AUGUST 1999 NOTES AND CORRESPONDENCE 2711 NOTES AND CORRESPONDENCE Annual Variation of Surface Pressure on a High East Asian Mountain and Its Surrounding Low Areas TSING-CHANG CHEN Atmospheric Science Program, Department of Geological and Atmospheric Sciences, Iowa State University, Ames, Iowa MING-CHENG YEN Department of Atmospheric Science, National Central University, Chung-Li, Taiwan 14 October 1998 and 9 February 1999 ABSTRACT An inverse annual variation is observed between surface pressure on the highest mountain, which has an elevation of approximately 4000 m, and in the lowlands of Taiwan (a subtropical island in east Asia). This inverse annual variation in surface pressure of high and low elevation in low latitudes reflects, essentially, a vertical phase reversal of the tropical circulation, which is illustrated with the annual variation in the vertical structure of tropical geopotential height. 1. Introduction Landmass is warmer than the ocean in the summer, and this land ocean thermal contrast is reversed during the winter. As a result, air mass (i.e., surface pressure divided by gravity) is heavier (lighter) over the landmass and lighter (heavier) over the ocean during the winter (summer). Evidently, the surface high (low) system is generally associated with low (high) surface temperature. Thus, seasonal variation in the land ocean surface pressure contrast is opposite to that in thermal contrast. Examining the seasonal variation of atmospheric mass, van den Dool and Saha (1993) pointed out that some high elevations may have an annual cycle of surface pressure opposite to that of the major landmass. Saha et al. (1994) elaborated further this out-of-phase relationship in the mean monthly surface pressure anomaly between the Tibetan plateau and its surrounding lowlevel landmass. This opposite-phase variation of surface pressure was later confirmed by Chen et al. (1997). Based upon some cross sections of geopotential height anomalies through Tibet, Saha et al. (1994) offered a trivial explanation of this interesting phenomenon, arguing that the annual variation of surface pressure in Tibet may be comparable to its own free atmospheric large-scale environment at 650 hpa. Corresponding author address: Tsing-Chang (Mike) Chen, Atmospheric Science Program, 3010 Agronomy Hall, Department of Geological and Atmospheric Sciences, Iowa State University, Ames, IA 50011. E-mail: tmchen@iastate.edu Taiwan is a subtropical east Asian island with an area of approximately 30 000 km 2 that is separated primarily into two major geographic entities (east and west) by a north south-oriented mountain range whose highest peak reaches 4000 m. Using different timescales in our recent analysis of the surface pressure variation of this island (Chen et al. 1998a,b), an opposite-phase annual variation of monthly mean surface pressure between the tallest mountain (Yu-Shan) of Taiwan and its surrounding lowlands and sea caught our attention as being consistent with Saha et al. s finding. However, the size of Taiwan makes this contrast of surface pressure undetectable to the global data assimilation system of any operational center. The purpose of this short note is to add to surface observations a new example of the inverse relationship between the annual variations of surface pressure on a tall tropical mountain and its lowlevel surroundings. The annual variation of the east Asian climate system primarily follows the alternation of two opposite monsoons (i.e., the southwesterly summer and the northeasterly winter east Asian monsoons) (e.g., Ramage 1971; Fein and Stephens 1987). An explanation of the inverse relationship of annual variation in surface pressure between high and low elevations in Taiwan is presented from a perspective of the annual variation of the large-scale monsoon circulation. 2. Annual variation of surface pressure Taiwan has a total of 26 surface meteorological stations: four stations on small islands around Taiwan, three mountain stations above an elevation of 1000 m, 1999 American Meteorological Society
2712 JOURNAL OF CLIMATE VOLUME 12 FIG. 1. Geography and orography of Taiwan with locations of two surface meteorological stations marked by open asterisks: Pen-Hu (WMO station 46735) and Yu-Shan (WMO station 46755). Elevation of the latter station is 3845.7 m above sea level. The orography (H) is indicated by the following convention: 500 m H 1000 m (light stippling), 1000 m H 2000 m (moderate stippling), 2000 m H 3000 m (heavy stippling), and 3000 m H (dark stippling). and 19 low-level stations. One of the mountain stations is located at Yu-Shan (Jade Mountain, the tallest mountain in Taiwan) whose elevation is 3845.7 m. The annual mean surface pressure and temperature of that station [marked by an open asterisk in Fig. 1 with the World Meteorological Organization (WMO) station number 46755] are 641.6 mb and 4.4 C, respectively, with their annual amplitudes of 2 mb and 2.8 C. While the surface pressure is comparable to Lhasa, Tibet, the temperature is somewhat warmer, particularly in winter. Regarding the annual variation, surface pressures of 25 stations below 3000 m are coherent. To save space, we present in Fig. 2 only station surface pressures of Pen-Hu (thin, solid line), a small island situated in the middle of the Taiwan Strait (marked by an open asterisk in Fig. 1 with WMO station number 46735) and Yu-Shan (thick, solid line). Let us denote surface pressures of these two stations as p s (Pen-Hu) and p s (Yu-Shan) for convenience. Their annual variations are characterized by the following salient features. 1) As indicated by p s (Pen-Hu), a summer minimum and a winter maximum stand out in the surface pressure of the lowlands and sea around Taiwan. The annual variation of surface pressure in the vicinity of this island basically follows that in the giant landmass of Asia. This relationship will be illustrated further in the next section. 2) The semiannual variation of tropical circulation depicted by previous studies (e.g., van Loon and Jenne 1970; Weickmann and Chervin 1988; Chen et al. 1996) is discernable in the p s (Pen-Hu) time series. Saha et al. (1994) depicted a relationship between the Tibetan plateau and its surroundings; a clear inverse relationship also exists between annual variations of p s (Pen-Hu) and p s (Yu-Shan).
AUGUST 1999 NOTES AND CORRESPONDENCE 2713 FIG. 2. Time series of monthly mean surface pressure ( p s ) at Yu-Shan (thick solid line) and Pen-Hu (thin solid line) for 1960 96. 3) Table 1 of Saha et al. (1994) showed that the amplitudes of the annual variation in surface pressure are much larger over India and China than over Tibet; the annual change of surface pressure between January and July is about 17 mb at Calcutta and only about 6 mb at Lhasa. This amplitude contrast is also true between Yu-Shan and Pen-Hu. 3. Annual variation of the low-latitude planetary circulation During the northern winter, major troughs exist to the east of the North American and east Asian continents, and a minor trough extends equatorward from central Eurasia to North Africa. Ridges associated with these troughs are located to the west. Because of their baroclinic nature, stationary waves representing these asymmetric components of the Northern Hemispheric circulation exhibit a well-organized westward tilting (Lau 1979). The existence of these baroclinic stationary waves in mid- and high latitudes is reflected by a horizontal phase change at 30 N. South of this latitude, a vertical phase reversal of stationary waves emerge (as revealed in the right column of Fig. 3). In the northern summer, the lower-tropospheric circulation is dominated by the North Pacific and North Atlantic anticyclones, and the Asian continental low (which covers the Asian monsoon region). These major low tropical circulation elements are overlaid by the two oceanic troughs and the Tibetan high (Krishnamurti 1971a,b). The vertical reversal of the summer circulation structure is clarified further by the vertical phase change of the summertime tropical stationary waves depicted by White (1982); this represents the monsoonal (instead of baroclinic) characteristics of the planetary-scale circulation in the Tropics. In view of the monsoonal nature of the tropical circulation, the low- and high-elevation surface pressure (or height) should vary out of phase, as long as the latter elevation is above the altitude of the vertical phase change of the tropical circulation and the mountain does not change the climate. If this argument is substantiated, the cause of the inverse relationship between the annual variations of surface pressure at the tall tropical mountain and its surrounding lowlands can be explained. To explore the cause of the opposite-phase annual variation of surface pressure suggested above, we display in Fig. 3 [with the National Centers for Environmental Prediction National Center for Atmospheric Research (NCEP NCAR) reanalysis data (Kalnay et al. 1996) for the period 1960 97] the seasonal mean departures of surface and upper-air geopotential heights from their yearly mean values. The major features of these anomaly charts are highlighted as follows. 1) The alternation of the summer continental low system (associated with the Asian monsoon) and the winter continental high system can be seen clearly from the contrast between Figs. 3d and 3h. From these two figures, one can observe that the annual variation of Taiwan surface pressure follows the continental-scale annual variation of Asian surface pressure. 2) The inverse annual variations of surface pressure between the Tibetan plateau and its surrounding lowlands are also apparent from the comparison between Figs. 3d and 3h. However, due to the small geographic size of Taiwan, the inverse annual variations of surface pressure over this island are revealed visibly in the contrast between these two figures. 3) The winter geopotential height is generally smaller than the summer value. Regardless of this seasonal
2714 JOURNAL OF CLIMATE VOLUME 12 FIG. 3. The summer (Jun Aug) and winter (Dec Feb) departures of surface pressure and geopotential height at different levels from their annual mean values. The Z (650 mb) chart is constructed with the average of Z (600 mb) and Z (700 mb) computed directly from the NCEP NCAR reanalysis data. Contour intervals of p s, Z (850 mb), Z (650 mb), and Z (200 mb) are 2 mb, 10 m, 20 m, and 50 m, respectively. Positive values of p s and Z are stippled, while values of p s 850 mb are darkened in (b) and (g). change, the vertical reversal of pressure (height) anomalies occurs between 700 and 600 mb over the Asian monsoon region. It is likely that the annual variation of p s (Yu-Shan) follows the circulation regime above this atmospheric layer. 4) The annual variation of surface pressure and upperair geopotential height over the Asian continent is coupled to the east with the alternation of the summer North Pacific anticyclone [indicated by positive p s ( Z) anomalies in the left column of Fig. 3] and the winter Aleutian low/east Asian trough [indicated by negative p s ( Z) anomalies in the right column of
AUGUST 1999 NOTES AND CORRESPONDENCE 2715 anomaly centers. This vertical phase reversal of Z (25 N) anomalies is consistent with the vertical structure of stationary waves portrayed by White (1982, his Fig. 4) at 30 N. As indicated by the orography of Yu-Shan, p s (Yu-Shan) belongs to the upper-tropospheric monsoon circulation regime, and p s (Pen-Hu) is a part of the lower-tropospheric monsoon circulation. The argument so far agrees with Fig. 4 that the inverse annual variations of p s (Pen-Hu) and p s (Yu-Shan) essentially reflect the out-of-phase annual variation of the lowerand upper-monsoon circulation in the Tropics. The illustration of the inverse annual variations between p s (Pen-Hu) and p s (Yu-Shan) with Figs. 3 and 4 is also applicable to Saha et al. s (1994) suggestion that the annual variation of the Tibetan surface pressure should be a part of the large-scale environment at the 650-mb level. FIG. 4. Longitude height cross sections of the summer and winter geopotential height departures from their annual mean values at 25 N, Z (25 N). The orography is darkened and Yu-Shan is denoted by a straight-up thick solid line. The contour interval of Z (25 N) is 25 m and positive values of Z (25 N) are stippled. Fig. 3]. Evidently, the annual variation of p s over the Asian monsoon region is a part of the annual change of the planetary-scale circulation. In order to illustrate further the cause of the inverse annual variations of p s (Pen-Hu) and p s (Yu-Shan), the longitude height cross sections of height anomalies at 25 N (constructed with the NCEP NCAR reanalysis of the period 1960 97) are shown in Fig. 4 where the orography is darkened. Yu-Shan is located at 23 29 N, 120 57 E, only slightly south of the cross section, and is denoted by a thick, solid line. The horizontal p s and Z anomaly charts in Fig. 3 reveal that the continental heating (cooling) in summer (winter) results in the landmass low (high) system. Thus, the Z (25 N) anomalies around major mountains are negative (positive) in summer (winter). These lower-tropospheric Z (25 N) anomaly centers are overlaid aloft by the opposite-sign 4. Remarks The contrast of the p s (Pen-Hu) and p s (Yu-Shan) annual variations in Taiwan offers us another example of the inverse annual variations of surface pressure between a tall tropical mountain and its surrounding lowlands in the Tropics as observed by Saha et al. (1994). The general perception that surface pressure should belong to the lower troposphere might be misleading. As demonstrated in this short study, the inverse relationship of annual variations in surface pressure between a tall tropical mountain and low lands is actually an indicator of the vertical phase change in the planetary-scale monsoon circulation. This inverse annual variation of surface pressure is not only interesting in its own right, but it may be also significant to the study of long-term climate change. The NCEP NCAR reanalysis data (Kalnay et al. 1996) are available at the present time only for the period 1958 97. However, some station observations from the tall mountains and lowlands in low latitudes may be traced back to the turn of the century, and this may allow some implication of the interdecadal variation of the tropical atmospheric circulation to be derived. Finally, we should also clarify the reason for the inverse annual variation of surface pressure not to occur in mid- and high latitudes. As pointed out previously, the baroclinic nature of asymmetric component of the mid high-latitude circulation is characterized by a vertical westward tilting (Lau 1979), instead of a vertical phase reversal in its vertical structure, as the asymmetric circulation component in the Tropics. Therefore, it is unlikely to have inverse annual variations of surface pressure between a tall mountain and its surrounding lowlands, in mid- and high latitudes. Acknowledgments. T.-C. Chen s effort for this study is supported by the NSF Grant ATM-9416954 and the NASA Grant NAG57530, while M.-C. Yen s effort is supported by the NSC Grant 88-2111-M-008-028-AP7
2716 JOURNAL OF CLIMATE VOLUME 12 of Taiwan. Comments given by Dr. Hung van den Dool on this paper were helpful in improving our presentation. We thank Mr. S.-Y. Wang and Mr. I.-S. Wong for their graphic assistance. The typing and editing support provided by Mrs. Reatha Diedrichs and Ms. Dana Baldridge, respectively, are highly appreciated. REFERENCES Chen, T.-C., H. van Loon, and M.-C. Yen, 1996: An observational study of the tropical subtropical semiannual oscillation. J. Climate, 9, 1993 2002., J.-M. Chen, S. Schubert, and L. L. Tackacs, 1997: Seasonal variation of global surface pressure and water vapor. Tellus, 49A, 613 621., M.-C. Yen, and R. Arritt, 1998a: Detection of semidiurnal wind oscillations with a radar profiler. Bull. Amer. Meteor. Soc., 79, 1921 1924.,, J.-C. Hsieh, and R. Arritt, 1998b: Preliminary research results of the rainfall measured by the Automatic Rainfall and Meteorological Telemetry System in Taiwan: Diurnal and seasonal variations. Bull. Amer. Meteor. Soc., in press. Fein, J. S., and P. L. Stephens, 1987: Monsoons. John Wiley and Sons, 632 pp. Kalnay, E. M., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437 471. Krishnamurti, T. N., 1971a: Observational study of the tropical motion field during the northern summer. J. Appl. Meteor., 10, 1066 1096., 1971b: Tropical east west circulations during the northern summer. J. Atmos. Sci., 28, 1342 1347. Lau, N.-C., 1979: The observed structure of tropospheric stationary waves and local balances of vorticity and heat. J. Atmos. Sci., 36, 996 1016. Ramage, C. S., 1971: Monsoon Meteorology. Academic Press, 296 pp. Saha, K., H. van den Dool, and S. Saha, 1994: On the annual cycle in surface pressure on the Tibetan Plateau compared to its surroundings. J. Climate, 7, 2014 2019. van den Dool, H., and S. Saha, 1993: Seasonal redistribution and concentration of atmospheric mass in a general circulation model. J. Climate, 6, 22 30. van Loon, H., and R. J. Jenne, 1970: On the half-yearly oscillations in the tropics. Tellus, 22, 391 398. Weickman, K. M., and R. M. Chervin, 1988: The observed and simulated atmospheric seasonal cycle. Part I: Global wind field modes. J. Climate, 1, 265 289. White, G. H., 1982: An observational study of the Northern Hemisphere extratropical summertime general circulation. J. Atmos. Sci., 39, 28 40.