Abstract

The summer (June–August) Asian–Pacific Oscillation (APO) measures the interannual variability of large-scale atmospheric circulation over the Asian–North Pacific Ocean sector. In this study, the authors assess the predictability of the summer APO index interannual variability and the associated atmospheric circulation anomalies using the 1959–2001 hindcast data from the European Centre for Medium-Range Weather Forecasts (ECMWF), Centre National de Recherches Météorologiques (CNRM), and the Met Office (UKMO) general circulation models from the Development of a European Multimodel Ensemble System for Seasonal-to-Interannual Prediction (DEMETER) project. The results show that these models predict the summer APO index interannual variability well and have higher skill for the North Pacific than for the Asian upper-tropospheric temperature. Meanwhile, the observed APO-related atmospheric circulation anomalies in the South Asian high, the tropical easterly wind jet over the Asian monsoon region in the upper troposphere, the subtropical anticyclone over the North Pacific, and the summer southwest monsoon over Asia in the lower troposphere are reasonably well predicted in their spatial patterns and intensities. Compared with the observations, however, these models display low skill in predicting the long-term varying trends of the upper-tropospheric temperature over the Asian–North Pacific sector or the APO index during 1959–2001.

1. Introduction

Understanding and predicting the variability of large-scale teleconnections are important to correctly predict atmospheric circulation and climate anomalies in different regions. There are many large-scale atmospheric teleconnections, such as the North Atlantic Oscillation (NAO; Walker and Bliss 1932), the North Pacific Oscillation (NPO; Rogers 1981), the Pacific–North American (PNA) pattern (Wallace and Gutzler 1981), and the Pacific–Japan (PJ) pattern (Nitta 1986, 1987). Recently, an extratropical large-scale teleconnection pattern, called the Asian–Pacific Oscillation (APO), was identified in the summer upper-tropospheric temperature over the Asian–North Pacific Ocean sector (Zhao et al. 2007). A positive (negative) APO phase corresponds to a higher (lower) tropospheric temperature over Asia and a lower (higher) tropospheric temperature over the North Pacific and indicates a strengthened (weakened) thermal contrast between Asia and the North Pacific (Zhao et al. 2012).

The interannual variability of the summer APO is closely associated with the large-scale atmospheric circulation anomalies, such as the South Asian high, the subtropical anticyclone over North Pacific, the tropical easterly jet over South Asia, and the Asian monsoon and rainfall (Zhao et al. 2007, 2012). Moreover, a close relationship between the APO and sea surface temperatures (SSTs) over the North Pacific is further observed (Zhao et al. 2010, 2012). On interdecadal time scales, the APO is also associated with atmospheric circulation and precipitation anomalies over the Asian–North Pacific region (Zhao et al. 2011).

Meanwhile, some climate models have been utilized to simulate the APO and associated climate anomalies. Using the National Center for Atmospheric Research (NCAR) Community Climate System Model, version 3.0 (CCSM3.0), and the Community Atmospheric Model, version 3.0 (CAM3.0), Nan et al. (2009) compared the relationship between the summer APO and Pacific SSTs in CCSM3.0 with observations, and Zhao et al. (2010) further compared the modeled summer APO and upper-tropospheric circulation anomalies with the observations. Moreover, Man and Zhou (2011) used the coupled climate system model, the Flexible Global Ocean–Atmosphere–Land System Model gridpoint, version 1.0 (FGOALS-gl.0), developed by the State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics and the Institute of Atmospheric Physics of the Chinese Academy of Sciences, to simulate a long-term change in the summer APO index during the last millennium, and further compared the model results with proxy-reconstructed data. The correlation coefficient between the reconstruction and the simulation during A.D. 1000–1985 is 0.50 (significant at the 99% confidence level). These studies demonstrated that the summer APO variability and associated major climate anomalous features may be well captured by some climate models.

In recent years, the multimodel ensemble seasonal forecasts have been effectively achieved though the Development of a European Multimodel Ensemble System for Seasonal-to-Interannual Prediction (DEMETER) project (Palmer et al. 2004). It has been demonstrated that DEMETER is intrinsically more useful and effective than the forecasts from any one (e.g., national) model (Doblas-Reyes et al. 2005; Hagedorn et al. 2005). In particular, the performances of El Niño–Southern Oscillation (ENSO; Palmer et al. 2004; Jin et al. 2008), the NAO and PNA patterns (Fil and Dubus 2005; Johansson 2007), tropical storm (Vitart 2006; Sun and Chen 2011), regional precipitation prediction (Diez et al. 2005; Kumar et al. 2005; Wang and Fan 2009; Chen et al. 2012; Liu and Fan 2013), Arctic and Antarctic Oscillation (Qian et al. 2011), East Asian winter monsoon (Li and Wang 2012), and some other atmospheric and oceanic processes in the DEMETER models have been investigated by many researchers. Furthermore, the DEMETER hindcasts have also been used to predict crop yield (Cantelaube and Terres 2005; Challinor et al. 2005; Marletto et al. 2005) and malaria (Hoshen and Morse 2004; Thomson et al. 2006).

Regarding the Asian summer climate prediction, many researchers have mainly focused on the prediction of the Asian summer monsoon and associated climate (e.g., Fan et al. 2008; Wang and Fan 2009; Fan and Wang 2009; Liu and Fan 2012, 2013). Figure 1 displays the correlation coefficients for summer [June–August (JJA)] precipitation during 1979–2001 between the observation [i.e., the Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP)] and the DEMETER models (see detailed data and method introductions in section 2). According to Fig. 1, the DEMETER models show almost no skill for the monsoon precipitation over Asia and only perform well over the tropical ocean areas. Actually, the low skill of DEMETER models in predicting the Asian summer monsoon has been attributed to the systematic bias (Ma et al. 2012) and the poor performance of the very long breaks–drought relationship (Joseph et al. 2010) and decadal change–related circulation (Fan et al. 2012). Thus, it is necessary to find new clues to predict the Asian summer climate. Based on the above-mentioned relationship between the Asian summer climate and the APO, investigating the predictability of the APO is meaningful.

Fig. 1.

The correlation coefficient between the observed (CMAP) and predicted JJA precipitation from 1979 to 2001 for the (a) ECMWF model, (b) CNRM model, (c) UKMO model, and (d) MME. Shading is significant at the 95% confidence level.

Fig. 1.

The correlation coefficient between the observed (CMAP) and predicted JJA precipitation from 1979 to 2001 for the (a) ECMWF model, (b) CNRM model, (c) UKMO model, and (d) MME. Shading is significant at the 95% confidence level.

The remainder of this paper is organized as follows. We describe the data and analysis methods applied in this study in section 2, and we discuss the climatological features of the summer tropospheric temperature and the APO index pattern’s predictability from the DEMETER general circulation models (GCMs) in section 3. In section 4, we examine the APO-related atmospheric circulation anomalies’ predictability. Finally, a summary and discussion are provided in section 5.

2. Data and methods

The DEMETER project has been funded under the European Union Fifth Framework Environment Programme for the period April 2000 to September 2003. The DEMETER system comprises seven global coupled ocean–atmosphere models that are initialized 4 times each year on the first day of February, May, August, and November. Each hindcast has been integrated for 6 months and comprises an ensemble of nine members (Palmer et al. 2004). In the present study, three coupled ocean–atmosphere DEMETER GCMs [the European Centre for Medium-Range Weather Forecasts (ECMWF), United Kingdom; Centre National de Recherches Météorologiques (CNRM), France; and the Met Office (UKMO), United Kingdom] were chosen since they have longer available hindcast data, and the longer the record of achieved model data, the better chance to examine the predictability. Particularly, the three GCMs with 43-yr (1959–2001) hindcast data initiated on 1 May are selected to assess the JJA APO predictability and the associated climate anomalies. The multimodel ensemble (MME) was calculated as the simple average of the ECMWF, CNRM, and UKMO hindcasts. Additionally, the monthly 40-yr ECMWF Re-Analysis (ERA-40) data (Uppala et al. 2005) are used as a verification dataset and the observed precipitation for the period 1979–2001 comes from CMAP (Xie and Arkin 1997).

Because our study primarily focuses on the interannual variability, a linear trend has been subtracted from the raw time series when performing a linear correlation and regression analysis. Moreover, the linear correlation analysis of the raw time series (with the trend) is also used for a comparison analysis. The statistical significance was assessed using a Student’s t test.

3. Prediction of the upper-tropospheric temperature and APO

Because the APO is defined based on the upper-tropospheric eddy temperature (Zhao et al. 2007), we first examine the features of the upper-tropospheric , in which , where T is the total air temperature and is the zonal-mean (symmetric) temperature. Relative to ERA-40 (Fig. 2a), DEMETER GCMs capture an extratropical pattern of the summer upper-tropospheric (500–200 hPa) T′ reasonably well, with positive values over Eurasia and negative values over the central-eastern North Pacific (Figs. 2b–e). Figure 3 further shows the difference of the climatological-mean T′ between the prediction and observation. In Fig. 3a, the positive values (warm bias) appear in the high latitudes of Eurasia, indicating an overestimated temperature, while negative values (cold bias) appear in the lower latitudes of Eurasia and the high latitudes of the North Pacific, indicating an underestimated temperature. A similar feature can also be found in Figs. 3b–d. Nevertheless, the predicted positive and negative values are generally similar to the observed values in intensity and position. This result suggests the high skill of the three GCMs in predicting the tropospheric T′.

Fig. 2.

The climatology of the JJA-mean upper-tropospheric (500–200 hPa) T′ (°C) during 1959–2001 for (a) ERA-40, (b) the ECMWF model, (c) the CNRM model, (d) the UKMO model, and (e) the MME, in which two boxes represent the regions for Asia and North Pacific, respectively.

Fig. 2.

The climatology of the JJA-mean upper-tropospheric (500–200 hPa) T′ (°C) during 1959–2001 for (a) ERA-40, (b) the ECMWF model, (c) the CNRM model, (d) the UKMO model, and (e) the MME, in which two boxes represent the regions for Asia and North Pacific, respectively.

Fig. 3.

The difference of the climatology (1959–2001) of JJA-mean upper-tropospheric (500–200 hPa) T′ (°C) between ERA-40 and (a) the ECMWF model, (b) the CNRM model, (c) the UKMO model, and (d) the MME.

Fig. 3.

The difference of the climatology (1959–2001) of JJA-mean upper-tropospheric (500–200 hPa) T′ (°C) between ERA-40 and (a) the ECMWF model, (b) the CNRM model, (c) the UKMO model, and (d) the MME.

Following Zhao et al. (2007), we define the Asian tropospheric temperature index (AI) and the North Pacific tropospheric temperature index (PI) as the regionally averaged upper-tropospheric (500–200 hPa) T′ over 15°–50°N, 60°–120°E and 15°–50°N, 180°–120°W (as indicated by the boxes in Fig. 2), respectively, and define the APO index as the arithmetic difference between the AI and PI. Moreover, the skill of the GCMs in predicting the interannual variability of APO is measured by the linear correlation coefficients (CCs) between the predicted and observed detrended APO indices (CC-I). Meanwhile, the CCs of the raw APO indices (with the trend) between the observation and model simulations (CC-II) are also calculated to further assess the predictive abilities of the GCMs (Table 1). Figure 4 displays the summer APO index from 1959 to 2001 for the observations and the GCMs. In the figure, the CC-I values are 0.49 (significant at the 99.9% confidence level), 0.45 (significant at the 99% confidence level), and 0.55 (significant at the 99.9% confidence level) for the ECMWF, CNRM, and UKMO models, respectively. These results indicate that the three DEMETER GCMs are skillful to predict the APO interannual variability during 1959–2001. The MME with the highest CC-I of 0.57 (significant at the 99.9% confidence level) shows the best summer APO prediction. Meanwhile, the ERA-40 APO index was mainly in a positive phase before 1975 and in a negative phase afterward (Fig. 4a), showing a linearly weakening trend of −0.043 (significant at the 99% confidence level) during 1959–2001. However, the linear trends of the predicted APO indices are not significant, and the CC-II values for the summer APO of 0.47, 0.43, 0.43, and 0.51 for the ECMWF, CNRM, and UKMO models, and the MME (Table 1), respectively, are further lower than CC-I. These results may imply a lower skill of these GCMs in predicting the long-term variability of the APO pattern.

Table 1.

The correlation coefficients (CC-II) between the observed and predicted raw summer APO index, AI, and PI (with the trend) during 1959–2001, in which boldface numbers are significant at the 95% confidence level.

The correlation coefficients (CC-II) between the observed and predicted raw summer APO index, AI, and PI (with the trend) during 1959–2001, in which boldface numbers are significant at the 95% confidence level.
The correlation coefficients (CC-II) between the observed and predicted raw summer APO index, AI, and PI (with the trend) during 1959–2001, in which boldface numbers are significant at the 95% confidence level.
Fig. 4.

The normalized JJA APO index during 1959–2001 for (a) ERA-40, (b) the ECMWF model, (c) the CNRM model, (d) the UKMO model, and (e) the MME, in which the correlation coefficient (CC-I) between the predicted and observed detrended summer APO index is shown. In (a) the solid line represents the linear trend of the APO index.

Fig. 4.

The normalized JJA APO index during 1959–2001 for (a) ERA-40, (b) the ECMWF model, (c) the CNRM model, (d) the UKMO model, and (e) the MME, in which the correlation coefficient (CC-I) between the predicted and observed detrended summer APO index is shown. In (a) the solid line represents the linear trend of the APO index.

Figure 5 shows the summer AI from 1959 to 2001 for the observations and the GCM simulations. It is seen that the CC-I values are 0.36 and 0.32 for the ECMWF and CNRM models, significant at the 98% and 95% confidence levels, respectively, while the CC-I (0.23) is not significant for the UKMO model. These results exhibit a higher skill of the ECMWF and CNRM models in predicting the AI interannual variability and a low skill of the UKMO model. Particularly, the MME with the highest CC-I value of 0.40 (significant at the 99% confidence level) improves the prediction skill of the summer AI interannual variability significantly. In addition, similar to the APO, the ERA-40 summer AI index also showed a significant decreasing trend (−0.046, significant at the 99% confidence level) during 1959–2001, with a positive phase before 1975 and a negative phase after that, further supporting the result of Wang (2001). His study showed that the upper-tropospheric temperature over the midlatitudes of Asia turned to a cold phase in the late 1970s, which may contribute to the weakening of the East Asian summer monsoon. However, this linearly decreasing trend is not significant in the predicted AI, and compared to CC-I the CC-II values for the summer AI with apparently lower values of 0.27, 0.23, 0.05, and 0.24 for the ECMWF, CNRM, and UKMO models, and the MME (Table 1), respectively, are also not significant. These results suggest a poor performance of the DEMETER GCMs to predict the long-term variability of AI. In Fig. 6, the CC-I values of the summer PI are 0.52, 0.51, and 0.64 for the ECMWF, CNRM, and UKMO models, respectively, significant at the 99.9% confidence level. This result illustrates that the GCMs perform very well in predicting the summer PI interannual variability. However, the MME with the value of 0.62 does not show the best performance to predict the interannual variability of summer PI, which may attribute to the outstanding skill (0.64) of the UKMO model (Yoo and Kang 2005). Furthermore, contrasting with the weakening trend of AI, during 1959–2001 the PI shows a significant increasing trend (0.034, significant at the 99% confidence level). According to Table 1, although some of the CC-II values for the summer PI are slightly higher than CC-I values, all the predicted PI trends during the entire time period (1959–2001) are not significant.

Fig. 5.

As in Fig. 4, but for the AI.

Fig. 5.

As in Fig. 4, but for the AI.

Fig. 6.

As in Fig. 4, but for the PI.

Fig. 6.

As in Fig. 4, but for the PI.

From the above analysis, the three DEMETER GCMs can predict well the climatological-mean pattern of the summer upper-tropospheric T′ over the Eurasian–North Pacific sector and the APO interannual variability, having a higher predictive skill for the North Pacific than for the Asian tropospheric temperature. The multimodel ensemble seems to help filter out model errors and thus improves the skill in predicting the interannual variability of APO and AI. Moreover, the long-term varying trends of the APO index during 1959–2001 are not well predicted by these models.

4. Predictions of the APO-related atmospheric circulation anomalies

Because the APO variability is closely linked with anomalies of the large-scale atmospheric circulations, such as the South Asian high, the westerly jet over extratropical Eurasia, the subtropical anticyclone over the North Pacific, and the Asian summer monsoon (Zhao et al. 2007, 2010, 2012), in this section, we also examine the predictability of the APO-related atmospheric circulation anomalies in the DEMETER GCMs.

Figure 7 displays the regression map of the summer 200-hPa horizontal winds with respect to the summer APO index during 1959–2001 for the observation and the GCM predictions. In Fig. 7a, when the observed APO index is positive, a large anomalous anticyclone spans the region north of 20°N, indicating the northward South Asian high; additionally, anomalous easterly winds prevail from the North Pacific to Eurasia between 25° and 40°N and anomalous westerly winds prevail over the high latitudes of Eurasia and the North Pacific, indicating a strengthened tropical easterly and extratropical westerly jet streams, respectively, in association. The three DEMETER GCMs generally capture these anomalous features, reasonably predicting a strong South Asian high, a tropical easterly wind, and extratropical westerly wind jet streams over the study region (Figs. 7b–d). The MME (Figs. 7e) shows comparable skill with these three models. However, the predicted anomalous westerly winds are weaker than the observations, particular in the ECMWF and CNRM models. Similar anomalous features with respect to the APO index are also observed from the regression map of the 200-hPa geopotential height (figures not shown).

Fig. 7.

The regression of JJA 200-hPa winds (m s−1) with reference to the normalized APO index for (a) ERA-40, (b) the ECMWF model, (c) the CNRM model, (d) the UKMO model, and (e) the MME, in which the red arrow indicates the prevailing wind direction. Shading is significant at the 95% confidence level.

Fig. 7.

The regression of JJA 200-hPa winds (m s−1) with reference to the normalized APO index for (a) ERA-40, (b) the ECMWF model, (c) the CNRM model, (d) the UKMO model, and (e) the MME, in which the red arrow indicates the prevailing wind direction. Shading is significant at the 95% confidence level.

At 850 hPa (Figs. 8a), associated with an observed positive APO phase, there is an anomalous cyclone over the subtropical western North Pacific and an anomalous anticyclone to the northeast. On the climatological field of the 850-hPa winds (figure not shown), the subtropical anticyclone appears over the North Pacific, with the circulation center near 30°N, 150°W. Thus, the anomalous pattern in Fig. 8a reflects a weakened and northward western North Pacific subtropical high (WPSH) when the APO index is above the normal. Because the WPSH is one of the most important North Pacific atmospheric circulations and has a great effect on the East Asian–North Pacific climate, the relationship between the WPSH and the APO index is further examined. We calculate the correlation coefficient between the WPSH intensity index and the APO index. Here, following the method of the National Climate Center of China (Chen 1999), the intensity index is defined as the accumulated value with a 500-hPa geopotential height greater than 586 dagpm within the domain of 10°–90°N, 110°E–180°. For the observations, the APO index has a significant negative correlation of −0.34 with the WPSH intensity index (see Table 2) during 1959–2001 that suggests a weaker WPSH in association with a positive-APO index. This link between the APO and WPSH is explained as follows. When the APO index is higher, a large-scale lower-tropospheric low pressure system over Asia strengthens, extends eastward, and leads to the eastward expansion of the low pressure circulation to the western North Pacific and weakens the subtropical anticyclone over the western North Pacific (Zhao et al. 2007). Additionally, the anomalous westerly winds prevail over a large tropical area extending from the Arabian Sea to the South China Sea, and anomalous southeasterly or southwesterly winds prevail over northern China (in Fig. 8a), indicating stronger South and East Asian summer monsoons. Moreover, the upper-tropospheric easterly and lower-tropospheric westerly anomalies over the South Asian monsoon region also indicate a strengthened zonal vertical shear in the local zonal wind, implying a strengthened Indian summer monsoon corresponding to a high APO index.

Fig. 8.

As in Fig. 7, but for 850 hPa (m s−1), in which C indicates anomalous cyclonic center, the red arrow indicates the prevailing wind direction, and the thick dashed lines indicate the topographic contour of 1500 m.

Fig. 8.

As in Fig. 7, but for 850 hPa (m s−1), in which C indicates anomalous cyclonic center, the red arrow indicates the prevailing wind direction, and the thick dashed lines indicate the topographic contour of 1500 m.

Table 2.

The correlation coefficients between the WPSH intensity index and the APO index for the observation and the DEMETER models; all numbers are significant at the 95% confidence level.

The correlation coefficients between the WPSH intensity index and the APO index for the observation and the DEMETER models; all numbers are significant at the 95% confidence level.
The correlation coefficients between the WPSH intensity index and the APO index for the observation and the DEMETER models; all numbers are significant at the 95% confidence level.

The DEMETER GCMs also predict the above observed anomalies well. The ECMWF model well predicts the anomalous anticyclone and cyclone over the North Pacific, anomalous westerly winds over the Asian tropical monsoon region, anomalous southerly winds over northern China, and a stronger Asian summer monsoon (Fig. 8b). Moreover, the correlation coefficient between the APO index and the WPSH intensity index is −0.62, also consistent with the observations. The CNRM (Fig. 8c) and UKMO (Fig. 8d) models also capture the subtropical anomalous cyclone over the North Pacific, the anomalous anticyclone to the northeast, the anomalous southerly wind over northern China, and the anomalous westerly winds over the tropical Arabian Sea, southern India, and the South China Sea. Significant negative correlation coefficients of −0.52 and −0.39 between the APO and WPSH intensity indices are predicted by the CNRM and UKMO, respectively. The MME (Fig. 8e) with a correlation coefficient of −0.52 also successfully reproduces the above-mentioned anomalous circulation. Meanwhile, compared with the observations, the models underestimate the anomalous anticyclone, which is also observed in 500-hPa winds (figures not shown), over the high latitudes of the North Pacific to some degree (shown in Fig. 8).

5. Summary and discussion

The summer Asian–Pacific Oscillation (APO) is a major mode of climate variations over the subtropics, especially in the Asian–Pacific sector, and is also a good index to measure large-scale atmospheric circulation anomalies (Zhao et al. 2007, 2012). In this study, we assess the predictability of the APO interannual variability in three DEMETER GCMs using the monthly ERA-40 data as verification. The result shows that all three DEMETER GCMs predict the summer APO index interannual variability during 1959–2001 reasonably well. In particular, three GCMs display a higher skill for the interannual variability of the North Pacific than for the Asian upper-tropospheric temperature index. The multimodel ensemble, as compared to the individual model, improves the prediction of the interannual variability of summer APO index and the Asian upper-tropospheric temperature index; in particular, the improvement is more visible in the latter prediction. We also note that these models do not successfully predict the observed long-term trends of the APO index or the upper-tropospheric temperature over Asia and the North Pacific during this period, suggesting a lower predictive ability for the long-term trends.

For the APO-related circulation anomalies, the GCMs capture well their spatial patterns, although there are some deficiencies in the amplitude. In particular, the GCMs successfully predict the subtropical cyclonic anomaly in the western North Pacific and an anticyclonic anomaly to the northeast in the lower troposphere associated with a high APO index, which indicates a weakened and northward WPSH. Moreover, the GCMs predict well the anomalous easterly winds over Eurasia and the North Pacific tropics in the upper troposphere and anomalous westerly or southerly winds over South Asia and East Asia in the lower troposphere with a strengthened vertical shear of the zonal wind over the South Asian monsoon region. These predicted anomalous features also indicate a stronger Asian summer monsoon when the APO index is higher, which is consistent with the observations.

The above analysis has shown that the DEMETER GCMs better predict the upper-tropospheric temperature variability over the North Pacific than over the Asian land. In fact, this phenomenon is also observed in the surface temperature. Figure 9 shows the correlation coefficients between the predicted and observed summer-mean land (ocean) surface temperatures. In general, significant positive correlations mainly appear in the ocean rather than on the land, although large-scale significant correlations also cover many areas of Eurasian land in the ECMWF model and the multimodel-ensemble results (Figs. 9a,d). In the upper troposphere (Fig. 10), large-scale significant positive correlations also appear in the North Pacific while there is almost no predictability over the Eurasian land. This lack of predictability over land is possibly associated with complicated land–atmosphere interactions over Asia and the drawbacks of the models in describing these interactions. Moreover, the atmospheric aerosol concentrations over Asia showed a remarkable increasing trend since the 1950s, particularly a great increase of atmospheric black carbon and sulfate aerosol concentrations since 1980 (Tegen et al. 2000); the increased aerosol concentrations could reduce tropospheric temperature (Wigley 1989; Li 2000). The DEMETER GCMs in this study use a constant aerosol concentration forcing fixed to the observed 1990 value, possibly leading to a low-latitude Asian tropospheric temperature underestimation. This also gives an explanation for the DEMTER GCMs poorly predicting the Asian tropospheric temperature during 1959–2001. Meanwhile, previous studies have documented that the winter/spring snow cover (depth) is one of the most important factors of surface thermal conditions over the Tibetan Plateau and can significantly affect the Asian summer climate (Blanford 1884; Hahn and Shukla 1976; Robock and Mu 2003; Zhang et al. 2004). The spring snow depth over the Tibetan Plateau further displays a distinct decadal increase after the late 1970s (Zhang et al. 2004). However, the DEMETER GCMs to predict the interannual variability and the decadal change of the winter/spring snow depth over the Tibetan Plateau remain at a low level (figures not shown), which gives another explanation for the lower skill in predicting the Asian upper temperature.

Fig. 9.

The correlation coefficient between the predicted and observed detrended JJA land (sea) surface temperature during 1959–2001 for the (a) ECMWF model, (b) CNRM model, (c) UKMO model, and (d) the MME. Shading is significant at the 95% confidence level. The contours have been smoothed.

Fig. 9.

The correlation coefficient between the predicted and observed detrended JJA land (sea) surface temperature during 1959–2001 for the (a) ECMWF model, (b) CNRM model, (c) UKMO model, and (d) the MME. Shading is significant at the 95% confidence level. The contours have been smoothed.

Fig. 10.

As in Fig. 9, but for the JJA upper-tropospheric (500–200 hPa) T′.

Fig. 10.

As in Fig. 9, but for the JJA upper-tropospheric (500–200 hPa) T′.

The Asian monsoon climate is closely associated with the tropospheric temperature contrasts between Asia and its adjacent areas. Although the predictability is lower over Asia, the interannual variability of the thermal contrast indicated by the APO index and the associated atmospheric circulation anomalies can be reasonably well predicted by the DEMTER GCMs. When the summer APO index is added as a new predictor to the present climate forecast system, to what extent is the Asian summer monsoon precipitation prediction improved? In addition, the new multimodel ensembles for the seasonal-to-annual predictions project, Ensemble-Based Predictions of Climate Changes and their Impacts (ENSEMBLES), show progress relative to DEMETER, including the reduction of the systematic errors, the improvement of the probabilistic forecast skill scores of tropical SSTs (Weisheimer et al. 2009), and the enhancement for the prediction of anomalous surface temperature conditions (Alessandri et al. 2011). However, can the skill of APO be improved in the ENSEMBLES hindcasts? Furthermore, considering the circulation anomalies associated with the APO, using the National Centers for Environmental Prediction Climate Forecast System (Saha et al. 2006) to produce the real-time APO prediction would be meaningful. These questions and works will be addressed in the future.

Acknowledgments

This research was supported by the Major State Basic Research Development Program of China (973 Program, Grant 2009CB421406) and the National Natural Science Foundation of China (Grant 41130103). We are grateful to the anonymous reviewers whose comments greatly improved the quality of the manuscript.

REFERENCES

REFERENCES
Alessandri
,
A.
,
A.
Borrelli
,
A.
Navarra
,
A.
Arribas
,
M.
Déqué
,
P.
Rogel
, and
A.
Weisheimer
,
2011
:
Evaluation of probabilistic quality and value of the ENSEMBLES multimodel seasonal forecasts: Comparison with DEMETER
.
Mon. Wea. Rev.
,
139
,
581
607
.
Blanford
,
H. F.
,
1884
:
On the externsion of the Himalaya snowfall with dry winds and seasons of drought in India
.
Proc. Roy. Soc. London
,
37
,
1
23
.
Cantelaube
,
P.
, and
J. M.
Terres
,
2005
:
Seasonal weather forecasts for crop yield modelling in Europe
.
Tellus
,
57A
,
476
487
.
Challinor
,
A. J.
,
J. M.
Slingo
,
T. R.
Wheeler
, and
F. J.
Doblas-Reyes
,
2005
:
Probabilistic simulations of crop yield over western India using the DEMETER seasonal hindcast ensembles
.
Tellus
,
57A
,
498
512
.
Chen
,
G.
,
1999
: The subtropical high. The Droughts and Floods in Summer in China and Background Fields (in Chinese), Z. G. Zhao, Ed., China Meteorological Press, 45–52.
Chen
,
H. P.
,
J. Q.
Sun
, and
H. J.
Wang
,
2012
:
A statistical downscaling model for forecasting summer rainfall in China from DEMETER hindcast datasets
.
Wea. Forecasting
,
27
,
608
628
.
Diez
,
E.
,
C.
Primo
,
J. A.
García-Moya
,
J. M.
Gutiérrez
, and
B.
Orfila
,
2005
:
Statistical and dynamical downscaling of precipitation over Spain from DEMETER seasonal forecasts
.
Tellus
,
57A
,
409
423
.
Doblas-Reyes
,
F. J.
,
R.
Hagedorn
, and
T. N.
Palmer
,
2005
:
The rationale behind the success of multi-model ensembles in seasonal forecasting–II. Calibration and combination
.
Tellus
,
57A
,
234
252
.
Fan
,
K.
, and
H. J.
Wang
,
2009
:
A new approach to forecasting typhoon frequency over the western North Pacific
.
Wea. Forecasting
,
24
,
974
986
.
Fan
,
K.
,
H. J.
Wang
, and
Y. J.
Choi
,
2008
:
A physically-based statistical forecast model for the middle-lower reaches of the Yangtze River Valley summer rainfall
.
Chin. Sci. Bull.
,
53
,
602
609
.
Fan
,
K.
,
Y.
Liu
, and
H. P.
Chen
,
2012
:
Improving the prediction of the East Asian summer monsoon: New approaches
.
Wea. Forecasting
,
27
,
1017
1030
.
Fil
,
C.
, and
L.
Dubus
,
2005
:
Winter climate regimes over the North Atlantic and European region in ERA-40 reanalysis and DEMETER seasonal hindcasts
.
Tellus
,
57A
,
290
307
.
Hagedorn
,
R.
,
F. J.
Doblas-Reyes
, and
T. N.
Palmer
,
2005
:
The rationale behind the success of multi-model ensembles in seasonal forecasting.–I. Basic concept
.
Tellus
,
57A
,
219
233
.
Hahn
,
D. G.
, and
J.
Shukla
,
1976
:
An apparent relationship between Eurasian snow cover and Indian monsoon rainfall
.
J. Atmos. Sci.
,
33
,
2461
2462
.
Hoshen
,
M.
, and
A.
Morse
,
2004
:
A weather-driven model of malaria transmission
.
Malar. J.
,
3
,
doi:10.1186/1475-2875-3-32
.
Jin
,
E. K.
, and
Coauthors
,
2008
:
Current status of ENSO prediction skill in coupled ocean–atmosphere models
.
Climate Dyn.
,
31
,
647
664
.
Johansson
,
Å.
,
2007
:
Prediction skill of the NAO and PNA from daily to seasonal time scales
.
J. Climate
,
20
,
1957
1975
.
Joseph
,
S.
,
A. K.
Sahai
, and
B. N.
Goswami
,
2010
:
Boreal summer intraseasonal oscillations and seasonal Indian monsoon prediction in DEMETER coupled models
.
Climate Dyn.
,
35
,
651
667
.
Kumar
,
K. K.
,
M.
Hoerling
, and
B.
Rajagopalan
,
2005
:
Advancing dynamical prediction of Indian monsoon rainfall
.
Geophys. Res. Lett.
,
32
,
L08704
,
doi:10.1029/2004GL021979
.
Li
,
C. Y.
,
2000
: Introduction to Climate Dynamics (in Chinese). China Meteorological Press, 32 pp.
Li
,
F.
, and
H. J.
Wang
,
2012
:
Predictability of the East Asian winter monsoon interannual variability as indicated by the DEMETER CGCMS
.
Adv. Atmos. Sci.
,
29
,
441
454
.
Liu
,
Y.
, and
K.
Fan
,
2012
:
Improve the prediction of summer precipitation in the Southeastern China by a hybrid statistical downscaling model
.
Meteor. Atmos. Phys.
,
117
,
121
134
,
doi:10.1007/s00703-012-0201-0
.
Liu
,
Y.
, and
K.
Fan
,
2013
:
A new statistical downscaling model for autumn precipitation in China
.
Int. J. Climatol.
,
33
,
1321
1336
,
doi:10.1002/joc.3514
.
Ma
,
S.
,
X.
Rodó
, and
F. J.
Doblas-Reyes
,
2012
:
Evaluation of the DEMETER performance for seasonal hindcasts of the Indian summer monsoon rainfall
.
Int. J. Climatol.
,
32
,
1717
1729
,
doi:10.1002/joc.2389
.
Man
,
W.
, and
T.
Zhou
,
2011
:
Forced response of atmospheric oscillations during the last millennium simulated by a climate system model
.
Chin. Sci. Bull.
,
56
,
3042
3052
.
Marletto
,
V.
, and
Coauthors
,
2005
:
Evaluation of downscaled DEMETER multi-model ensemble seasonal hindcasts in a northern Italy location by means of a model of wheat growth and soil water balance
.
Tellus
,
57A
,
488
497
.
Nan
,
S. L.
,
P.
Zhao
,
S.
Yang
, and
J. M.
Chen
,
2009
:
Springtime tropospheric temperature over the Tibetan Plateau and evolutions of the tropical Pacific SST
.
J. Geophys. Res.
,
114
,
D10104
,
doi:10.1029/2008jd011559
.
Nitta
,
T.
,
1986
:
Long-term variations of cloud amount in the western Pacific region
.
J. Meteor. Soc. Japan
,
64
,
373
390
.
Nitta
,
T.
,
1987
:
Convective activities in the tropical western Pacific and their impact on the Northern Hemisphere summer circulation
.
J. Meteor. Soc. Japan
,
65
,
373
390
.
Palmer
,
T. N.
, and
Coauthors
,
2004
:
Development of a European Multimodel Ensemble System for Seasonal-to-Interannual Prediction (DEMETER)
.
Bull. Amer. Meteor. Soc.
,
85
,
853
872
.
Qian
,
Z. L.
,
H. J.
Wang
, and
J. Q.
Sun
,
2011
:
The hindcast of winter and spring Arctic and Antarctic oscillation with the coupled climate models
.
Acta Meteor. Sin.
,
25
,
340
354
.
Robock
,
A.
,
M.
Mu
,
K.
Vinnikov
, and
D.
Robinson
,
2003
:
Land surface conditions over Eurasia and Indian summer monsoon rainfall
.
J. Geophys. Res.
,
108
,
4131
,
doi:10.1029/2002jd002286
.
Rogers
,
J. C.
,
1981
:
The North Pacific Oscillation
.
J. Climatol.
,
1
,
39
57
.
Saha
,
S.
, and
Coauthors
,
2006
:
The NCEP Climate Forecast System
.
J. Climate
,
19
,
3483
3517
.
Sun
,
J. Q.
, and
H. P.
Chen
,
2011
:
Predictability of western North Pacific typhoon activity and its factors using DEMETER coupled models
.
Chin. Sci. Bull.
,
56
,
3474
3479
.
Tegen
,
I.
,
D.
Koch
,
A. A.
Lacis
, and
M.
Sato
,
2000
:
Trends in tropospheric aerosol loads and corresponding impact on direct radiative forcing between 1950 and 1990: A model study
.
J. Geophys. Res.
,
105
(
D22
),
26 971
26 989
.
Thomson
,
M. C.
,
F. J.
Doblas-Reyes
,
S. J.
Mason
,
R.
Hagedorn
,
S. J.
Connor
,
T.
Phindela
,
A. P.
Morse
, and
T. N.
Palmer
,
2006
:
Malaria early warnings based on seasonal climate forecasts from multi-model ensembles
.
Nature
,
439
,
576
579
.
Uppala
,
S. M.
, and
Coauthors
,
2005
:
The ERA-40 Re-Analysis
.
Quart. J. Roy. Meteor. Soc.
,
131
,
2961
3012
,
doi:10.1256/qj.04.176
.
Vitart
,
F.
,
2006
:
Seasonal forecasting of tropical storm frequency using a multi-model ensemble
.
Quart. J. Roy. Meteor. Soc.
,
132
,
647
666
.
Walker
,
G. T.
, and
E. W.
Bliss
,
1932
:
World weather V
.
Mem. Roy. Meteor. Sci.
,
4
,
53
84
.
Wallace
,
J. M.
, and
D. S.
Gutzler
,
1981
:
Teleconnections in the geopotential height field during the Northern Hemisphere winter
.
Mon. Wea. Rev.
,
109
,
784
812
.
Wang
,
H. J.
,
2001
:
The weakening of the Asian monsoon circulation after the end of 1970s
.
Adv. Atmos. Sci.
,
18
,
376
386
.
Wang
,
H. J.
, and
K.
Fan
,
2009
:
A new scheme for improving the seasonal prediction of summer precipitation anomalies
.
Wea. Forecasting
,
24
,
548
554
.
Weisheimer
,
A.
, and
Coauthors
,
2009
:
ENSEMBLES: A new multi-model ensemble for seasonal-to-annual predictions—Skill and progress beyond DEMETER in forecasting tropical Pacific SSTs
.
Geophys. Res. Lett.
,
36
,
L21711
,
doi:10.1029/2009GL040896
.
Wigley
,
T. M. L.
,
1989
:
Possible climate change due to SO2-derived cloud condensation nuclei
.
Nature
,
339
,
365
367
.
Xie
,
P.
, and
P. A.
Arkin
,
1997
:
Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs
.
Bull. Amer. Meteor. Soc.
,
78
,
2539
2558
.
Yoo
,
J. H.
, and
I.-S.
Kang
,
2005
:
Theoretical examination of a multi-model composite for seasonal prediction
.
Geophys. Res. Lett.
,
32
,
L18707
,
doi:10.1029/2005GL023513
.
Zhang
,
Y.
,
T.
Li
, and
B.
Wang
,
2004
:
Decadal change of the spring snow depth over the Tibetan Plateau: The associated circulation and influence on the East Asian summer monsoon
.
J. Climate
,
17
,
2780
2793
.
Zhao
,
P.
,
Y. N.
Zhu
, and
R. H.
Zhang
,
2007
:
An Asian-Pacific teleconnection in summer tropospheric temperature and associated Asian climate variability
.
Climate Dyn.
,
29
,
293
303
.
Zhao
,
P.
,
Z. H.
Cao
, and
J. M.
Chen
,
2010
:
A summer teleconnection pattern over the extratropical Northern Hemisphere and associated mechanisms
.
Climate Dyn.
,
35
,
523
534
.
Zhao
,
P.
,
S.
Yang
,
H. J.
Wang
, and
Q.
Zhang
,
2011
:
Interdecadal relationships between the Asian–Pacific Oscillation and summer climate anomalies over Asia, North Pacific, and North America during a recent 100 years
.
J. Climate
,
24
,
4793
4799
.
Zhao
,
P.
,
B.
Wang
, and
X.
Zhou
,
2012
:
Boreal summer continental monsoon rainfall and hydroclimate anomalies associated with the Asian-Pacific Oscillation
.
Climate Dyn.
,
39
,
1197
1207
,
doi:10.1007/s00382-012-1348-6
.