite similar to each other. They find that the strength of positive SSTA might play a more important role to the summer climate rather than the positive SSTA duration. Therefore the greater attention should be paid to the stronger positive SSTA even if its duration might be shorter. Yu et al. (2001) use the monthly precipitation data (1951-1998) of the 115 stations of Eastern China and global SST data of COADS to calculate lag correlation coefficient between the regional precipitation and the regional SST, and to study the oceanic predominant region in forcing the eastern China summer monsoon rainfall. They find that the seasonal variations of rainfall at any one of the 4 sub-regions are similar to each other in the eastern China, with a monsoon rainfall peak in Jane or July. They work out the respective lag correlation coefficients between summer monsoon rainfall in the Changjiang River sub-region and the mean SST in each of the 7 sub-regions in the Pacific Ocean and the Indian Ocean with SST leading the rainfall, and obtain the regression equation of the summer monsoon rainfall for the sub-region, which depends upon mean SST at each oceanic sub-region with respective leading times. The hindcast results for the summer monsoon rainfall by use of the regression equation are consistent with the observed rainfall. Li and He (2001) study the interdecadal features of East Asia monsoon and the monsoon rainfall over North China, and their relations with North Pacific SSTA by use of the NCEP/NCAR reanalysis data (2.5°´2.5°), the SST data and rainfall data of National Climate Center of China. They find that the East Asia summer monsoon had been stronger before 1926 together with the rainfall abundance of North China. And it becomes weaker after 1976 together with rainfall deficiency of North China. They also find that the correlation between the rainfall of North China and the SSTA in the North Pacific Ocean had been much better before 1976
2. Interaction between Asia Monsoon and ENSO
A large number of papers concentrate on the interaction between Asia monsoon and ENSO. Huang (2001) uses rainfall data (1951-1999) of China to analyze the association of drought or flood with various phases of a ENSO cycle. He finds that the summer monsoon rainfall tends to be abundant in the Changjiang River and Huaihe River Basin of China, Japan, and Korea if it coincides with the developing phase of an ENSO cycle, and it tends to be deficient in North China and South China. A reverse summer rainfall distribution can be found in the ENSO decaying phase. Therefore heavy disaster of drought or flood often occurs in China in the ENSO developing phase or decaying phase (Ni and Sun, 2000). Jin and Tao (1999) also find that the rainfall in East China is closely related to ENSO cycle, Tao and Chang (1998) analyze the geopotential height fields and atmospheric circulation patterns respectively for El Nino and La Nina, indicating that a certain circulation pattern corresponds to winter monsoon weakening in East Asia in an ENSO year. To explain the abnormality of Asia monsoon circulation in an ENSO cycle, Ren and Huang (1999) attribute it to the abnormal convection pattern forced by the SSTA distribution related to ENSO in the tropical Pacific Ocean, i. e. the thermal convection strengthening in the middle and eastern equatorial Pacific Ocean and the weakening in the western equatorial Pacific Ocean result in a dipole of the convective activity or the thermal source over the tropical Pacific Ocean which is favorable for strengthening anticyclonic circulation over South China Sea and southwest monsoon air-flow over the Changjiang River Basin and Huaihe River Basin. Wang et al. (2001) further indicate the ENSO effect on the East Asia summer monsoon circulation lags behind the ENSO evolution, i.e. the anti-cyclonic circulation becomes strengthened in Northeast Asia, and the subtropical high becomes further westward extended in summer after an ENSO event has reached its mature phase. As the Asia summer monsoon usually brings a lot of water vapor into India, South Asia, and East Asia, influencing the vaper transport for the above area, ENSO effects on the summer monsoon circulation will be an important role in the regional vapor transport. Huang et al. (1998) find that the longitudinal vapor transport by East Asia monsoon is much greater than the latitudinal, and water vapor convergence, which causes the monsoon rainfall, depends upon both the monsoon circulation pattern and the water vapor advection path.
In addition to studies of vapor transport and its sources related to summer monsoon, the coincidence of ENSO occurrence with strengthening of the winter monsoon in East Asia is investigated by means of the observational data (1950-1989) analysis (Mu and Li 1999). Mu and Li find that the winter monsoon has become stronger (weaker) in East Asia before an El Nino (La Nina) event occurs. They indicate that ENSO signals are better reflected and included in the interannual anomaly of the winter monsoon in East Asia. Zou, Wu and Ni (2002) further confirm that the Asia winter monsoon coming down southward can excite the strong convective activities in the western tropical Pacific Ocean and influence the ENSO cycle. They point out that the influence is reflected in the quasi-biennial oscillation mode. Therefore the Asia winter monsoon plays an important role in ENSO occurrence. On the other hand, ENSO will influence East Asia monsoon. Long and Li (1999), using AGCM heated by the typical SSTA of the eastern Pacific Ocean during ENSO, study the effect of a typical ENSO event on the following monsoon activity in East Asia. They find that in the summer after an ENSO event has occurred, the subtropical high becomes stronger. Its ridge tends to be displaced northward and its area extends further westward causing rainfall deficiency in East China and North China. In winter after La Nina has occurred, rainfall deficiency happens in the Changjiang River and Huaihe River Basin, causing a different rainfall distribution.
A review of the studies for the interaction between the ENSO cycling and the Asia monsoon was published (Huang and Chen 2002) and more details about the recent progress in this aspect can be found there.
3. South China Sea Monsoon
Zhao et al. (2000a) analyze the SSTA field in May for the South China Sea and the eastern tropical Pacific Ocean by means of EOF method to study the South China Sea monsoon onset and its association with the SSTA field. They find that the positive maximum value of the first characteristic vector is centered at Sumatra of Indonesia showing its closest relation with the monsoon onset. The South China Sea monsoon starts earlier (later) when SST becomes warmer (cooler) in the sea area around Sumatra than usual. They also (2000b) analyse the seasonal and the intrannual variations of the South China Sea warm pool (with sea temperatare greater than 28℃) and discuss their relation with the South China Sea monsoon onset. The results are that the warm pool has an obvious seasonal variation with its largest area from June to May and the smallest area from December to February. The rapid transitional months are respectively March-April and October-November, leading to the monsoon seasonal reverse. The monsoon onset can be closely related to the warm pool interannual variability. The summer monsoon starts later (earlier) than usual if the warm pool becomes warmer (cooler) in the previous winter and spring. The subsurface variability of the tropical Pacific-Indian Ocean and its correlation with the strength of the South China Sea monsoon are analysed by Zhang, Li and Wang (2001, 2002). They find (2001) that the thermocline depths of the central equatorial Pacific Ocean and Bay of Bangal are positively and closely correlated with the strength of the monsoon and can be used as a predictor for monsoon prediction. They also (2002) find that the sea temperature at the 120 m water depth is the best correlated with the monsoon strength. They classify all the South China Sea monsoon cases (1995-1998) into 4 kinds and discuss the subsurface sea temperature field at the 120 m depth and the wind field on 850 hPa level by means of the composite analysis method respectively for the 4 kinds indicating their scientific significances for the monsoon and ENSO prediction.
IV. DECADAL-INTERDECADAL VARIABILITY IN THE PACIFIC OCEAN
Data analysis for surface air temperature, the subtropical high activities over the Northwestern Pacific, and the SSTA in the equatorial Pacific show that there are clear quasi-decadal oscillation signals (C. Li 1998). He also find that the climate variations in East Asia, such as the surface air temperature, the precipitation and the onset date of Meiyu in the Yangtze (Changjiang) River Basin have the quasi-decadal oscillation similar to that mentioned above. He deduces from the data analysis that it is the quasi-decadal oscillations of the whole air-sea system that influence those climate variations and the oceanic variations.
Signals of the interdecadal variability in the Pacific Ocean are examined by use of CEOF method and composite analysis method for XBT temperature data obtained in the period from 1950 to 1993 (Wang and Liu 2000). Such signals are propagated southwestward from the central North Pacific and subducted to the subtropical region. They find that the themal anormalies subducted from the central North Pacific to the east of dateline can only reach 18°N in the western tropical ocean. There has been no further southward propagation than 18°N due to a certain barrier. However the origin of the interdecadal oceanic signal in the western tropical Pacific can be traced to the southern tropical Pacific. These variabilities reflect the nature of the thermocline circalation. Yang et al. (2002) analyse the data (1950-1998) from the Maryland Ocean Data Assimilation and the NCEP/NCAR atmospheric reanalysis data to study spatio-temporal structures of interannual and interdecadal variations in the global ocean-atmosphere system. They conclude that the interannual variations in the global upper oceans are characterized by the ENSO mode, which dominated in the tropical Pacific while the oceanic interdecadal variations are mainly confined to the mid-to-high latitudes, the off-equator Pacific and Atlantic. On the other hand, the atmospheric interannual and interdecadal variations both appear in the mid-to-high latitudes especially over the polar regions, with less coherence between the interannual air temperature variation and the interannual sea level pressure variation but better coherence between the interdecadal two. The interannual variation of the mid-to-high latitude atmosphere seems to be an instinct variation, which can not be related to the interannal variation of the tropical upper ocean with ENSO features. They also find that an abrupt change in the global air-sea system happened in 1970s, such as air temperature increase in a large area of the world especially in Siberia and the polar regions, SST increase in the eastern tropical Pacific Ocean and the western coastal waters of North America, South America and Africa, air-temperature decrease in the North Tibet and Greenland, and SST decrease in mid latitude water of the north Pacific Ocean and high latitude water of the Southern Hemisphere.
Zhu et al. (2002) analyse the data (1951-1995) of the geopotential height of 500 hPa and monthly mean SST by means of SVD method to study the coupled pattern of the mid and low latitude atmosphere/ocean interaction and its decadal variation. They find that the PNA teleconnection pattern is closely related to El Nino and the Pacific-Japan (PJ) teleconnection pattern to La Nina. They also find that the abrupt changes of the coupled patterns happened in the middle 1970s, with the significant SST increase in the central and eastern tropical Pacific Ocean and the decrease in the middle north Pacific Ocean. The PNA pattern in winter becomes strengthened after the middle 1970s and no PJ pattern appears before the middle 1970s.
More details of the interdecadal variability of the air-sea interaction system in the Pacific region can be found in the paper by H. Li et al. (2001).
V. AIR-SEA INTERACTION IN THE NORTH PACIFIC OCEAN
Zhang, Yu and Liu (1998) analyse the sea surface turbulent heat flux anomalies (including sensible and latent) and their effects on SSTA in the North Pacific Ocean in wintertime using the data extracted from a long-term integration of a coupled ocean-atmosphere general circulation model. The relative importance of the sea surface heat flux components for determining the SSTA variability has been learnt to be compared with the other oceanic dynamic processes. They find that the turbulent heat fluxes play a major role to the SST variability in the most of the extratropical North Pacific Ocean except in a central patch of the basin where the effect of the oceanic horizontal advection on SST can not be negligible. The latent and sensible heat flux anomalies for the model are closely related to the SST tendency rather than SSTA in the extratropical Pacific Ocean during wintertime. Their results coincide with the analysis results by Cayan from COADS and by Reynolds et al from the NCEP data. It is supported by the model runs that the atmospheric role on the ocean is dominant in the air-sea interaction over the extratropical ocean in wintertime rather than the oceanic forcing on the above atmosphere, while ocean plays a dominant role in the tropical ocean. The first EOF mode of the SST tendency from the model output and its correlations with the sea level pressure are similar to the results by Wallace et al. from the observational data. Therefore it is the anomalous large-scale atmospheric circulation that affects SSTA in the most of the extratropical Pacific Ocean especially in the western Pacific through the sea surface turbulence heat fluxes. Jiang (1998) studies the heat transfer between air and sea in the mid-latitude North Pacific Ocean. He finds that the heat transfer from sea to air has obvious seasonal features becoming the greatest in December and the least in June. The spatial distribution of the heat transfer is closely related to the distribution of oceanic currents. It is greater in the warm current area and less in the cold current. The seasonal variations of the heat transfer between air and sea is more obvious near the west boundary where Kuroshio or Oyashio flows than near the east boundary where Alaska current or California current flows. The meridional gradient of the heat transfer becomes the greatest in 37.5°-43.5°N, west of 160° in March and almost vanishes in June. The sea surface heat budget also has seasonal variation with the net heat from sea to air in December in any area of the North Pacific Ocean and the net heat from air to sea in June in any area of the ocean. The heat budget ratio, which depends upon the latitude of the sea area, has a different tendency in December from the tendency in June.
In addition to the extratropical water heat budget studies, Xie and Liu (2001) detect a wind wake trailing westward behind the Hawaiian Islands for 3000 km, many times longer than observed anywhere else. The wind wake drives an eastward oceanic current and draws warm water 8000 km away from the Asia coast, leaving marked changes in surface and sub-surface temperature. Located in the steady trade wind zone, Hawaii triggers a kind of air-sea interaction, which provides the feedback to sustain the influence of these small islands over a long stretch of the North Pacific.
VI. AIR-SEA INTERACTIONS IN THE INDIAN OCEAN
Zhou and Zhang (2002) study the air-sea heat flux exchange of the Indian Ocean by use of the UWM/COADS data. They find that SSTA is closely related to the net sea surface heat flux in the tropical Indian Ocean, especially in the wintertime in the central-eastern Indian Ocean and in the summertime in the western and northern Indian Ocean. However in the other area of the ocean, the evolution of SSTA has less correlation with the net sea surface heat flux. It is suggested that SSTA evolution is dominated by the oceanic dynamical processes in the other area. The normalized covariance analysis shows that the latent heat prevails over the sensible heat, short wave radiation or long wave radiation in forcing the atmosphere in the tropical Indian Ocean. Although the sensible heat flux is less in quantity than the latent in 20°N-20°S, its zonal mean is more significantly correlated to SST rather than the latent heat flux.
Li and Mu (2001a and 2001b) study the SST variation with its dipole oscillation feature in the equatorial Indian Ocean by use of the observational data as long as about 100 years. The dipole oscillation feature with higher SST in the west of the ocean and lower SST in the east during its positive phase and with higher SST in the east and lower SST in the west during its negative phase becomes more obvious in September -November and less obvious in January-April. Generally speaking, the amplitude of the positive phase is greater than the negative. The interannual variation of the dipole has a 4-5 year periodicity and the interdecadal a 25-30 year periodicity. Significant impact of the dipole on the Asian monsoon activity can be found because the wind field in the lower troposphere over South Asia, the Tibetan high in the upper troposphere and the subtropical high over the northwest Pacific Ocean are all related to the dipole. There is also some scientific evidence for the dipole effects on the atmospheric circulation in North America and the southern Indian region (including Australia and South Africa). Wu et al. (2000) use the thermal adaption theory to study the impact of SSTA in the Indian Ocean on the weather of South China and the subtropical high over the western Pacific Ocean. It is shown that in the first stage of the atmospheric thermal adaption to the SSTA in the ocean the in situ anomalous cyclonic circulation of the lower troposphere and the deep convective precipitation to its east due to the southerlies strengthening will be generated and in the second stage of the adaption to the latent heating from the condensation of the precipitation anomalous anticyclonic circulation will be producecd at 500 hPa over the western Pacific Ocean and at 200 hPa over South Asia. Therefore their conclusion is that the both stages of the atmosphieric thermal adaption to the SSTA in the Indian Ocean are important mechanisms for the abnormality of the subtropical high with the corresponding climate variability over the East Asian monsoon region.
Xiao et al. (2000) use the IAP-GCM9L model to simulate the atmospheric response to the SSTA pattern during ENSO occurrence phase with warm SST in the western Indian Ocean and cool SST in the eastern and to learn the effects of the SSTA pattern on Asia climate. It is already known from published literature that the equatorial Pacific SSTA pattern during ENSO brings about the global climatic abnormality. If the SSTA pattern of the Indian Ocean is taken into account the climate abnormality will become enhanced in Bengal Bay, Indo-China Peninsula, Indonesia, India, and China. The main abnormality in climate is that the Indio-China Peninsula becomes much drier, rainfall decreases in North China and increases in a zone along the Changjiang River from Southwest China to Southeast China.