Abstract

The influence of the El Niño–Southern Oscillation (ENSO) and Pacific decadal oscillation (PDO) interference on the dry and wet conditions in the Great Plains of the United States has been examined using monthly observational datasets. It is shown that both ENSO and PDO can generate a similar pattern of atmospheric and oceanic anomalies over the eastern part of the North Pacific and western North America that has significant impact on the climate over the Great Plains. Furthermore, the relationship between ENSO–PDO and climate anomalies in the Great Plains is intensified when ENSO and PDO are in phase (El Niño and warm PDO or La Niña and cold PDO). On average, anomalies over the Great Plains favor wet (dry) conditions when both ENSO and PDO are in the positive (negative) phase. However, when ENSO and PDO are out of phase, the relationship is weakened and the anomalies over the Great Plains tend toward neutral. Without ENSO, PDO alone does not affect the North American climate significantly. The relationship is quite robust for different seasons, with the strongest effects for the months of spring and the weakest effects for the months of autumn, whereas the months of winter and summer fall in between. The seasonality of the relationship may be associated with the seasonal dependence of the anomalies of general circulation and the pattern of mean seasonal cycle in the North Pacific.

The contrasting impact of the interference of ENSO and PDO on the climate anomalies in the Great Plains is associated with differences in the ocean–atmosphere anomalies. When ENSO and PDO are in phase, the sea surface temperature (SST) anomalies extend from the equatorial Pacific to the higher latitudes of the North Pacific via the eastern ocean. The distribution of the corresponding anomalies of sea level pressure (SLP) and the wind at 1000 hPa form an ellipse with a southeast–northwest orientation of the long axis because the SST anomalies promote coherent changes in SLP in the central North Pacific. In the upper troposphere, a similar teleconnection pattern with the same sign generated by ENSO and PDO is overlapped and enhanced, which favors anomaly (dry and wet) conditions in the Great Plains. However, when ENSO and PDO are out of phase, the SST anomalies have the same sign in the tropical and central North Pacific, which is opposite to the anomalies near the west coast of North America. The anomalously cyclonic circulation over the North Pacific is weaker in the out-of-phase situation than in the in-phase situation. The distribution of the anomalies of SLP and the wind at 1000 hPa resembles a circle. Meanwhile, in the upper troposphere, ENSO and PDO generate a similar teleconnection pattern with opposite sign, causing cancellation of the anomalous circulation and favoring neutral climate in the Great Plains.

1. Introduction

Persistent dry or wet climate, particularly severe and long drought periods, such as those occurring during the 1930s, 1950s, and 1999–2003, significantly affects the economy and society of the United States. Therefore, it is important to understand the processes that cause these abnormal conditions and to enhance our capability to forecast their occurrence. Previous studies have shown that dry and wet events in the United States are directly associated with large-scale atmospheric general circulation anomalies in the midlatitudes of the Northern Hemisphere (NH). For example, Namias (1955) linked summer droughts in 1952–54 over the United States with anomalies of quasi-stationary planetary waves over the NH. Indeed, the heat waves and droughts over the central Great Plains are mainly associated with a pattern of zonal teleconnection extending from the central Pacific to the North Atlantic (Chang and Wallace 1987). These circulation anomalies generate anomalous moisture transport into the North American continent and cause convergence or divergence, leading to dry or wet spells over specific geographic areas (Mo et al. 1997). Conceptually, these circulation anomalies can be further decomposed into internal variability and externally forced components. The former originates from dynamical evolution within the atmosphere, which is usually chaotic in nature and becomes unpredictable beyond a month (Shukla 1981). The latter, however, is usually the forced atmospheric response to anomalous conditions in the ocean, which are more predictable on seasonal and interannual time scales. For example, using model simulations, Hoerling and Kumar (2003) found that the 1998–2003 droughts over the United States, southern Europe, and Southwest Asia were linked with remarkably persistent cold sea surface temperature (SST) anomalies in the eastern tropical Pacific and warm SSTs in the western tropical Pacific and Indian Oceans during this period. Schubert et al. (2004a,b) demonstrated that the low-frequency (time scales longer than 6 yr) rainfall variations in the Great Plains in an ensemble mean of general circulation model simulations are forced by a pan-Pacific pattern of SST variability.

Among these SST anomalies associated with climate variability in the United States, the most coherent signals are in the tropical eastern Pacific, connected with the El Niño–Southern Oscillation (ENSO). The connection of climate variability in the United States with ENSO has been recognized for decades (Ropelewski and Halpert 1986; Sittel 1994; Green et al. 1997; Hoerling and Kumar 2003; Schubert et al. 2004a; Yang et al. 2007; Wang et al. 2007). Ropelewski and Halpert (1986) investigated the typical patterns of North American precipitation and temperature anomalies associated with ENSO. They found that above-normal precipitation in the southeastern United States and the Great Basin area of the western United States tend to occur in different phases of an ENSO cycle. Negative temperature anomalies in parts of the southeastern United States near the Gulf of Mexico are connected with the warm phase of ENSO (El Niño). Sittel (1994) and Green et al. (1997) showed the patterns of seasonal surface temperature and precipitation anomalies in the United States averaged for El Niño and La Niña events, respectively. They suggested that the patterns vary with seasons and the impact was not symmetric with respect to warm and cold events. The impacts of El Niño during summer over North America are weaker and more variable compared with the impact of La Niña (Wang et al. 2007). During summers with La Niña, a continental-scale anomalous high dominates over most of North America, leading to hot and dry summers over the central United States. On average, the largest correlation occurs when the Niño-3.4 SST leads the precipitation in the Great Plains by one month. The ENSO–precipitation relationship has a strong seasonal dependence, with the greatest significance in summer (Yang et al. 2007). Beyond the general patterns of monthly or seasonal climate anomalies, there is also evidence that ENSO can affect the extremes of U.S. climate, such as daily extremes of winter surface air temperature (Higgins et al. 2002) and peak wind gust magnitudes (Enloe et al. 2004).

In addition, the impact of ENSO on U.S. climate depends on the phase of ENSO cycle, regions of the United States, seasons, and even climate variables. Moreover, the impacts also experience an interdecadal change (Hu and Feng 2001; McCabe et al. 2004; Larkin and Harrison 2005). This probably reflects the fact that the ENSO impact on the North American climate is modulated by other lower-frequency signals (McCabe et al. 2004), such as decadal variability in the Pacific. For example, Gershunov and Barnett (1998) found that typical ENSO signals in the United States tend to be stronger and more stable during preferred phases of the North Pacific Oscillation (NPO). Typical El Niño response patterns around the North American continent (e.g., low pressure over the northeastern Pacific, dry northwest, wet southwest, etc.) are strong and coherent only during the high phase of NPO, which corresponds to an anomalous cold northwestern Pacific and warm eastern North Pacific. The generally reversed sea level pressure (SLP) and precipitation patterns are dominant during La Niña winters only when the low NPO phases prevail. However, climatic anomalies tend to be weak and spatially incoherent in winters when the combinations of low NPO–El Niño and high NPO–La Niña occur.

Actually, the impact of different combinations of ENSO and NPO or the Pacific decadal oscillation (PDO) on U.S. climate varies with region, season, and even the air–sea coupling in the tropical Pacific. For example, McCabe and Dettinger (1999) argued that in the western United States, different precipitation anomalies in the warm season were observed in different combinations of the air–sea coupling in the tropical Pacific and PDO. They found that ENSO teleconnection with precipitation is strong when the Southern Oscillation index (SOI) and Niño-3 index are out of phase and PDO is negative and it is weak when SOI and Niño-3 index are weakly correlated and PDO is positive. Brown and Comrie (2004) suggested that spatial inconsistencies in the relationship between ENSO and winter precipitation across the western United States are due to PDO phase shifts. In Idaho, dry and wet conditions are most likely associated with the combination of La Niña–cold PDO and El Niño–warm PDO, whereas rainfall anomalies are largest during La Niña years with cold PDO conditions (Harshburger et al. 2002). In the Upper Colorado River, PDO also modulates the relationship between ENSO and basin hydroclimatic variations (Hidalgo and Dracup 2003). In Arizona, when the neutral ENSO years are split by cold and warm PDO, the resulting winter precipitation patterns in Arizona are spatially similar to those that occur during years of La Niña–cold PDO and—to a lesser extent—years of El Niño–warm PDO (Goodrich 2007). In the U.S. Southwest, before 1977, negative winter precipitation anomalies are strongly tied to La Niña years, but El Niño years do not systematically lead to positive precipitation anomalies; after 1977, this asymmetry is reversed and positive precipitation anomalies predictably follow El Niño years, but La Niña years yield no precipitation predictability (Gutzler et al. 2002). In the central United States, the teleconnection between ENSO and the regional rainfall is best developed during the early part of the warm season in association with El Niño events, but modest spatial shifts in the teleconnection occur depending upon phase of PDO (Englehart and Douglas 2002). However, there is little evidence for a teleconnection between La Niña events and the regional rainfall. Furthermore, Englehart and Douglas (2003) indicated that warm season droughts in the central United States have no simple or strong teleconnection with ENSO but have systematic links with PDO.

From this review, it is suggested that U.S. climate variability is affected by multiple factors, including ENSO and PDO. The influence of ENSO and PDO on U.S. climate depends on the combination of these two factors and varies with region, season, and variable. Including only one factor and ignoring the other factor or factors in a climate forecast could be a major source of inaccuracy. Studying the combination and interference of these different factors may improve climate prediction. For example, McCabe et al. (2004) demonstrated that U.S. droughts on multidecadal time scales are affected by PDO, the Atlantic multidecadal oscillation, and the NH temperature trend. Inclusion of all these three factors in multivariate regression equations is crucial for accurately simulating the historical 20-yr patterns of drought frequency in the US. One of the tasks of observational climate analyses is to find such statistically significant relationships from historical data and to establish their physical links.

Although it has been recognized that the impact of ENSO on U.S. climate variability is modulated by PDO background, it will be interesting to examine the impact of various combinations of monthly mean ENSO and PDO on dry and wet conditions in the Great Plains, the seasonality of the impact, and any possible physical mechanism. It is expected that a better prediction of the dry and wet conditions over the Great Plains might be achieved by examining this problem. In this work, the dry and wet months in the Great Plains are first identified and the associated atmosphere circulation and SST anomalies are analyzed. Then the impact of ENSO, PDO, and their interference on the dry and wet conditions is examined through composites of atmosphere circulation and SST anomalies based on ENSO, PDO, and various combinations of ENSO and PDO. The paper is organized as follows: in section 2, the data and the analysis strategy are briefly described. Section 3 defines dry and wet conditions in the Great Plains and shows the associated large-scale anomalies. The role of ENSO and PDO interference on dry and wet conditions, as well as its seasonality, is investigated in section 4. Section 5 provides a summary and further discussion.

2. Data and analysis strategy

Monthly global analyses of precipitation over land on a 2.5° × 2.5° grid, referred to as the precipitation reconstruction (PREC; Chen et al. 2002), are used in this study. The grid data are defined by optimal interpolation of gauge observations over land, which are collected from gauge observations over 17 000 stations by the Global Historical Climatology Network, version 2, and the Climate Anomaly Monitoring System datasets. Chen et al. (2002) demonstrated that the mean distribution and annual cycle of PREC showed good agreement with those in several other published gauge-based datasets. They also found that the anomalous patterns associated with ENSO from this dataset resemble those identified in previous studies.

Monthly global soil wetness (SW) dataset on a 0.5° × 0.5° grid is used to define wet, dry, and neutral conditions in the Great Plains. The soil wetness data are derived from a land model (Fan and Van den Dool 2004) of a one-layer bucket water balance (Huang et al. 1996) with the driving input fields of monthly global PREC precipitation and temperature from global reanalysis. Huang et al. (1996) showed that, under such boundary conditions, the model-calculated soil wetness fits well with the observed soil wetness in the top 1.3 m of soil. Moreover, the annual cycle and interannual variability of the simulated soil wetness are reasonably close to the limited observations in different regions (Fan and Van den Dool 2004).

In addition to the precipitation and soil wetness data, other analyzed/reanalyzed atmosphere variables and SST data are also used. The monthly mean SLP and wind at 1000 and 200 hPa on a 2.5° × 2.5° grid were derived from the reanalysis of the National Centers for Environmental Prediction and National Center for Atmospheric Research (Kalnay et al. 1996). Version 2 of the reconstructed SST data on a 2° × 2° grid used in SST composites and Niño-3.4 index calculations are from Reynolds et al. (2002). The Niño-3.4 index is defined as the averaged SST anomalies within 5°S–5°N, 170°–120°W.

To analyze the connection between dry and wet conditions and SST variability in the North Pacific, a PDO index is also adopted. The PDO index is defined as the leading principal component (PC) of North Pacific monthly SST variability (poleward of 20°N; Zhang et al. 1997; Mantua et al. 1997). The global averages of the monthly mean SST anomalies are removed prior to the PC decomposition to separate PDO-related variability from any global warming signals that may be present in the SST data. The PDO index represents a long-lived El Niño–like pattern of the SST anomalies in the tropics with broader meridional scale, which is linked to a North Pacific SST pattern with opposite SST anomalies between the central ocean and eastern boundary (see Figs. 1, 2 of Mantua et al. 1997). The standardized values of the PDO index are calculated by Nathan Mantua and downloaded from the Joint Institute for the Study of the Atmosphere and Ocean Web site (available online at http://jisao.washington.edu/pdo/PDO.latest).

All the previously mentioned data that are used in the analyses are monthly means, which span the 56-yr period of January 1950–December 2005. Our analyses are directly applied to the data in all months, except in section 4c, where monthly data in different seasons are examined separately. In particular, we have used the original monthly PDO index as the basis of composite analyses without presuming that the phenomenon it characterizes is on decadal and interdecadal time scales. Using a monthly index in this work is different from most of the previous work, in which those investigations were conducted based on long-term epochs of PDO (e.g., Gershunov and Barnett 1998; McCabe and Dettinger 1999; Gutzler et al. 2002; Englehart and Douglas 2002, 2003; Brown and Comrie 2004; Goodrich 2007). Actually, besides the interdecadal time scale variation, PDO also has large variability at interannual and interseasonal time scales (see next section). Thus, it is reasonable to use monthly data to examine the influence of ENSO and PDO on the climate variations in the Great Plains.

3. Wet and dry conditions in the Great Plains and large-scale anomalies

a. Wet and dry conditions in the Great Plains

In this work, dry, wet, and neutral conditions in the Great Plains are defined based on the soil wetness. Figure 1a shows the time series of monthly anomalies of precipitation and soil wetness averaged in a box that is roughly coincident with the Great Plains (30°–50°N, 95°–105°W). The regressions of the indices onto their corresponding gridpoint values (Figs. 1b,c) show that the time series of the monthly anomalies of precipitation and soil wetness averaged in the Great Plains well represents the anomalies in the central part of the conterminous United States. Compared with the result of Diaz (1983), it is found that regions with high regression values (Figs. 1b,c) are collocated with those of high frequency of dry and wet spells (see Fig. 4 of Diaz 1983).

Fig. 1.

(a) Time series of monthly SW (%, shading) and precipitation (mm day−1, curve) anomalies averaged over the Great Plains (30°–50°N, 95°–105°W) for January 1950–December 2005. (b),(c) Regression of (b) SW and (c) precipitation onto the normalized corresponding time series in (a). The contour interval (CI) is 2% in (b) and 0.2 mm day−1 in (c). The Great Plains region for (a) is shown by the rectangles in (b),(c).

Fig. 1.

(a) Time series of monthly SW (%, shading) and precipitation (mm day−1, curve) anomalies averaged over the Great Plains (30°–50°N, 95°–105°W) for January 1950–December 2005. (b),(c) Regression of (b) SW and (c) precipitation onto the normalized corresponding time series in (a). The contour interval (CI) is 2% in (b) and 0.2 mm day−1 in (c). The Great Plains region for (a) is shown by the rectangles in (b),(c).

Nevertheless, the definition of dry and wet conditions at a location can still be an issue of dispute. Normally, dry or wet episodes can be characterized by anomalies of precipitation. However, monthly rainfall amount alone may not fully describe a flood or drought condition. The precipitation anomalies are related to the temperature anomalies, particularly in the midlatitudes of NH summer. For instance, in the central Great Plains, little rain falls during a heat wave, so summer droughts can also be characterized by anomalously high surface air temperature (Chang and Wallace 1987). Soil wetness (moisture) is a variable combining the information of both precipitation and temperature anomalies to some extent. Moreover, given the much higher persistence of soil wetness anomalies compared with that of precipitation anomalies (not shown), soil wetness represents the integrated effects of the precipitation and surface air temperature, which smooths out high-frequency atmospheric fluctuations and also makes it less seasonally dependent (see section 4c). The soil wetness (moisture) has been used in previous works to define dry and wet anomalies. For example, Chang and Wallace (1987) used soil moisture anomalies to define the droughts over the central Great Plains. Andreadis et al. (2005) used soil moisture and runoff to construct a drought history in the conterminous United States from 1920 to 2003. This suggests that soil wetness is a good factor for defining dry and wet anomalies. In this work, a wet (dry) month in the Great Plains is defined when the averaged soil wetness anomalies ≥0.5 (≤−0.5) standard deviation (STDV). According to this criterion, about 33%, 33%, and 34% of the total months during 1950–2005 are cataloged as dry, wet, and neutral states in the Great Plains, respectively.

Actually, the precipitation and soil wetness indices of the Great Plains shown in Fig. 1a are significantly correlated. Most of the correlation seems to come from their coherence at lower frequencies of the two time series. The simultaneous correlation coefficient is 0.36, and the maximal correlation coefficient is 0.54 when the precipitation anomalies lead the soil wetness anomalies by one month (not shown). Moreover, all the correlations are significant at the 99% significance level of the t test when the precipitation anomalies lead the soil wetness anomalies by 1–12 months, indicating the persistent effect of precipitation on soil wetness in the following months. However, when the soil wetness anomalies lead the precipitation anomalies, no correlations are significant at the 99% level (not shown).

b. Associated with large-scale anomalies

The dry and wet spells in the Great Plains are associated with large-scale atmosphere circulation and SST anomalies. Figure 2 shows the simultaneous composite of SST and SLP anomalies based on the dry and wet months in all months, as well as the differences of the composite between the means of the wet and dry months. For the SST (Fig. 2, top), the most pronounced anomalies are in the tropical central and eastern Pacific centered on the equator and extending poleward along the eastern boundary, which are mostly linked to ENSO. Wet (dry) conditions in the Great Plains are associated with positive (negative) SST anomalies in the tropical central and eastern Pacific. This is consistent with previous results, as reviewed in section 1.

Fig. 2.

Composites of (top) SST and (bottom) SLP based on the (a) wet (averaged SW ≥ 0.5 STDV) and (b) dry (averaged SW ≤ −0.5 STDV) months in the Great Plains and (c) wet − dry [(a) − (b)]. The (CI) is 0.1°C for SST and 0.2 hPa for SLP, and the zero line is omitted.

Fig. 2.

Composites of (top) SST and (bottom) SLP based on the (a) wet (averaged SW ≥ 0.5 STDV) and (b) dry (averaged SW ≤ −0.5 STDV) months in the Great Plains and (c) wet − dry [(a) − (b)]. The (CI) is 0.1°C for SST and 0.2 hPa for SLP, and the zero line is omitted.

Another region with clear and opposite SST anomalies between the wet and dry composite is the North Pacific. For the wet composite, negative SST anomalies are seen in the central North Pacific and positive ones are seen in the region from the southeastern extended to the northern part of the North Pacific near the coasts. The SST anomaly pattern in the North Pacific for the dry composite is similar to that for the wet composite with opposite sign and more prominent signals near the coast. The SST anomalies in the North Pacific imply a possible connection between dry and wet conditions in the Great Plains and PDO. Basically, wet (dry) conditions in the Great Plains are favorable when the SST anomalies are below (above) normal in the central North Pacific and above (below) normal in the region from the southeastern extended to northern part of the North Pacific. This is generally consistent with some previous works, such as Mantua et al. (1997), Gershunov and Barnett (1998), Barlow et al. (2001), Englehart and Douglas (2002), and McCabe et al. (2004).

The SLP anomalies (Fig. 2, bottom) associated with wet and dry conditions in the Great Plains show a consistent pattern with the corresponding SST anomalies (Fig. 2, top). In the tropics, a Southern Oscillation–like anomaly pattern is seen. For the wet (dry) composite, the SLP anomalies favor the negative (positive) phase of SO. In the middle and high latitudes, major anomalies over the oceans are connected with the SST anomalies in the North Pacific. During the wet episode, an area of negative anomaly centered near 45°N and east of the date line extends southeastward to the coast of California (Fig. 2a, bottom). During the dry episode, the pattern of SLP anomalies is largely reversed, except for an eastward shift of the SLP center (Fig. 2b, bottom). Interestingly, there seems to be a certain symmetry of the midlatitude SLP with respect to the equator. As a result, an SLP center also emerges over the Southern Ocean. This is consistent with Schubert et al. (2004a,b). They suggested that the ENSO-related global symmetry of atmospheric circulations with respect to the equator may be associated with multiyear droughts. Other pronounced regions of SLP anomalies are in the high latitudes, suggesting that the negative (positive) phase of Arctic Oscillation (AO) and the North Atlantic Oscillation (NAO) is also linked to wet (dry) conditions in the Great Plains, which is consistent with the results of daily surface temperature composite of Higgins et al. (2002). In this work, we focus on the influence of ENSO and PDO interference on the dry and wet conditions in the Great Plains, although admitting that high-latitude anomalies, such as the AO/NAO, may also make potentially significant contributions to the dry and wet conditions.

c. Influence of ENSO and PDO

The composites of SST and SLP based on dry and wet conditions in the Great Plains in the previous subsection suggest a possible connection of ENSO and PDO with dry and wet conditions. As we know, the variability of PDO is mainly at decadal and interdecadal time scales (Figs. 3a,c), which contrasts with the characteristics of ENSO, which is dominated by interannual variations (Figs. 3a,b). Nevertheless, the spectra of PDO at around 1 yr are also significant at the 90% level of significance besides the significance spectra at interdecadal time scales (Fig. 3c). However, it is not significant at the 95% level of significance at both time scales because of the short record of the data. The power spectrum results suggest the multiscale feature of PDO. The power spectra are estimated based on the autocorrelation functions of PDO and ENSO indices. Furthermore, our calculation shows that the variances of the PDO index are 0.30, 0.43, and 0.40 for time scales longer than 10 yr, between 1 and 10 yr, and shorter than 1 yr, respectively. The variance is 0.22 at ENSO time scales (between 2 and 7 yr). Therefore, the short time scale variations of PDO are not an ignorable component, and it may be reasonable to use monthly data, instead of only considering the interdecadal phase of PDO, to examine the influence of PDO and ENSO on the climate variability over the Great Plains.

Fig. 3.

(a) Time series of monthly PDO (shading) and Niño-3.4 (curve) indices for January 1950–December 2005, and the power spectra of the (b) Niño-3.4 and (c) PDO indices. The blue shadings in (b) and (c) are significant at the level (SL) of 90%.

Fig. 3.

(a) Time series of monthly PDO (shading) and Niño-3.4 (curve) indices for January 1950–December 2005, and the power spectra of the (b) Niño-3.4 and (c) PDO indices. The blue shadings in (b) and (c) are significant at the level (SL) of 90%.

Actually, there are significance correlations between the Niño-3.4 and PDO indices (Fig. 3) with a simultaneous correlation coefficient of 0.58 (not shown). The maximal correlation coefficient is 0.61 when the Niño-3.4 index leads the PDO index by 2 months. Also, compared with the PDO index, the Niño-3.4 index shows a shorter persistence (not shown). The positive correlations at lag can last 11 months for the Niño-3.4 index and 14 months for the PDO index (not shown). The close association of ENSO and PDO indicated here is consistent with previous investigations. For example, it has been shown that a substantial amount of PDO can be recovered from a reconstruction of North Pacific SST anomalies based on a first-order autoregressive model and forcing by variability of the Aleutian low, ENSO, and oceanic zonal advection anomalies in the Kuroshio–Oyashio Extension (Schneider and Cornuelle 2005). They further proposed a hypothesis that PDO is not a dynamical mode but arises from the superposition of SST fluctuations with different dynamical origins. Recently, An et al. (2007) suggested that the generation of decadal variability in the North Pacific resulted from the asymmetry of ENSO and the nonlinearity of the atmospheric tropical–extratropical teleconnection.

Nevertheless, the connection of ENSO and PDO with the soil wetness anomalies in the Great Plains also has some difference (Fig. 4). The maximal leading and lagging correlation coefficients are observed when the Niño-3.4 index leads the soil wetness anomalies by 1–2 months (Fig. 4a). This is generally consistent with Harshburger et al. (2002) and Yang et al. (2007), considering the 1-month lag of soil wetness to the precipitation in the Great Plains (not shown). Harshburger et al. (2002) suggested that the teleconnection between Niño-3.4 SSTs and precipitation in the western United States involves a 2–3-month lag time, whereas Yang et al. (2007) noticed the strongest relationship when the Niño-3.4 SST leads the precipitation in the Great Plains by 1 month. However, the situation is different for the impact of PDO. The strongest correlation appears when the PDO index lags the soil wetness anomalies by 1 month (Fig. 4b). These results might suggest that these ocean signals affect climate variations in the United States, probably through atmospheric bridge in a short time interval. The PDO index lagging the soil wetness may imply that PDO is a longer persistent system, compared with ENSO. Comprehensively considering the leading and lagging feature of ENSO and PDO impact on dry and wet conditions in the Great Plains, we examine the contemporary composites of different variables in the following to explore the connection between the dry and wet conditions and ocean–atmosphere anomalies.

Fig. 4.

Leading and lagging corrections of SW anomalies averaged over the Great Plains with (a) Niño-3.4 index and (b) PDO index. The dashed lines are the SL at 99% of the t test.

Fig. 4.

Leading and lagging corrections of SW anomalies averaged over the Great Plains with (a) Niño-3.4 index and (b) PDO index. The dashed lines are the SL at 99% of the t test.

The influence of ENSO and PDO on dry and wet conditions in North America is further demonstrated by the composite of soil wetness based on the Niño-3.4 and PDO indices (Fig. 5, right). To simplify the description, we use El Niño, La Niña, and neutral ENSO to refer to the situations when the Niño-3.4 index is ≥0.5, ≤−0.5, and between −0.5 and 0.5, respectively. When the PDO index is ≥0.5, ≤−0.5, and between −0.5 and 0.5 refers to warm PDO, cold PDO, and neutral PDO, respectively. The basic pattern of the soil wetness anomaly is similar and the differences are small between the composite based on either the Niño-3.4 or PDO indices (Fig. 5). On average, dry (wet) conditions are favorable in the north (45°–60°N) and wet (dry) conditions are mostly seen in the south (20°–40°N), when either the Niño-3.4 or PDO indices are positive (negative). The influence of ENSO and PDO on soil wetness anomalies is most noticeable and robust in the Great Plains, which is another reason to focus this work on the Great Plains. However, the influence of PDO–ENSO is not robust in the west of the Rockies, even for the in-phase situation. This may be partly due to using soil wetness instead of precipitation, because the soil moisture may not be accurately simulated in places such as the desert Southwest.

Fig. 5.

Composites of SLP (shading and contour) and (left) wind at 1000 hPa (vector), (middle) SST, and (right) SW anomalies: (a) mean for El Niño minus mean for La Niña; (b) mean for warm PDO minus mean for cold PDO; and (c) their difference, (a) − (b). The (CI) is 1 hPa for SLP, 0.3°C for SST, and 2% for SW, and the zero line is omitted.

Fig. 5.

Composites of SLP (shading and contour) and (left) wind at 1000 hPa (vector), (middle) SST, and (right) SW anomalies: (a) mean for El Niño minus mean for La Niña; (b) mean for warm PDO minus mean for cold PDO; and (c) their difference, (a) − (b). The (CI) is 1 hPa for SLP, 0.3°C for SST, and 2% for SW, and the zero line is omitted.

The similarity in the soil wetness anomaly composite is consistent with the similarity of the corresponding composites of SLP, wind at 1000 hPa (Fig. 5, left), and SST (Fig. 5, middle) based on the Niño-3.4 and PDO indices (Fig. 5). In the composites of El Niño–La Niña and warm–cold PDO (Figs. 5a,b), anomalously cyclonic circulation is dominant in the North Pacific. The anomalous southwest wind in the south of the anomalously cyclonic circulation favors moisture transport from the ocean to the land and its convergence in the southern part of the conterminous United States. Similarly, north of the anomalously cyclonic circulation, the anomalous northeast wind blocks moisture transport from the ocean to the land and results in divergence and drying in the northern part of North America. The SST anomalies are also similar to the composite based on the Niño-3.4 and PDO indices, which is consistent with the results of Zhang et al. (1997).

However, there are some differences between the composite based on the Niño-3.4 and PDO indices (Fig. 5). As expected, both anomalously cyclonic circulation and negative SST anomalies in the North Pacific are stronger in the composite based on the PDO index than in that based on the Niño-3.4 index. In contrast, the positive SST anomalies in the tropical Pacific are much larger in the composite based on the Niño-3.4 index than in that based on the PDO index. However, it is interesting to note that the contributions from both factors are nearly equal in North America, so that the dry and wet conditions there could be sensitive to the different combination of the two factors, which is shown in the next section.

4. Role of ENSO and PDO interference on the dry and wet conditions

a. Bivariate composites

To investigate the influence of ENSO and PDO interference on the dry and wet conditions in North America, we constructed a composite based on the Niño-3.4 index with different phases of PDO. There are about 33% of total months with in-phase relation and about 10% with out-of-phase relation between the PDO and ENSO anomalous months. Because a total of 56 yr of monthly data were used, the sample is large enough to make the following composites even for the out-of-phase cases. It is noted that positive (negative) Niño-3.4 index is associated with positive (negative) SST anomalies in the tropical Pacific, whereas positive (negative) PDO index is tied to negative (positive) SST anomalies in the central North Pacific. When ENSO and PDO are in phase (Figs. 6a, 7a), anomalies of atmospheric circulation, SST, and soil wetness display a similar spatial pattern with opposite polarity, depending on whether the indices are positive or negative. On average, dry (wet) conditions appear in the north (45°–60°N) and wet (dry) conditions appear in the south (20°–40°N) in the composites of El Niño and warm PDO (La Niña and cold PDO), which is similar to the pattern of composites based solely on either the Niño-3.4 or PDO index (Figs. 5a, 5b). On average, when ENSO and PDO are out of phase (in phase), the soil wetness anomalies in the Great Plains are small (large; Figs. 6, 7). If the ENSO (PDO) effect is absent, PDO (ENSO) alone generates a weaker soil wetness anomaly in North America (Figs. 6c, 7c), compared with the in-phase impact (Figs. 6a, 7a). These results are generally consistent with the previous investigations (see section 1).

Fig. 6.

Composites of SLP (shading and contour) and (left) wind at 1000 hPa (vector), (middle) SST, and (right) SW anomalies: (a) mean for El Niño and warm PDO, (b) mean for La Niña and warm PDO, and (c) mean for neutral ENSO and warm PDO. The CI is 1 hPa for SLP, 0.3°C for SST, and 2% for SW, and the zero line is omitted.

Fig. 6.

Composites of SLP (shading and contour) and (left) wind at 1000 hPa (vector), (middle) SST, and (right) SW anomalies: (a) mean for El Niño and warm PDO, (b) mean for La Niña and warm PDO, and (c) mean for neutral ENSO and warm PDO. The CI is 1 hPa for SLP, 0.3°C for SST, and 2% for SW, and the zero line is omitted.

Fig. 7.

Composites of SLP (shading and contour) and (left) wind at 1000 hPa (vector), (middle) SST, and (right) SW anomalies: (a) mean for La Niña and cold PDO, (b) mean for El Niño and cold PDO, and (c) mean for neutral ENSO and cold PDO. The CI is 1 hPa for SLP, 0.3°C for SST, and 2% for SW, and the zero line is omitted.

Fig. 7.

Composites of SLP (shading and contour) and (left) wind at 1000 hPa (vector), (middle) SST, and (right) SW anomalies: (a) mean for La Niña and cold PDO, (b) mean for El Niño and cold PDO, and (c) mean for neutral ENSO and cold PDO. The CI is 1 hPa for SLP, 0.3°C for SST, and 2% for SW, and the zero line is omitted.

The large-scale ocean–atmospheric conditions dictate this response to a large extent. The key factor seems to be the location of the anomalous SLP center over the North Pacific, which is in turn closely tied to the SST anomalies (Figs. 6a, 7a). For instance, positive SST anomalies around the California coast generated by an El Niño event usually enhance the warm water near the coast in a positive PDO phase (Fig. 6a, middle). This enhanced warm sea surface will reduce local SLP and expand the low SLP center in the North Pacific closer to the coast (Fig. 6a, left). On the other hand, a La Niña event occurring in a positive PDO episode will send cold SST anomalies to the California coast, canceling out the warming associated with PDO (Fig. 6b, middle). As a result, the low SLP center is weakened and pushed westward, making it harder to affect the U.S. climate (Fig. 7b, left). The role of ENSO in a negative PDO episode follows the same general principle (Figs. 7a,b). It is interesting that in the cases of ENSO in neutral conditions, the SLP center in the central Pacific Ocean cannot exert a strong control over the North American continent (Figs. 6c, 7c).

Furthermore, in the upper troposphere, there are clear differences in stationary wave activity between the composite of the in-phase ENSO and PDO and that of the out-of-phase ENSO and PDO. The high-latitude equilibrium response to low-latitude SST anomalies can be explained by Rossby wave propagation forced by convective heating anomalies that develop locally in response to the SST anomalies (Hoskins and Karoly 1981; Schneider et al. 2003; Wu et al. 2003). Jin and Hoskins (1995) suggested that upper-tropospheric meridional wind υ perturbations provide a useful diagnostic for tracing the origins and regions of influence of these waves (Schneider et al. 2003). Actually, the connection of Rossby wave propagation with droughts in the United States has been recognized (Chang and Wallace 1987; Lyon and Dole 1995; Chen and Newman 1998). Chang and Wallace (1987) identified a pattern of zonal teleconnection extending from the central Pacific to the North Atlantic. Lyon and Dole (1995) found that the propagation and sources of anomalous wave activity are different for summer droughts in the United States between 1980 and 1988. Chen and Newman (1998) emphasized the role of Rossby wave propagation originating in the west Pacific at subseasonal time scales in the initiation of anomalous anticyclones resulting in the 1988 drought in the United States.

To demonstrate the in-phase enhancement and out-of-phase cancellation of PDO- and ENSO-forced large-scale anomalies, Fig. 8 shows the composites of the zonal departure anomalies of υ at 200 hPa based on ENSO and PDO. For both composites based on ENSO or PDO (Figs. 8a,b), a similar wave pattern is seen over the Pacific and North America. The pattern suggests a Rossby wave propagation from tropical central and western Pacific to the North American continent via the North Pacific that is generally consistent with Chen and Newman (1998). The in-phase occurrence of ENSO and PDO (El Niño and warm PDO or La Niña and cold PDO) results in an enhancement of the pattern (Fig. 8c). There are positive values in the eastern part of North Pacific and the southeastern coast and negative ones in the central part of North America (Figs. 8a–c), favoring wet conditions in the Great Plains (Figs. 5a,b, 6a). The similar pattern with opposite sign is associated with dry conditions (Figs. 5a,b, 7a). However, the out-of-phase occurrence of ENSO and PDO (El Niño and cold PDO or La Niña and warm PDO) largely reduces the large-scale anomalies over the Pacific and North America (Fig. 8d), which does not favor the generation of dry or wet conditions in the Great Plains (Figs. 5c, 6b, 7b).

Fig. 8.

Anomaly zonal departure of v based on the composites of (a) El Niño–La Niña, (b) warm–cold PDO, as well as (c) their combination, (a) + (b), and (d) their difference, (a) − (b). The CI is 1 m s−1, and the zero line is omitted.

Fig. 8.

Anomaly zonal departure of v based on the composites of (a) El Niño–La Niña, (b) warm–cold PDO, as well as (c) their combination, (a) + (b), and (d) their difference, (a) − (b). The CI is 1 m s−1, and the zero line is omitted.

b. Frequency analyses

From these analyses, it is suggested that, on average, the anomalous conditions are likely (unlikely) to occur when ENSO and PDO are in (out of) phase. This point is further demonstrated when the two indices are plotted as a scatter diagram with different colors and shapes representing dry, neutral, and wet conditions in the Great Plains (Fig. 9). It is seen from Fig. 9 that wet (dry) conditions are indicated in the combination of El Niño and warm PDO (La Niña and cold PDO). The interference of ENSO and PDO on dry and wet conditions is further examined by calculating partial correlations, using the following formula (Zar 1998; Wu and Kirtman 2007):

 
formula

where Rab,c is the partial correlation between variables a and b without the influence of variable c; Rij refers to the conventional correlation between i and j; and indices a, b, and c represent the specific variables, Niño-3.4 index, PDO index, and the soil wetness anomaly averaged in the Great Plains, respectively. The correlation between Niño-3.4 index and the soil wetness drops from 0.31 to 0.18 after eliminating the effect of PDO (Table 1). Similarly, the correlation between PDO index and the soil wetness decreases from 0.32 to 0.21 after removing the effect of ENSO (Table 1). These correlations confirm the in-phase enhancement and the out-of-phase cancellation of the influence of ENSO and PDO on the dry and wet conditions in the Great Plains.

Fig. 9.

Scatterplot of Niño-3.4 index vs PDO index. The closed green squares represent wet months with the SW in the Great Plains ≥0.5 STDV, the opened red squares are dry months with the SW ≤ −0.5 STDV, and the closed black circles show neutral months with the SW between −0.5 and 0.5 STDV. The larger marks represent the corresponding values of the two indices averaged for dry, neutral, and wet months in the Great Plains.

Fig. 9.

Scatterplot of Niño-3.4 index vs PDO index. The closed green squares represent wet months with the SW in the Great Plains ≥0.5 STDV, the opened red squares are dry months with the SW ≤ −0.5 STDV, and the closed black circles show neutral months with the SW between −0.5 and 0.5 STDV. The larger marks represent the corresponding values of the two indices averaged for dry, neutral, and wet months in the Great Plains.

Table 1.

Simultaneous correlations between monthly mean SW in the Great Plains, Niño-3.4 index, and PDO index.

Simultaneous correlations between monthly mean SW in the Great Plains, Niño-3.4 index, and PDO index.
Simultaneous correlations between monthly mean SW in the Great Plains, Niño-3.4 index, and PDO index.

These results are also consistent with the results of relative frequencies (RF; Fig. 10, top) and absolute frequencies (AF; Fig. 10, bottom) of dry, neutral, and wet months in different combinations of the two indices. From AF, it is noted that two outstanding numbers are associated with the in-phase combinations of ENSO and PDO. AF values are about 9%–10%, which represents the percentages of contemporary occurrence of wet, El Niño, and warm PDO (dry, La Niña, and cold PDO) to all possible combinations of soil wetness, ENSO, and PDO. This is also consistent with RF numbers. When ENSO and PDO are in phase, the largest RF is wet (64%) for El Niño and warm PDO and dry (51%) for La Niña and cold PDO. The situation is different when ENSO and PDO are out of phase. The most likely condition is neutral (51%) for El Niño and cold PDO. For cold PDO and neutral ENSO combination, it is less likely to be wet (10%) than dry (38%) and neutral (52%) conditions. The differences of RF are small for all other combinations.

Fig. 10.

(top) Relative and (bottom) absolute frequency (%) of dry, neutral, and wet conditions in the Great Plains in different combinations of the Niño-3.4 and PDO indices: (a) warm PDO, (b) cold PDO, and (c) neutral PDO. The total absolute frequency (%) of the different combinations is marked at the top of each group. In each plot, the left, middle, and right groups are for El Niño, La Niña, and neutral ENSO, respectively. In each group, left (red), central (black), and right (green) bars represent dry, neutral, and wet conditions in the Great Plains, respectively, based on the definition in Fig. 9.

Fig. 10.

(top) Relative and (bottom) absolute frequency (%) of dry, neutral, and wet conditions in the Great Plains in different combinations of the Niño-3.4 and PDO indices: (a) warm PDO, (b) cold PDO, and (c) neutral PDO. The total absolute frequency (%) of the different combinations is marked at the top of each group. In each plot, the left, middle, and right groups are for El Niño, La Niña, and neutral ENSO, respectively. In each group, left (red), central (black), and right (green) bars represent dry, neutral, and wet conditions in the Great Plains, respectively, based on the definition in Fig. 9.

c. Seasonality of the association

As has been noted in previous works, there is a seasonal dependence of ENSO impacts on seasonal mean and daily extremes of precipitation and temperature in some parts of the United States (see, e.g., section 1; Sittel 1994; Green et al. 1997; Higgins et al. 2002; Enloe et al. 2004; Wang et al. 2007; Yang et al. 2007). Seasonality is important both from a viewpoint of the relationship between the variations of soil wetness and those of circulation and also from potential applications of the composites to operational climate forecasts. Because of the different features of the persistence of soil wetness compared with temperature and precipitation, it is particularly interesting to compare seasonal differences of the PDO and ENSO impacts on wetness anomalies in the Great Plains.

The seasonal dependence of the association is demonstrated in Fig. 11 and Table 1 by using monthly data in different seasons. It is seen that the general scatter patterns are similar for the months in different seasons (Fig. 11). This may be due to using soil wetness instead of precipitation to define dry and wet conditions, resulting in less seasonal dependence of the results. Nevertheless, there are quantitative differences in the correlations between the soil wetness in the Great Plains and Niño-3.4–PDO indices in different seasons (Table 1). The numbers in Table 1 suggest that the strongest correlations are for the months of spring [March–May (MAM)] and weakest for the months of autumn [September–November (SON)], whereas it is in between for the months of winter [December–February (DJF)] and summer [June–August (JJA)]. Meanwhile, the largest drop of the correlations between Niño-3.4 (PDO) index and the soil wetness is observed in spring when eliminating the influence of PDO (Niño-3.4). The correlation in spring between the Niño-3.4 index and the soil wetness decreases from 0.40 to 0.25 when the PDO influence is removed. Similarly, the correlation between the PDO index and the soil wetness drops from 0.42 to 0.27 when the ENSO influence is suppressed.

Fig. 11.

As in Fig. 9, but for (a) DJF, (b) MAM, (c) JJA, and (d) SON.

Fig. 11.

As in Fig. 9, but for (a) DJF, (b) MAM, (c) JJA, and (d) SON.

The seasonality of the relationship may be connected with the seasonal dependence of the anomalies and the pattern of mean seasonal cycle in the North Pacific. There is a strong seasonal cycle of the large-scale pattern of SLP in the North Pacific (Fig. 12). In spring, the subtropical high is close to its annual mean strength in the subtropical North Pacific and the Aleutian low occupies the high latitudes of the North Pacific (Fig. 12a). In the following season (summer), the subtropical high strengthens significantly and extends northward, whereas the SLP decreases over the Asian and North American continents (Fig. 12b). The pattern of SLP in autumn (Fig. 12c) returns to a similar state to that in spring (Fig. 12a), and the pattern in winter (Fig. 12d) is almost opposite to that in summer, with the subtropical high at its weakest, whereas the Aleutian low is dominant (Fig. 11b). Overall, the amplitudes in the anomalous seasonal cycle of SLP are much larger in summer and winter than in spring and autumn (Fig. 12). Thus, the effect of anomalous signals associated with ENSO and PDO on soil wetness may not add much to those associated with the seasonal cycle in summer and winter, implying a weaker impact of ENSO and PDO on dry/wet conditions in the Great Plains in these two seasons.

Fig. 12.

Seasonal cycle of anomalous (shading) and mean SLP (contours) averaged for (a) MAM, (b) JJA, (c) SON, and (d) DJF in January 1950–December 2005. The CI is 3 hPa.

Fig. 12.

Seasonal cycle of anomalous (shading) and mean SLP (contours) averaged for (a) MAM, (b) JJA, (c) SON, and (d) DJF in January 1950–December 2005. The CI is 3 hPa.

However, this argument still cannot explain the fact that although the anomalous seasonal cycle of SLP is smaller in autumn than in winter and summer (Figs. 12b–d), its SST–soil wetness relationship is the weakest among seasons. On the other hand, our composites based on seasons of the anomalous fields show that the anomalous pattern is weakest and most disorganized in autumn. The anomalies of SST in the tropical Pacific associated with wet/dry conditions in the Great Plains in autumn are the smallest among the four seasons (Fig. 13). Furthermore, negative anomalies of SLP are dominant in the North Pacific, particularly in the central and eastern parts, in all the seasons, except for autumn (Fig. 13). When the wind anomalies, associated with the negative anomalies of SLP, are parallel to the coast, the Ekman effect suppresses local upwelling and causes a local warming that favors the northward extension of the positive SST anomalies along the west coast, which is the case of MAM and JJA (Figs. 13a,b). In DJF, the wind anomalies are parallel to the coast in the northwestern part of North America, causing positive SST anomalies there. However, in the central and southern parts, the wind anomalies are perpendicular to the coast, resulting in small local SST anomalies (Fig. 13d). On the contrary, positive (negative) anomalies of SLP are observed in the eastern (western) part of the North Pacific in autumn (Fig. 13c), which may prevent the northward extension of SST anomalies along the west coast because of the Ekman effect and push the main positive anomalies of SST in the North Pacific off the west coast to the central North Pacific (Fig. 13c), causing the weakest relationship between the dry/wet conditions in the Great Plains and ENSO–PDO. We also notice that the anomalous pattern, especially the SLP, is strongest in spring (Fig. 13), which is consistent with our result that the SST–soil wetness relation is the strongest in the season.

Fig. 13.

Monthly mean anomalous composites based on SW (°C) anomaly over the Great Plains averaged in 3 months (wet month mean minus dry month mean) in (a) MAM, (b) JJA, (c) SON, and (d) DJF. Shading (contours) represents SST (SLP). The CI is 0.5 hPa.

Fig. 13.

Monthly mean anomalous composites based on SW (°C) anomaly over the Great Plains averaged in 3 months (wet month mean minus dry month mean) in (a) MAM, (b) JJA, (c) SON, and (d) DJF. Shading (contours) represents SST (SLP). The CI is 0.5 hPa.

5. Summary and discussion

In this study, we have examined the impact of monthly mean ENSO and PDO on dry and wet conditions in the Great Plains and the associated large-scale atmosphere and ocean conditions. Our focus is on the influence of ENSO and PDO interference on dry and wet conditions. It is shown that both ENSO and PDO can generate a similar pattern of atmospheric and oceanic anomalies over the eastern part of the North Pacific and the western North America continent, which strongly influences the regional dry and wet conditions over the Great Plains. Moreover, the relationship between ENSO–PDO and climate anomalies in the Great Plains is intensified when ENSO and PDO are in phase. On average, the Great Plains displays anomalous wet (dry) conditions with the combination of El Niño and warm PDO (La Niña and cold PDO). However, when ENSO and PDO are out of phase, the relationship is weakened and the Great Plains tends toward neutral. Without ENSO, PDO alone does not affect the North American climate significantly. The relationship is quite robust for different seasons, with the strongest relationship for the months of spring, and the weakest relationship for the months of autumn, whereas the months of winter and summer are in between. The seasonality of the relationship may be associated with the seasonal dependence of the anomalies of general circulation and the pattern of mean seasonal cycle in the North Pacific. The maximal correlation coefficients occur when the Niño-3.4 index leads the soil wetness anomalies by 1–2 months and when the PDO index lags the soil wetness anomalies by 1 month.

The contrasting impact of the interference of ENSO and PDO on the climate anomalies in the Great Plains is associated with the differences of the ocean–atmosphere anomalies in the lower troposphere. When ENSO and PDO are in phase, the SST anomalies are extended from the equatorial Pacific to the higher latitudes of the North Pacific via the eastern ocean. The distribution of the corresponding SLP anomalies is in the shape of an ellipse with a southeast–northwest orientation of the long axis because the SST anomalies promote coherent SLP change over the central North Pacific. However, when ENSO and PDO are out of phase, the SST anomalies have the same sign in the tropical and central North Pacific, which is opposite to the anomalies near the west coast of North America. The anomalously cyclonic circulation over the North Pacific is weaker in the out-of-phase composite than in the in-phase composite. The distribution of the anomalies of SLP and the wind at 1000 hPa resembles a circle. Compared with the out-of-phase composite, the pronounced differences of atmosphere circulation in the in-phase composite are mainly located in the eastern part of the North Pacific, which shows a southeastward shift of the anomalously cyclonic circulation over the North Pacific.

Meanwhile, in the upper troposphere, there are clear differences in stationary wave activity between composites of the in-phase and out-of-phase ENSO and PDO, which are associated with the dry/wet climate in the Great Plains. When they are in phase, a similar wave activity pattern with the same sign generated by ENSO and PDO is overlapped and enhanced, which favors anomalous (dry or wet) conditions in the Great Plains. In contrast, when out of phase, they generate a similar wave activity pattern with opposite sign, causing cancellation of the anomalous circulation and favoring neutral conditions in the Great Plains.

In addition, the SST differences in the eastern part of the North Pacific between the in-phase and out-of-phase composites based on ENSO and PDO may also suggest an overlap of the local upwelling signals associated with PDO- and ENSO-induced coastal Kelvin waves. When ENSO and PDO are in phase, the signal propagating from the tropical to high latitudes along the North Pacific coast has the same sign as the one in the eastern part of the North Pacific associated with PDO. The two signals reinforce each other. The overlap of the same sign signal in the eastern part of the North Pacific enhances the anomalies in the North Pacific and intensifies the climate anomalies in the Great Plains. However, when ENSO and PDO are out of phase, the signal propagating from the tropical Pacific has a sign opposite to the one associated with PDO, resulting in a cancellation of the overall anomalies in the eastern part of the North Pacific, which is unfavorable for forcing climate anomalies in the Great Plains.

There are a few caveats when applying the results. First, because the property of an ENSO event is quite different in terms of its strength and its position of maximum center, the soil wetness anomaly in the Great Plains may mostly depend on the state of ENSO rather than the ENSO–PDO relationship. It is not easy to separate the role of PDO from the ENSO influences based on the analysis in observations. It is needed to test the hypothesis using the idealized model experiments. Furthermore, because the composites of El Niño–La Niña and warm–cold PDO are examined, only the symmetric part of the ENSO and PDO influence on the climate in the Great Plains was considered in this work. As many previous studies have pointed out, nonlinear response to ENSO and PDO might also have an important impact on the climate in the Great Plains. Finally, PDO is very poorly sampled at interdecadal time scales; essentially, it has one positive phase and one negative phase during 1950–2005. Thus, it is unclear what impact PDO intensity has on dry and wet conditions in the Great Plains. In addition, the PDO has shown more interannual variability than multidecadal variability since 1998, which may lessen its usefulness as a long-range forecast tool.

Acknowledgments

The authors thank Editor Prof. A. Pitman and anonymous reviewers, as well as Paul Dirmeyer, V. Krishnamurthy, Song Yang, Peitao Peng, Yun Fan, and Renguang Wu for their suggestions and comments to significantly improve the manuscript. Dan Paolino edited the manuscript. This work was supported by the NOAA CVP Program (NA07OAR4310310).

REFERENCES

REFERENCES
An
,
S. I.
,
J. S.
Kug
,
A.
Timmermann
,
I. S.
Kang
, and
O.
Timm
,
2007
:
The influence of ENSO on the generation of decadal variability in the North Pacific.
J. Climate
,
20
,
667
680
.
Andreadis
,
K. M.
,
E. A.
Clark
,
A. W.
Wood
,
A. F.
Hamlet
, and
D. P.
Lettenmaier
,
2005
:
Twentieth-century drought in the conterminous United States.
J. Hydrometeor.
,
6
,
985
1001
.
Barlow
,
M.
,
S.
Nigam
, and
E. H.
Berbery
,
2001
:
ENSO, Pacific decadal variability, and U.S. summertime precipitation, drought, and stream flow.
J. Climate
,
14
,
2105
2128
.
Brown
,
D. P.
, and
A. C.
Comrie
,
2004
:
A winter precipitation ‘dipole’ in the western United States associated with multidecadal ENSO variability.
Geophys. Res. Lett.
,
31
,
L09203
.
doi:10.1029/2003GL018726
.
Chang
,
F-C.
, and
J. M.
Wallace
,
1987
:
Meteorological conditions during heat waves and droughts in the United States Great Plains.
Mon. Wea. Rev.
,
115
,
1253
1269
.
Chen
,
M.
,
P.
Xie
,
J. E.
Janowiak
, and
P. A.
Arkin
,
2002
:
Global land precipitation: A 50-yr monthly analysis based on gauge observations.
J. Hydrometeor.
,
3
,
249
266
.
Chen
,
P.
, and
M.
Newman
,
1998
:
Rossby wave propagation and the rapid development of upper-level anomalous anticyclones during the 1988 drought.
J. Climate
,
11
,
2491
2504
.
Diaz
,
H. F.
,
1983
:
Some aspects of major dry and wet periods in the contiguous United States, 1895–1981.
J. Climate Appl. Meteor.
,
22
,
3
16
.
Englehart
,
P. J.
, and
A. V.
Douglas
,
2002
:
On some characteristic variations in warm season precipitation over the central United States (1910–2000).
J. Geophys. Res.
,
107
,
4286
.
doi:10.1029/2001JD000972
.
Englehart
,
P. J.
, and
A. V.
Douglas
,
2003
:
Assessing warm season drought episodes in the central United States.
J. Climate
,
16
,
1831
1842
.
Enloe
,
J.
,
J. J.
O’Brien
, and
S. R.
Smith
,
2004
:
ENSO Impacts on peak wind gusts in the United States.
J. Climate
,
17
,
1728
1737
.
Fan
,
Y.
, and
H.
Van den Dool
,
2004
:
Climate Prediction Center global monthly soil moisture data set at 0.5° resolution for 1948 to present.
J. Geophys. Res.
,
109
,
D10102
.
doi:10.1029/2003JD004345
.
Gershunov
,
A.
, and
T. P.
Barnett
,
1998
:
Interdecadal modulation of ENSO teleconnections.
Bull. Amer. Meteor. Soc.
,
79
,
2715
2725
.
Goodrich
,
G. B.
,
2007
:
Influence of the Pacific decadal oscillation on winter precipitation and drought during years of neutral ENSO in the western United States.
Wea. Forecasting
,
22
,
116
124
.
Green
,
P. M.
,
D. M.
Legler
,
C. J.
Miranda
V
, and
J. J.
O’Brien
,
1997
:
The North American climate patterns associated with the El Niño-Southern Oscillation.
COAPS Project Rep. Series 97–1, 8 pp. [Available online at http://www.coaps.fsu.edu/lib/booklet/]
.
Gutzler
,
D. S.
,
D. M.
Kann
, and
C.
Thornbrugh
,
2002
:
Modulation of ENSO-based long-lead outlooks of Southwestern U.S. winter precipitation by the Pacific decadal oscillation.
Wea. Forecasting
,
17
,
1163
1172
.
Harshburger
,
B.
,
H.
Ye
, and
J.
Dzialoski
,
2002
:
Observational evidence of the influence of Pacific SSTs on winter precipitation and spring stream discharge in Idaho.
J. Hydrol.
,
264
,
157
169
.
doi:10.1016/S0022-1694(02)00072-0
.
Hidalgo
,
H. G.
, and
J. A.
Dracup
,
2003
:
ENSO and PDO effects on hydroclimatic variations of the Upper Colorado River basin.
J. Hydrometeor.
,
4
,
5
23
.
Higgins
,
R. W.
,
A.
Leetmaa
, and
V. E.
Kousky
,
2002
:
Relationships between climate variability and winter temperature extremes in the United States.
J. Climate
,
15
,
1555
1572
.
Hoerling
,
M.
, and
A.
Kumar
,
2003
:
The perfect ocean for drought.
Science
,
299
,
691
694
.
Hoskins
,
B. J.
, and
D. J.
Karoly
,
1981
:
The steady linear response of a spherical atmosphere to thermal and orographic forcing.
J. Atmos. Sci.
,
38
,
1179
1196
.
Hu
,
Q.
, and
S.
Feng
,
2001
:
Variations of teleconnection of ENSO and interannual variation in summer rainfall in the central United States.
J. Climate
,
14
,
2469
2480
.
Huang
,
J.
,
H. M.
Van den Dool
, and
K. P.
Georgakakos
,
1996
:
Analysis of model- calculated soil moisture over the United States (1931–1993) and applications to long-range temperature forecasts.
J. Climate
,
9
,
1350
1362
.
Jin
,
F-F.
, and
B. J.
Hoskins
,
1995
:
The direct response to tropical heating in a baroclinic atmosphere.
J. Atmos. Sci.
,
52
,
307
319
.
Kalnay
,
E.
, and
Coauthors
,
1996
:
The NCEP/NCAR 40-Year Reanalysis Project.
Bull. Amer. Meteor. Soc.
,
77
,
437
471
.
Larkin
,
N. K.
, and
D. E.
Harrison
,
2005
:
On the definition of El Niño and associated seasonal average weather anomalies.
Geophys. Res. Lett.
,
32
,
L13705
.
doi:10.1029/2005GL022738
.
Lyon
,
B.
, and
R. M.
Dole
,
1995
:
A diagnostic comparison of the 1980 and 1988 U.S. summer heat wave-droughts.
J. Climate
,
8
,
1658
1675
.
Mantua
,
N. J.
,
S. R.
Hare
,
Y.
Zhang
,
J. M.
Wallace
, and
R. C.
Francis
,
1997
:
A Pacific interdecadal climate oscillation with impacts on salmon production.
Bull. Amer. Meteor. Soc.
,
78
,
1069
1079
.
McCabe
,
G. J.
, and
M. D.
Dettinger
,
1999
:
Decadal variations in the strength of ENSO teleconnections with precipitation in the western United States.
Int. J. Climatol.
,
19
,
1399
1410
.
McCabe
,
G. J.
,
M. A.
Palecki
, and
J. L.
Betancourt
,
2004
:
Pacific and Atlantic Ocean influences on multidecadal drought frequency in the United States.
Proc. Natl. Acad. Sci. USA
,
101
,
4136
4141
.
Mo
,
K. C.
,
J. N.
Paegle
, and
R. W.
Higgins
,
1997
:
Atmospheric processes associated with summer floods and droughts in the central United States.
J. Climate
,
10
,
3028
3046
.
Namias
,
J.
,
1955
:
Some meteorological aspects of drought with special reference to the summers of 1952–1954 over the United States.
Mon. Wea. Rev.
,
83
,
199
205
.
Reynolds
,
R. W.
,
N. A.
Rayner
,
T. M.
Smith
,
D. C.
Stokes
, and
W.
Wang
,
2002
:
An improved in situ and satellite SST analysis for climate.
J. Climate
,
15
,
1609
1625
.
Ropelewski
,
C. F.
, and
M. S.
Halpert
,
1986
:
North American precipitation and temperature patterns associated with the El Niño/Southern Oscillation (ENSO).
Mon. Wea. Rev.
,
114
,
2352
2362
.
Schneider
,
E. K.
,
L.
Bengtsson
, and
Z-Z.
Hu
,
2003
:
Forcing of Northern Hemisphere climate trends.
J. Atmos. Sci.
,
60
,
1504
1521
.
Schneider
,
N.
, and
B. D.
Cornuelle
,
2005
:
The forcing of the Pacific decadal oscillation.
J. Climate
,
18
,
4355
4373
.
Schubert
,
S. D.
,
M. J.
Suarez
,
P. J.
Pegion
,
R. D.
Koster
, and
J. T.
Bacmeister
,
2004a
:
Causes of long-term drought in the U.S. Great Plains.
J. Climate
,
17
,
485
503
.
Schubert
,
S. D.
,
M. J.
Suarez
,
P. J.
Pegion
,
R. D.
Koster
, and
J. T.
Bacmeister
,
2004b
:
On the cause of the 1930s Dust Bowl.
Science
,
303
,
1855
1859
.
Shukla
,
J.
,
1981
:
Dynamical predictability of monthly means.
J. Atmos. Sci.
,
38
,
2547
2572
.
Sittel
,
M.
,
1994
:
Differences in the means of ENSO extremes for maximum temperature and precipitation in the United States.
COAPS Tech. Rep. 94-2, 50 pp. [Available from Center for Ocean-Atmospheric Prediction Studies, Florida State University, Tallahassee, FL 32306-2840]
.
Wang
,
Z.
,
C-P.
Chang
, and
B.
Wang
,
2007
:
Impacts of El Niño and La Niña on the U.S. climate during northern summer.
J. Climate
,
20
,
2165
2177
.
Wu
,
R.
, and
B. P.
Kirtman
,
2007
:
Observed relationship of spring and summer East Asian rainfall with winter and spring Eurasian snow.
J. Climate
,
20
,
1285
1304
.
Wu
,
R.
,
Z-Z.
Hu
, and
B. P.
Kirtman
,
2003
:
Evolution of ENSO-related rainfall anomalies in East Asia and the processes.
J. Climate
,
16
,
3742
3758
.
Yang
,
S.
,
X.
Ding
,
D.
Zheng
, and
Q.
Li
,
2007
:
Depiction of the variations of Great Plains precipitation and its relationship with tropical central-eastern Pacific SST.
J. Appl. Meteor. Climatol.
,
46
,
136
153
.
Zar
,
J. H.
,
1998
:
Biostatistical Analysis.
4th ed. Prentice Hall, 663 pp
.
Zhang
,
Y.
,
J. M.
Wallace
, and
D. S.
Battisti
,
1997
:
ENSO-like interdecadal variability: 1900–93.
J. Climate
,
10
,
1004
1020
.

Footnotes

Corresponding author address: Zeng-Zhen Hu, Center for Ocean-Land-Atmosphere Studies, 4041 Powder Mill Road, Suite 302, Calverton, MD 20705. Email: hu@cola.iges.org