Abstract

Decadal variability of summertime Great Plains surface temperature is probed from the perspective of the Great Plains low-level jet (GPLLJ). GPLLJ variability modes 2 and 5 are shown to be most influential on the evolution and magnitude of Great Plains surface temperature anomalies over the latter half of the twentieth century, including the development of the summertime warming hole and are further linked to the Pacific decadal oscillation (PDO) and Atlantic multidecadal oscillation (AMO), respectively. The connection between GPLLJ variability and Great Plains surface temperature is strongest when the PDO and AMO are oppositely phased, and in the case of the warming hole, a preference for a positive (negative) PDO (AMO).

The influence of remote SST variability on the central U.S. warming hole is broadly consistent with previous modeling studies. However, the pivotal role that GPLLJ variability plays in linking the hemispheric-wide SST variability (through the AMO and PDO) to the regional warming hole is an expanded and clarified perspective. These findings unify the results of recent studies from the U.S. Climate Variability and Predictability (CLIVAR) Drought Working Group and have implications for decadal climate prediction efforts.

1. Introduction

During the latter half of the twentieth century a warm season negative surface temperature anomaly developed over the central U.S. in the midst of continental warming. This observed cooling trend dubbed the “warming hole” (Kunkel et al. 2006) reached a peak in the 1990s, however, it has weakened during the most recent decade, suggesting that decadal climate variability may be influential in the temporal fluctuations of Great Plains surface temperature anomalies. The extent to which decadal climate variability is linked to regional Great Plains surface temperature variations is an important question in efforts to clarify the nature of the warming hole and its future trajectory.

Attempts to uncover the mechanisms that led to the warming hole have been largely model based, with general circulation models indicating a role for both external forcing and natural variability internal to the climate system (Kunkel et al. 2006; Robinson et al. 2002). Decadal sea surface temperature (SST) variability in the Pacific basin was implicated in the regionality and seasonality of observed and simulated surface temperature trends over the United States during 1950–2000 (Wang et al. 2009). Given the primacy of Great Plains low-level jet (GPLLJ) fluctuations in generating regional patterns of summertime hydroclimate variability (Weaver and Nigam 2008; Weaver et al. 2009a) it is pertinent to investigate the decadal variability of the GPLLJ and its regional surface temperature impacts, especially since modeling studies indicate that decadal SST patterns in the Atlantic and Pacific Oceans are influential in GPLLJ variability (Weaver et al. 2009b).

In this study the summertime warming hole is revisited from the perspective of GPLLJ variability modes and their regional surface temperature impacts. The assessment of GPLLJ variability relies on observationally constrained data (i.e., reanalysis) as opposed to model simulations in an effort to clarify the pertinent physical mechanisms that give rise to the observed summertime Great Plains surface temperature anomalies.

The primary motivation for the analysis is provided in Fig. 1, which depicts the 1950–2010 July–September (JAS) surface temperature trend. A positive surface temperature trend is entrenched over portions of the western and southeast United States and nearly all of southern Canada. While much of the eastern one-third of the United States has experienced a warming trend over the last 61 years, it is quite weak when compared to that over the western United States and southern Canada, consistent with the west–east gradient in annual U.S. surface temperature trends (Meehl et al. 2009). The warming hole is characterized by a large region of negative-to-neutral surface temperature trends over the central United States. The focus here is on JAS since that “season” exhibits the strongest warming hole amplitude (seasonal comparison not shown).

Fig. 1.

Linear trend of JAS surface temperature for 1950–2010. Surface temperature is shaded in K.

Fig. 1.

Linear trend of JAS surface temperature for 1950–2010. Surface temperature is shaded in K.

The data sources and methodology will be described in section 2. GPLLJ variability and its relationship to Great Plains surface temperature anomalies and decadal variability modes will be presented in section 3, while section 4 is left for the discussion.

2. Data and methodology

GPLLJ variability for the 1950–2010 period is assessed by conducting an empirical orthogonal function (EOF) analysis on seasonal anomalies of JAS 850-hPa meridional wind field over the domain 105°–80°W; 20°–50°N in the National Centers for Environmental Prediction (NCEP)–National Center for Atmospheric Research (NCAR) reanalysis (Kalnay et al. 1996). As in Weaver and Nigam (2008) a covariance-based analysis on the latitudinally weighted field was performed. The EOFs are not rotated given the limited analysis domain. The principal components (PCs) obtained from this analysis are used in relating GPLLJ variability to the warming hole via surface temperature anomaly reconstructions. The time series reconstructions are performed by taking the product of a given GPLLJ PC surface temperature regression pattern and the value of that GPLLJ PC for a given year, for each year of the 1950–2010 time period. The annual JAS reconstructions are then subject to 5 iterations of a 1–2–1 smoother to highlight the decadal variability. The choice of 5 iterations is to adequately smooth out a majority of the interannual variability without removing extensive portions of the early and latter parts of the data record. Varying the number of iterations has no appreciable impact on the results. The 2-m surface air temperature data is from the Global Historical Climate Network (GHCN)/Climate Anomaly Monitoring System (CAMS) dataset developed at the Climate Prediction Center (Fan and van den Dool 2008). All surface temperature and reconstructed time series are with respect to the area average within the enclosed region depicted in Fig. 2, which is bound by 32°–45°N and 102°–90°W. Seasonal mean JAS values are calculated from the monthly Atlantic multidecadal oscillation (AMO) and Pacific decadal oscillation (PDO) indices obtained from the National Oceanic and Atmospheric Administration (NOAA) Earth System Research Laboratory (ESRL) (http://www.esrl.noaa.gov/psd/data/climateindices/list/).

Fig. 2.

The 1950–2010 JAS GPLLJ PC regressions to 850-hPa meridional wind (contoured) and surface temperature (shaded) for modes 1–5. The 850-hPa meridional wind is contoured at 0.2 m s−1 and surface temperature is shaded at 0.2 K. The red box outlines the target study area for subsequent analyses.

Fig. 2.

The 1950–2010 JAS GPLLJ PC regressions to 850-hPa meridional wind (contoured) and surface temperature (shaded) for modes 1–5. The 850-hPa meridional wind is contoured at 0.2 m s−1 and surface temperature is shaded at 0.2 K. The red box outlines the target study area for subsequent analyses.

3. Results

a. Spatial variability

Shown in Fig. 2 are EOFs 1–5 of JAS 850-hPa meridional wind anomalies (contoured) and their corresponding PC regressions to surface temperature anomalies (shaded). Modes 1–5 explain 37%, 18%, 12%, 9%, and 6% of the meridional wind variance respectively. The meridional wind structures generally depict an in place amplitude modulation (modes 1, 4, and 5) and spatial shifting (modes 2 and 3) of the GPLLJ. The surface temperature regressions from PCs 1 and 4 exhibit a characteristic surface temperature pattern similar to the observed twentieth-century west–east annual surface temperature trend dipole (Trenberth et al. 2007; Meehl et al. 2009). Conversely GPLLJ modes 2, 3, and 5 have strong surface temperature impacts over the central United States, especially in the case of GPLLJ modes 2 and 5.

Even though all 5 GPLLJ modes have anomalous surface temperature footprints over the core region of interest, each mode may not contribute equally to the total surface temperature variability over the 1950–2010 period. An analysis of the various contributions from each of the first 5 GPLLJ modes reveals that modes 2 and 5 are the most important for Great Plains surface temperature variability. This is highlighted in Fig. 3, which shows the reconstructed fraction of surface temperature variance from GPLLJ modes 1, 3, and 4 (upper) and modes 2 and 5 (lower). The sum of modes 1, 3, and 4 show a negligible contribution to the total surface temperature variance over the warming hole region, however, are quite important in the continental scale surface temperature variability outside of the Great Plains region, especially in western Canada where the fraction of surface temperature variance exceeds 70%. Conversely, GPLLJ modes 2 and 5 exert their strongest influence over the surface temperature variability in the warming hole region with a weak contribution elsewhere.

Fig. 3.

Fraction of variance of GPLLJ reconstructed surface temperature anomalies expressed as a percentage of total surface temperature variance for (top) modes 1, 3, and 4 and (bottom) modes 2 and 5. Shading is in intervals of 10%.

Fig. 3.

Fraction of variance of GPLLJ reconstructed surface temperature anomalies expressed as a percentage of total surface temperature variance for (top) modes 1, 3, and 4 and (bottom) modes 2 and 5. Shading is in intervals of 10%.

To be sure, a comparison of Figs. 1 and 3 indicates that the GPLLJ modes may have a marginal contribution to the surface temperature cooling trend maxima over central Texas. A further examination of the amplitude of reconstructed surface temperature trends based on GPLLJ variability modes confirms this, since the first 5 modes of GPLLJ variability combined explain at most 15% of the central Texas surface temperature maxima (not shown). However, the analysis also reveals that GPLLJ variability modes can account for up to 50% of surface temperature trend amplitude maxima in northwestern Missouri and nearly 100% elsewhere over the Great Plains.

The pathway through which GPLLJ modes 2 and 5 influence the JAS surface temperature is likely via changes in precipitation, given the dominance of the GPLLJ in governing warm season rainfall variability (Higgins et al. 1997; Weaver and Nigam 2008). An assessment of similarly regressed precipitation patterns (not shown) reveals a preference for positive precipitation anomalies over the entire Great Plains and upper Midwest for mode 2 and the central Great Plains and southeast for mode 5, consistent with central U.S. cooling, and likely due to surface energy balance changes from the combination of reduced insolation due to enhanced cloud cover and increased evaporative cooling from augmented contemporaneous soil moisture storage (Zhao and Khalil 1993).

b. Temporal variability

Clarifying the evolution of the low-frequency surface temperature fluctuations of consequence to the recently observed warming hole motivates an assessment of the temporal variability of the GPLLJ and its related impact on the evolution of surface temperature anomalies over the 1950–2010 time period. Figure 4 (top) shows the smoothed time series of GPLLJ variability for modes 2 and 5 with the interannual variation displayed with bars. Smoothed PC 2 (blue) hovers around zero during the early part of the record, however, is also characterized by an increasing trend between the 1950s and 1990, with a sudden and dramatic shift toward negative values thereafter. Smoothed PC 5 (red) appears multidecadal-like with negative values dominating from around 1965 to the mid-1980s. A similar analysis of PCs 1, 3, and 4 (not shown) indicates that PCs 1 and 3 are consistently out of phase with each other, while PC 4 is inherently weak throughout the period.

Fig. 4.

(top) Smoothed (1–2–1) GPLLJ PC time series for modes 2 (blue) and 5 (red). Interannual variation for modes 2 and 5 are shown in the blue and red bars respectively. (bottom) Smoothed (1–2–1) GPLLJ PC surface temperature anomaly reconstructions. The red and blue lines denote the PC 1–5 and PC 2 and 5 only surface temperature reconstructions respectively. The solid black line shows the total observed smoothed surface temperature anomaly while the gray bars indicate the total surface temperature anomaly interannual variation. All values are in K. Five applications of the 1–2–1 smoothing truncate the record by 5 years on each end.

Fig. 4.

(top) Smoothed (1–2–1) GPLLJ PC time series for modes 2 (blue) and 5 (red). Interannual variation for modes 2 and 5 are shown in the blue and red bars respectively. (bottom) Smoothed (1–2–1) GPLLJ PC surface temperature anomaly reconstructions. The red and blue lines denote the PC 1–5 and PC 2 and 5 only surface temperature reconstructions respectively. The solid black line shows the total observed smoothed surface temperature anomaly while the gray bars indicate the total surface temperature anomaly interannual variation. All values are in K. Five applications of the 1–2–1 smoothing truncate the record by 5 years on each end.

To assess the GPLLJ contribution to the temporal evolution of the warming hole, Fig. 4 (bottom) shows the time series of smoothed reconstructed surface temperature anomalies as a function of PCs 1–5 (red) and PCs 2 and 5 only (blue). The interannual variation in surface temperature (i.e., unsmoothed) is shown in gray bars with the corresponding smoothed anomaly in black. The complete five-mode reconstructed surface temperature anomaly is nearly identical to that derived from PCs 2 and 5 only (2 + 5), highlighting the importance of these 2 modes on the warming hole region. The time series correlation between the total surface temperature anomaly and its GPLLJ reconstructed counterpart is 0.53 (0.56) for PC 2 + 5 (PC 1–5). Consequently, including modes 1, 3, and 4 in the reconstruction is trivial.

The appearance of decadal fluctuations in the surface temperature anomalies (total and reconstructed), and GPLLJ PCs 2 (multidecadal trend) and 5 (multidecadal variation) in Fig. 4, suggests a role for natural decadal variability on the GPLLJ and its influence on the warming hole. Recent evidence from the U.S. Climate Variability and Predictability (CLIVAR) Drought Working Group indicates that varying polarities of the PDO and AMO SST patterns influences JAS GPLLJ variability (Weaver et al. 2009b). Specifically, a positive PDO SST pattern combined with a negative AMO SST pattern tends to strengthen the total 850-hPa meridional wind anomaly, and hence the GPLLJ, while the opposite diminishes it. In the present context, PC 5 appears susceptible to changes in the AMO (cf. red curves in Figs. 4 and 5), and has a time series correlation of 0.28 (smoothed correlation 0.59), while PC 2 (blue) appears more closely associated with the PDO (blue), with a 0.26 correlation (smoothed correlation 0.64). In the case of PC 2 the relationship is not exclusive to the PDO as this GPLLJ mode is also correlated to the AMO at −0.24. All correlations are significant at the 95% level based on a t test.

Fig. 5.

Smoothed (1–2–1) AMO (red) and PDO (blue) indices. The blue and red bars indicate the interannual variation in the PDO and AMO respectively. The anomalous SST structure is given for selected sub periods by Pacific (P), Atlantic (A), cold (c), and warm (w), while each basin’s positive or negative contribution to the southerly meridional wind anomaly over the Great Plains is given by + and −, respectively as diagnosed by Weaver et al. (2009b). Five applications of the 1–2–1 smoothing truncate the record by 5 years on each end.

Fig. 5.

Smoothed (1–2–1) AMO (red) and PDO (blue) indices. The blue and red bars indicate the interannual variation in the PDO and AMO respectively. The anomalous SST structure is given for selected sub periods by Pacific (P), Atlantic (A), cold (c), and warm (w), while each basin’s positive or negative contribution to the southerly meridional wind anomaly over the Great Plains is given by + and −, respectively as diagnosed by Weaver et al. (2009b). Five applications of the 1–2–1 smoothing truncate the record by 5 years on each end.

The somewhat weak (although statistically significant) correlations between the GPLLJ modes 2–5 and the PDO–AMO over the full 1950–2010 time period may mask some potentially useful information regarding the low-frequency variations of the GPLLJ influenced surface temperature variability, especially if there are interdecadal variations in the strength of the correlation. Focusing on the most recent 3 decades shows a remarkable coherence between the smoothed total surface temperature anomaly and the GPLLJ reconstruction prior to 1961 and from ~1977 onward. At the beginning of the recent subperiod there is a switch from a negative to positive PDO, while the AMO remains negative (Fig. 5). This leads to a nearly 10-yr long period with a warm Pacific and cold Atlantic (PwAc), an ideal SST configuration to promote GPLLJ strengthening (Weaver et al. 2009b), much like the +mode-2 and –mode-5 patterns in Fig. 2. Despite the PC 2 sign switch around 1977 and its cooling impact, a positive surface temperature anomaly is maintained from the late 1970s-to-early 1980s on account of the strong warming contribution by the negative phase of PC 5 associated with the negative AMO during this period. This highlights the subtle complexities and a clarified view when examining the GPLLJ from the modal perspective, as opposed to the total meridional wind anomaly (as in Weaver et al. 2009), specifically, the opposing influence of the AMO on GPLLJ modes 2 and 5 and their contributions to the warming hole.

Despite this quick excursion of the total surface temperature anomaly into positive territory, a negative surface temperature anomaly rapidly redevelops and strengthens during the mid-1980s to mid-1990s as a continually amplifying PC 2 takes on a more prominent role, in concert with the strengthening positive PDO (Fig. 5). The AMO sign switch from negative to positive during this time coincides with PC 5 rapidly becoming positive (cooling), contributing to a swift and substantial augmentation of the warming hole. Evidently, it is the combined contribution of PCs 2 and 5 and their rapid ascension that contributed to such a strong warming hole during the 1990s. Similarly, a weakening of the warming hole during the most recent decade is partially attributable to a rapid sign reversal of GPLLJ PC 2 (warming contribution), combined with a weakening of PC 5, which is coincident with both the shift to a strong warm AMO and a return to a cold PDO. This opposes the 1990s SST configuration, and is consistent with a weakened GPLLJ response to a cold Pacific and warm Atlantic (Weaver et al. 2009b).

Equally as interesting as the remarkable temporal phasing between the total and reconstructed surface temperature anomalies during the latter decades, is the 1960s and 1970s phase interruption. This mismatch can also be partially attributed to natural decadal variability of the GPLLJ, for although PC2 is nearly zero at this time (negligible surface temperature influence), the negative phase of PC5, related to the cold phase of the AMO, promotes a positive meridional wind anomaly over the central United States and a warming contribution (Fig. 2), opposing the negative observed total surface temperature anomaly and creating the mismatch. Despite the warming contribution by PC 5, it is not sufficient to reverse the sign of the total surface temperature anomaly, suggesting a mechanism outside of the direct influence of the PDO and AMO natural variability modes is in play, either as yet unknown low frequency variability or of anthropogenic origin.

4. Discussion

Variability of summertime surface temperature is analyzed to illuminate the role of GPLLJ variability and its connectivity to large-scale decadal variability modes in the development and maintenance of the central U.S. warming hole. The influence of remote SST variability on the central U.S. warming hole is broadly consistent with previous modeling studies (Kunkel et al. 2006, Robinson et al. 2002; Wang et al. 2009). However, the pivotal role that GPLLJ variability plays in linking the hemispheric-wide SST variability (through the AMO and PDO) to the regional warming hole is an expanded and clarified perspective.

It is found that while the first 5 modes of GPLLJ variability contribute substantially to the surface temperature fluctuations over North America, it is GPLLJ modes 2 and 5 that demonstrate the strongest link to the temporal and spatial variations of the recently observed summertime warming hole over the Great Plains. This linkage evidently occurs as a result of the connectivity of GPLLJ modes 2 and 5 to the PDO and AMO respectively. Furthermore, the GPLLJ contributions to central U.S. cooling is strongest when the AMO and PDO are in opposing phases, specifically a positive (negative) PDO (AMO), as occurred from the late 1970s to the mid-1990s.

These findings expand upon those of recent studies from the U.S. CLIVAR Drought Working Group. Wang et al. (2009) show that observed changes in SST are largely responsible for U.S. surface temperature trends over the latter half of the twentieth century, with model experiments indicating that Pacific decadal variability (i.e., PDO) is the primary influence in both the regionality and seasonality of these trends, while Atlantic multidecadal variability (i.e., AMO) is most influential during the summer and fall. Weaver et al. (2009b) linked these identical SST patterns to GPLLJ variability and its related precipitation impacts, finding that warm Pacific and cold Atlantic decadal SST variability strengthens the GPLLJ induced moisture flux convergence leading to increased precipitation over the Great Plains, which during summer leads to surface cooling.

This feature is evident in the spatial pattern of GPLLJ mode 2, however, the mode 5 pattern shows no such GPLLJ strengthening, and instead produces a weakened GPLLJ over the central Great Plains. At first glance it may seem counterintuitive that a highly localized negative GPLLJ anomaly will promote cooling over the central United States during summer, since northerly GPLLJ anomalies typically induce low-level divergence and drying, and thermal advection is likely to be marginal given the weak summertime meridional temperature gradient. However, the location of precipitation induced surface cooling is dependent upon the location of strongest moisture convergence. In the case of GPLLJ mode 5 the northerly anomaly imposed on the terminus region of the mean GPLLJ actually produces stronger moisture convergence in the jet exit region, which enhances precipitation and consequent cooling.

The finding that GPLLJ variability provides a strong link between the large scale natural decadal variability modes and the warming hole is noteworthy in furthering our understanding of regional manifestations of decadal climate variability. However, given that these modes (i.e., PDO–AMO) exert their influence on decadal time scales means that externally forced influences cannot be ruled out since anthropogenic global warming may be modulating the temporal phasing of the underlying decadal modes themselves. Meehl et al. (2009) show that the mid-1970s Pacific climate shift may have occurred a decade earlier in the absence of anthropogenically induced climate forcing. This delayed Pacific warming, as evident in the negative PDO index during the 1960s and early 1970s (Fig. 5), may partially explain the cause for the phase interruption between the GPLLJ reconstructed surface temperature anomaly time series and the total surface temperature anomaly time series during this same time period. Had the Pacific shift occurred earlier it is likely that GPLLJ mode 2 would have become positive sooner, strengthening the GPLLJ cooling contribution and limiting the temporal extent of the phase interruption.

Characteristics such as these have significant implications for efforts aimed at regional decadal climate prediction. The convolving of internal and external forcing of the climate system is a significant challenge in this regard (Solomon et al. 2011). Even if the internally and externally generated large-scale influences were to be cleanly separable and understood on decadal time scales it is unlikely that regional transfer mechanisms such as the GPLLJ and its surface temperature impacts would be adequately represented. Significant climate model biases exist in summer and are particularly acute over the Great Plains, arising from unrealistic partitioning of the water and energy cycles over the central United States (Ruiz-Barradas and Nigam 2006, 2010) and from inadequate geographic placement of the GPLLJ and its role in the regional water and energy cycles (Ghan et al. 1996; Weaver et al. 2009a,b). Any significant advances in regional decadal climate prediction will need to account for these biases and include improvements in the simulation of these critical regional mechanisms in addition to untangling the large-scale natural and anthropogenic forcing impacts.

Acknowledgments

The author wishes to thank Drs. Hui Wang and Arun Kumar for reviewing an early version of the manuscript, Dr. Kerry Cook for her editorial guidance, and three anonymous reviewers whose suggestions were instrumental to improving the quality of the manuscript.

REFERENCES

REFERENCES
Fan
,
Y.
, and
H.
van den Dool
,
2008
:
A global monthly land surface temperature analysis for 1948-present
.
J. Geophys. Res.
,
113
,
DO1103
,
doi:10.1029/2007JD008470
.
Ghan
,
S. J.
,
X.
Bian
, and
L.
Corsetti
,
1996
:
Simulation of the Great Plains low-level jet and associated clouds by general circulation models
.
Mon. Wea. Rev.
,
124
,
1388
1408
.
Higgins
,
R. W.
,
E. S.
Yarosh
,
J. E.
Janowiak
, and
K. C.
Mo
,
1997
:
Influence of the Great Plains low-level jet on summertime precipitation and moisture transport over the central United States
.
J. Climate
,
10
,
481
507
.
Kalnay
,
E.
, and
Coauthors
,
1996
:
The NCEP/NCAR 40-Year Reanalysis Project
.
Bull. Amer. Meteor. Soc.
,
77
,
437
471
.
Kunkel
,
K. E.
,
X.-Z.
Liang
,
J.
Zhu
, and
Y.
Lin
,
2006
:
Can GCMs simulate the twentieth-century “warming hole” in the central United States
.
J. Climate
,
19
,
4137
4153
.
Meehl
,
G. A.
,
A.
Hu
, and
B. D.
Santer
,
2009
:
The mid-1970s climate shift in the Pacific and the relative roles of forced versus inherent decadal variability
.
J. Climate
,
22
,
780
792
.
Robinson
,
W.
,
R.
Reudy
, and
J.
Hansen
,
2002
:
General circulation model simulations of recent cooling in teh east-central United States
.
J. Geophys. Res.
,
107
,
4748
,
doi:10.1029/2001JD001577
.
Ruiz-Barradas
,
A.
, and
S.
Nigam
,
2006
:
IPCC’s twentieth-century climate simulations: Varied representations of North American hydroclimate variability
.
J. Climate
,
19
,
4041
4058
.
Ruiz-Barradas
,
A.
, and
S.
Nigam
,
2010
:
Great Plains precipitation and its SST links in twentieth century climate simulations, and twenty-first- and twenty-second-century climate projections
.
J. Climate
,
23
,
6409
6429
.
Solomon
,
A.
, and
Coauthors
,
2011
:
Distinguishing the roles of natural and anthropogenically forced decadal climate variability: Implications for prediction
.
Bull. Amer. Meteor. Soc.
,
92
,
141
156
.
Trenberth
,
K. E.
, and
Coauthors
,
2007
:
Observations: Surface and atmospheric climate change. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 235–336
.
Wang
,
H.
,
S.
Schubert
,
M.
Suarez
,
J.
Chen
,
M.
Hoerling
,
A.
Kumar
, and
P.
Pegion
,
2009
:
Attribution of the seasonality and regionality in climate trends over the United States during 1950–2000
.
J. Climate
,
22
,
2571
2590
.
Weaver
,
S. J.
, and
S.
Nigam
,
2008
:
Variability of the Great Plains low-level jet: Large-scale circulation context and hydroclimate impacts
.
J. Climate
,
21
,
1532
1551
.
Weaver
,
S. J.
,
A.
Ruiz-Barradas
, and
S.
Nigam
,
2009a
:
Pentad evolution of the 1988 drought and 1993 flood over the Great Plains: A NARR perspective on the atmospheric and terrestrial water balance
.
J. Climate
,
22
,
5366
5384
.
Weaver
,
S. J.
,
S.
Schubert
, and
H.
Wang
,
2009b
:
Warm season variations in the low-level circulation and precipitation over the central U.S. in observations, AMIP simulations, and idealized SST experiments
.
J. Climate
,
22
,
5401
5420
.
Zhao
,
W.
, and
M. A. K.
Khalil
,
1993
:
The relationship between precipitation and surface temperature over the contiguous United States
.
J. Climate
,
6
,
1232
1236
.