Abstract

The purpose of this study was to quantify dipole events (a drought year followed by a pluvial year) for various spatial scales including the nine Oklahoma climate divisions and the author-defined regions of the U.S. Southern Great Plains (SGP), High Plains (HP), and Northern Great Plains (NGP). Analyses revealed that, on average, over twice as many standard deviation (STDEV) dipoles existed in the latter half of the dataset (1955–2013) relative to the first half (1896–1954), suggesting that dramatic increases in precipitation from one year to the next within the Oklahoma climate divisions are increasing with time. For the larger regions within the Great Plains of the United States, the percent chance of a significant pluvial year following a significant drought year was approximately 25% of the time for the SGP and NGP and approximately 16% of the time for the HP. The STDEV dipole analyses further revealed that the frequency of dipoles was consistent between the first and second half of the dataset for the NGP and HP but was increasing with time in the SGP. The temporal periods of anomalous precipitation during relative pluvial years within the STDEV dipole events were unique for each region whereby October occurred most frequently (70%) within the SGP, September occurred most frequently (60%) within the HP, and May occurred most frequently (62%) within the NGP.

1. Introduction

The Great Plains is a region that is heavily utilized for agriculture, livestock, and crop farming purposes. A report by the U.S. Global Change Research Program stated that 70% of the land area within the Great Plains is used for agriculture while 10% of the land area is designated for wildlife habitat and recreation (Karl et al. 2009). In addition, according to a past report from the U.S. Census Bureau, the population within the Great Plains has nearly doubled from 4.9 million in 1950 to 9.9 million people in 2007 (http://www.census.gov/newsroom/releases/archives/population/cb09-107.html, as referenced in http://www.epa.gov/climatechange/impacts-adaptation/greatplains.html, accessed in 2013). By increasing the population of a region that is highly influenced by agriculture and recreation, the effects of a severe drought are exacerbated and felt by many.

The frequent occurrence of severe drought within the Great Plains of North America has motivated the scientific community to study the causes and impacts of drought across the region (Warrick 1984; Chang and Wallace 1987; Atlas et al. 1993; Chang and Smith 2001; Ji and Peters 2003; Schubert et al. 2004; Basara et al. 2013). While not all causes and mechanisms indicative of drought are fully understood, some important contributors to drought in the Great Plains include extratropical and tropical Pacific and Atlantic Ocean sea surface temperature anomalies, subsidence from a mid-to-upper-level ridge centered over the central United States, Pacific and Atlantic anticyclones, changes in the extent and strength of the Great Plains low-level jet, deficits in soil moisture, and poor farming practices (Schubert et al. 2004; Nigam et al. 2011; Basara et al. 2013). These contributors are often the focus of studies over single major drought events, such as atmospheric dust in the “Dust Bowl” drought of the 1930s (Cook et al. 2008), anomalous sea surface temperatures in the drought of the 1950s (Cook et al. 2011), or the Atlantic multidecadal oscillation during the springs of the Dust Bowl drought in comparison with a period of abundant rainfall during the autumns of the 1980s (Nigam et al. 2011). Whereas most studies emphasize the influence of a particular variable or a set of variables related to drought events, Dong et al. (2011) conducted a comparative study over a year of intense drought in 2006 followed by a year of excessive rainfall in 2007 across the southern Great Plains. A key conclusion of the Dong et al. (2011) study was that a dipole relationship exists between drought and pluvial (abundant rainfall) periods. While Dong et al. (2011) quantified this dipole relationship during 2006–07, the study did not examine previous or subsequent periods to determine the frequency and variability of dipole relationships in the climatological record. However, the occurrence of significant precipitation following drought periods is critical to the recovery of water resources of a region. Therefore, the purpose of this study was to quantify dipole drought and pluvial events within the Great Plains and to create frequency distributions of such events by spatial regions.

2. Method

a. Definitions and data

Two definitions were utilized to examine the relationship and frequency of dipole drought and pluvial events. The first dipole was defined as a year with rainfall that is 10% or more below the annual average followed by a year with rainfall that is 10% or more above the annual average. This dipole was labeled as a significant drought to a significant pluvial, that is, an SD-to-SP dipole. This dipole definition was used to capture dipoles with a year of significantly below average precipitation followed by a year with significantly above average precipitation. The second dipole was defined as 1 standard deviation or more above the mean positive change in annual precipitation and was labeled as a standard deviation (STDEV) dipole. This definition encapsulated dipoles with the largest positive precipitation changes from one year to the next without the requirement of a severe drought or pluvial year within the dipole event. The STDEV dipole analyses were used to examine the temporal spacing of large annual precipitation swings for regions within this study. These dipole cases were analyzed by first calculating the change in annual rainfall between the years in the dataset. Only years with a positive change in annual rainfall from one year to the next were used for the calculation of the standard deviation because large positive increases in annual precipitation were the focus for this study. The sample standard deviation of the positive changes in annual rainfall was then calculated as

 
formula

where “cap” is the change in the annual precipitation. If capi is not positive it is discarded from the calculation, and only includes positive changes when calculating the mean. In this equation, a year labeled i represents the change in annual precipitation between that current year and the previous year. If the change in annual precipitation for a year i from its predecessor year was greater than or equal to 1 standard deviation above the mean positive change in annual precipitation, that is,

 
formula

then those two years were designated as an STDEV dipole.

To complete the analyses, this study utilized rainfall values from the National Climatic Data Center (NCDC, now the National Centers for Environmental Information) nClimDiv dataset for the period spanning 1896–2013. This dataset was used instead of NCDC’s traditional climate division (CD) dataset because the nClimDiv dataset includes 1) a grid-based calculation and incorporates more stations from the pre-1930s era as well as 2) utilizing NCDC’s modern array of quality-control algorithms (http://www.ncdc.noaa.gov/monitoring-references/maps/us-climate-divisions.php, accessed in 2014). Although precipitation estimates in the Parameter–Elevation Regressions on Independent Slopes Model (PRISM; http://www.prism.oregonstate.edu) are useful for large-scale precipitation normals, Schneider and Ford (2013) found that monthly analyses of PRISM precipitation estimates had potentially significant differences when compared with Fort Reno gauge averages in central Oklahoma. Because dipole events in this study were analyzed on a monthly basis, precipitation data from PRISM were not used because of the potential for significant error.

The MATLAB software package was used to process the data and retrieve any dipoles that fit the criteria and thresholds described in the following sections. The analyses spanned the hydrological year, which begins 1 October and ends 30 September of the following year (Nalbantis and Tsakiris 2009). As such, when referencing a particular year in this study, such as 1987, we refer to a year that begins in October of 1986 and ends in September of 1987. This method is also consistent with Dong et al. (2011), who utilized the hydrological year in examining the 2006–07 dipole event.

b. Spatial scales and averaging methods

Little exists in the literature as to what exactly defines the geographic boundaries for the Great Plains region. For example, Rossum and Lavin (2000) noted that “no single, clear boundary exists no matter how detailed the study, or accurate the measurements.” However, the Great Plains region is characterized by a large west–east-oriented precipitation gradient (Borchert 1950 and Fig. 1). As such, CDs that captured this precipitation gradient anywhere from the east of the Rocky Mountains to the west of the Mississippi River (excluding the states of Louisiana, Arkansas, Missouri, Iowa, and Minnesota) constituted the Great Plains in this study. Furthermore, the authors used the CDs to subdivide the Great Plains into the Southern Great Plains (SGP), High Plains (HP), and Northern Great Plains (NGP) study regions, as shown in Fig. 2. It is important to note that only CDs 1–4 were chosen from Texas for the SGP region. CDs farther south were not selected because of the presence of the Edwards Plateau in western Texas and Gulf Coastal climates in southeastern Texas. CDs that composed the HP region were chosen to capture the limited precipitation gradient on the western edge of the Great Plains (Fig. 2). CDs within the NGP were chosen to observe how dipole events might vary in more northern latitudes (Fig. 2). Note that CD 4 within South Dakota was excluded from this analysis because it includes the Black Hills region.

Fig. 1.

The 30-yr normal annual precipitation from 1981 to 2010 across the United States (the map was taken from http://www.prism.oregonstate.edu/normals/).

Fig. 1.

The 30-yr normal annual precipitation from 1981 to 2010 across the United States (the map was taken from http://www.prism.oregonstate.edu/normals/).

Fig. 2.

CDs that composed the SGP, HP, and NGP regions.

Fig. 2.

CDs that composed the SGP, HP, and NGP regions.

Utilizing the dipole definitions, a weighted averaging method was applied to the CDs within the SGP, HP, and NGP. The weighted averaging method takes into account the areas of each CD to provide a more accurate depiction of rainfall amounts in larger versus smaller CDs. The weighted averaging method is also used by NCDC for rainfall datasets that span larger regions (http://www.ncdc.noaa.gov/monitoring-references/maps/us-climate-divisions.php, accessed in 2014).

Although very little literature exists on what specifically defines a pluvial, Findell and Delworth (2010) defined a pluvial month as a month with rainfall that is 80% or more above the average for that specific month. Thus, a pluvial month was defined as a month with rainfall that is 80% or more above the average for smaller spatial scales (i.e., the Oklahoma CDs). However, for larger spatial scales, such as the SGP, HP, and NGP, a pluvial month was defined as a month with 40% or more above the average rainfall. This reduction in pluvial threshold for larger spatial scales was necessary because, once larger spatial scales were reached, an artifact of the averaging made large departures from normal difficult to obtain over the larger regions. Thus, 40% was chosen to capture pluvial months that contributed to the pluvial year for larger spatial scales.

When examining dipole events within the Great Plains, various spatial scales were considered. While the study focused on larger regions (SGP, HP, NGP), the nine CDs within Oklahoma were also individually analyzed for dipole events. The Oklahoma CDs were selected for small-scale analysis 1) because Oklahoma has the largest annual precipitation gradient across the state in comparison with all other analyzed states and 2) for comparison with Dong et al. (2011).

3. Results

a. Oklahoma climate divisions

Beginning with the SD-to-SP analysis for the Oklahoma CDs (example of SD-to-SP analysis shown in Fig. 3), the distribution of SD-to-SP dipoles for each CD within Oklahoma during the study period is shown in Fig. 4. A gradient in the number of dipoles existed across Oklahoma, with a general trend of a greater number of SD-to-SP dipoles in the northwestern CDs relative to the southeastern CDs.

Fig. 3.

The SD-to-SP dipoles for Oklahoma CD 2. The green dashed lines represent the annual rainfall for Oklahoma CD 2. The solid blue lines represent the SD-to-SP dipoles identified by the analysis.

Fig. 3.

The SD-to-SP dipoles for Oklahoma CD 2. The green dashed lines represent the annual rainfall for Oklahoma CD 2. The solid blue lines represent the SD-to-SP dipoles identified by the analysis.

Fig. 4.

The SD-to-SP dipole count for the Oklahoma CDs between 1896 and 2013.

Fig. 4.

The SD-to-SP dipole count for the Oklahoma CDs between 1896 and 2013.

To increase the understanding of when the transition from SD to SP occurred from an annual perspective, the specific months that contributed to the overall pluvial years in the SD-to-SP analyses were analyzed, quantified, and plotted in Fig. 5. CDs were grouped by their geographic location, with CDs 1, 4, and 7 composing the western CDs, CDs 2, 5, and 8 composing the central CDs, and CDs 3, 6, and 9 composing the eastern CDs. Such a grouping was used to analyze pluvial-month relationships between CDs with relatively low annual rainfall (western CDs), median annual rainfall (central CDs), and relatively high annual rainfall (eastern CDs). In addition, a pluvial month for this analysis was defined as a month with rainfall that is 80% or more above the average rainfall for that specific month.

Fig. 5.

Pluvial months in pluvial years for the west (CDs 1, 4, and 7), central (CDs 2, 5, and 8), and east (CDs 3, 6, and 9) Oklahoma CDs in the SD-to-SP dipole analysis between 1896 and 2013.

Fig. 5.

Pluvial months in pluvial years for the west (CDs 1, 4, and 7), central (CDs 2, 5, and 8), and east (CDs 3, 6, and 9) Oklahoma CDs in the SD-to-SP dipole analysis between 1896 and 2013.

The overall results demonstrated that the climatological period of autumn through early winter was critical during SD-to-SP transitions, with November being the most likely month to be classified as a pluvial month (37.8% of the time for western CDs, 36.8% of the time for central CDs, and 33.33% of the time for eastern CDs). For the western CDs, September and December also yielded significant precipitation corresponding to pluvial months in over 30% of the SD-to-SP transition periods. In contrast, July and August were minimum values for the western CDs, and the two summer months rarely contributed as pluvial months.

The SD-to-SP transition periods within the central CDs revealed distinct local maxima during November and June, with over 40% of the pluvial years experiencing a June pluvial month and over 35% of the pluvial years experiencing a November pluvial month. However, the transition from SD to SP rarely included July and August, which is a stark contrast when compared with June.

For the eastern CDs, the trend in pluvial months was similar to that for the central CDs, with November and June being preferred months for significantly above average rainfall. In addition, July and August were rarely classified as pluvial months and were not significant contributors to the SD-to-SP transition periods. Unlike the western and central CDs, October seldom contributed as a pluvial month for the pluvial years in the SD-to-SP dipole transition periods for the eastern CDs.

Although this analysis demonstrated that periods of significant precipitation and pluvial periods can occur following a significant drought period, this is not always the case. Thus, the frequency of SD-to-SP dipoles formed during all significant drought periods for the Oklahoma CDs was analyzed (Fig. 6). The analysis revealed a gradient oriented southeast to northwest across Oklahoma, similar to the CD SD-to-SP dipole count. Further, Fig. 6 displays the percent chance of SD-to-SP dipole events occurring during periods of significant drought; CD 8 yielded the highest percentage, with SD-to-SP dipoles occurring approximately 35% of the time during periods of significant drought, whereas CD 9 had the lowest percentage, with SD-to-SP dipoles occurring approximately 19% of the time during periods of significant drought. Additionally, CDs 1, 2, and 8 had a significant pluvial year follow a significant drought year for approximately one out of every three significant drought years.

Fig. 6.

The percent of SD-to-SP dipoles occurring from a significant drought year in the Oklahoma CDs between 1896 and 2013. This figure represents the percent of significant pluvial years following a significant drought year for all significant drought years between 1896 and 2013.

Fig. 6.

The percent of SD-to-SP dipoles occurring from a significant drought year in the Oklahoma CDs between 1896 and 2013. This figure represents the percent of significant pluvial years following a significant drought year for all significant drought years between 1896 and 2013.

In addition to the SD-to-SP dipole analysis, an STDEV dipole analysis was also completed for the Oklahoma CDs. The STDEV dipoles were compared between the first half and second half of the dataset, with the first half comprising the years from 1896 through 1954 and the second half comprising the years between 1955 and 2013. The results shown in Fig. 7 revealed that anywhere from two to four more STDEV dipole events existed for all of the Oklahoma CDs for 1955–2013 relative to 1896–1954.

Fig. 7.

STDEV dipole count for the Oklahoma CDs for the years between 1896–1954 and 1955–2013.

Fig. 7.

STDEV dipole count for the Oklahoma CDs for the years between 1896–1954 and 1955–2013.

b. The Southern Great Plains, High Plains, and Northern Great Plains study regions

The method applied in section 3a was applied to the large subregions within the Great Plains, and a count of SD-to-SP dipoles for each region was produced (Fig. 8). For regions falling within the significant west–east-oriented precipitation gradient (SGP and NGP), numerous SD-to-SP dipoles occurred over the subregions, with the SGP having nine and the NGP having eight. Conversely, the HP region, a region that is characterized by a reduced precipitation gradient, had only four SD-to-SP dipoles. Furthermore, the likelihood of an SD-to-SP dipole “initiating” during a significant drought year in these three subregions was approximately one out of every four within the SGP and NGP regions, whereas approximately one SD-to-SP dipole formed out of every six significant drought years within the HP region (Fig. 9).

Fig. 8.

As in Fig. 4, but for the SGP, HP, and the NGP.

Fig. 8.

As in Fig. 4, but for the SGP, HP, and the NGP.

Fig. 9.

As in Fig. 6, but for the SGP, HP, and the NGP.

Fig. 9.

As in Fig. 6, but for the SGP, HP, and the NGP.

Of particular note, the STDEV dipole analysis for the SGP subregion depicted in Fig. 10 displays a trend of increasing STDEV dipole events with time. For example, only three events were noted between 1896 and 1954, with the average temporal spacing between the three dipoles being approximately 17 yr. However, seven events occurred between 1955 and 2013, with an average temporal spacing of 9 yr, with the temporal spacing between the most recent STDEV dipole events in 2006–07 and 2011–12 being only 5 yr. The monthly periods in the SGP subregion most likely to yield significantly above normal precipitation (pluvial months) during STDEV dipole events are shown in Fig. 11. As previously stated, a pluvial month for this analysis was defined as a month with rainfall that is 40% or more above the average for that specific month. In particular, the autumn months (especially October) and the months of late winter represent the temporal window during which the transition to significant precipitation is most likely. Conversely, the summer months of July and August yielded very few significantly above normal precipitation periods during the dry–wet transition period.

Fig. 10.

STDEV dipoles for the SGP. The green dashed lines represent the annual rainfall for the SGP. The solid blue lines represent the STDEV dipoles identified by the analysis.

Fig. 10.

STDEV dipoles for the SGP. The green dashed lines represent the annual rainfall for the SGP. The solid blue lines represent the STDEV dipoles identified by the analysis.

Fig. 11.

Pluvial months in relative pluvial years for the SGP, HP, and NGP in the STDEV dipole analysis between 1896 and 2013.

Fig. 11.

Pluvial months in relative pluvial years for the SGP, HP, and NGP in the STDEV dipole analysis between 1896 and 2013.

For the NGP, the STDEV dipole analysis revealed a consistent number of STDEV dipoles between the first half and the second half of the dataset, with six STDEV dipoles occurring in the NGP from 1896 through 1954 and seven occurring between 1955 and 2013 (Fig. 12). The analysis of pluvial months that contributed to the pluvial year for the STDEV dipole analysis within the NGP revealed that May was a critical month and occurred as a significant precipitation anomaly during 62% of the STDEV dipole events for the subregion. Months that also appeared frequently included the adjacent months of April, June, and July, which appeared as pluvial months in approximately 31% of pluvial years in STDEV NGP dipole events.

Fig. 12.

As in Fig. 10, but for the NGP.

Fig. 12.

As in Fig. 10, but for the NGP.

For the HP, the STDEV dipole analysis revealed five STDEV dipoles from 1896 through 1954 and five between 1955 and 2013 (Fig. 13). Pluvial months that contributed to the pluvial year for the STDEV dipole analysis revealed that September appeared as a pluvial month more than any other month in the HP, with September occurring as a pluvial month approximately 60% of the time in STDEV HP dipole pluvial years. In addition, the period spanning April–June was a secondary peak and included monthly percentages between 30% and 40% during the STDEV transition periods.

Fig. 13.

As in Fig. 10, but for the HP.

Fig. 13.

As in Fig. 10, but for the HP.

4. Discussion

Unique patterns for dipoles and for pluvial months were quantified across varying spatial scales within the Great Plains region. For comparison with the results of Dong et al. (2011), the initial analysis focused on the Oklahoma CDs. The results of the SD-to-SP dipole analysis revealed an increase in the number of SD-to-SP dipoles from the southeast to northwest (Fig. 4). It is reasonable to conclude that a higher number of SD-to-SP dipoles implies a higher variability in rainfall. Taking this into account, it is worth noting that CDs with lower annual rainfall [CDs 1, 2, and 4: ~51–76 cm (20–30 in.) yr−1] in the northwest had a higher frequency in SD-to-SP dipoles while CDs with higher annual rainfall [CDs 6, 8, and 9: ~89–127 cm (35–50 in.) yr−1] in the southeast had a lower frequency in SD-to-SP dipoles. Furthermore, it is expected that CDs with lower annual rainfall exhibit drier soils than CDs with higher annual rainfall. Likewise, a study by Schubert et al. (2008) found that “dryer [sic] soils associated with drought conditions tend to promote greater evaporation variability and thus greater precipitation variability than do the wetter soils associated with pluvial conditions.” Although Schubert et al. (2008) were discussing increased precipitation variability in relation to drought conditions, the concept of drier soils experiencing greater precipitation variability than wetter soils is a plausible explanation as to why CDs with lower annual rainfall exhibited a higher number of SD-to-SP dipoles relative to CDs with higher annual rainfall.

For the central and eastern CDs within Oklahoma, June was found to be a prominent pluvial month within pluvial years for the SD-to-SP dipole analysis (Fig. 5). Although the period spanning from March through June is typically the climatological peak for precipitation (Fig. 14; approximately 44% of annual rainfall falls between March and June for the central CDs and approximately 42% for the eastern CDs), the results suggest that if the wet season is extended through June with abnormally significant precipitation at the end of the wet season, the odds of experiencing a pluvial year are favorable. As an example, approximately 19 cm (7.4 in.) or more of precipitation is needed to fall for June to be classified as a pluvial month for the central CDs and 21 cm (8.3 in.) or more is needed for such to be true for the eastern CDs. Barandiaran et al. (2013) found that extreme precipitation events in June over eastern Oklahoma and northeastern Texas were related, in part, to the increased strength of the Great Plains low-level jet. Thus, the fact that June was a prominent pluvial month for both the central and eastern Oklahoma CDs could be due, in part, to the strength of the Great Plains low-level jet during June as found by Barandiaran et al. (2013). However, further research utilizing past reanalyses is needed to confirm this relationship.

Fig. 14.

Mean monthly precipitation for the west (CDs 1, 4, and 7), central (CDs 2, 5, and 8), and east (CDs 3, 6, and 9) Oklahoma CDs between 1901 and 2000.

Fig. 14.

Mean monthly precipitation for the west (CDs 1, 4, and 7), central (CDs 2, 5, and 8), and east (CDs 3, 6, and 9) Oklahoma CDs between 1901 and 2000.

Statistically, the STDEV dipoles represented approximately the highest 16% in positive annual rainfall changes from one year to the next. Furthermore, the STDEV dipole analyses revealed where large positive precipitation transitions occurred from one year to the next regardless of 1) whether drought conditions persisted or 2) whether the precipitation was sufficient to reach nondrought conditions. Although the NGP and HP STDEV dipole analyses revealed a consistent number of STDEV dipoles between the first half (1896–1954) and second half (1955–2013) of the dataset (Figs. 12 and 13), the SGP STDEV dipole analysis, similar to the Oklahoma CD analysis, exhibited an increase in STDEV dipoles over time (Fig. 10), which may be attributed to the expected precipitation extremes associated with climate change that have been noted for the region (Kharin et al. 2007; O’Gorman and Schneider 2009).

Each of the subregions within the Great Plains (SGP, HP, and NGP) exhibited a unique annual distribution for the occurrence of pluvial months during pluvial-year transitions during STDEV dipole events (Fig. 11). For the SGP, October normally receives 6.50 cm, or 2.56 in. (an average amount of precipitation when compared with the other months throughout the year; Fig. 15). However, for the STDEV dipole analysis, October appeared most frequently as a pluvial month (70% of the time). As precipitation impacts due to El Niño–Southern Oscillation (ENSO) affect the United States primarily during the cold season, beginning in the month of October (https://www.climate.gov/news-features/blogs/enso/united-states-el-niño-impacts-0, accessed in 2015), the abundance of precipitation during October might suggest an ENSO-related influence. For example, Garbrecht and Rossel (2002) found that ENSO affected precipitation within the Great Plains by influencing planetary waves and consequently, Pacific storm tracks during the cold season, which coincides with the observed peak in pluvial months during STDEV transitions. However, the magnitude of ENSO’s influence on Great Plains precipitation is still uncertain, as Ropelewski and Halpert (1986) found no clear ENSO-related precipitation response within the southern Great Plains from 1875 to 1980. For this study, the pluvial years were compared with past ENSO data from Florida State University’s Center for Ocean-Atmospheric Prediction Studies, which includes records dating back to 1868 (http://coaps.fsu.edu/jma, accessed in 2014). Of the 10 SGP pluvial years noted in the STDEV analysis (Fig. 10), 4 corresponded to years classified as ENSO, 5 corresponded to years classified as neutral, and 1 was classified as La Niña. Because equally as many neutral years as ENSO years corresponded to pluvial years within this study, no conclusive statement can be made concerning ENSO’s influence on the frequency of the autumn–winter pluvial months. Although Garbrecht and Rossel (2002) noted that ENSO was responsible for ending a number of major drought events within the Great Plains during the twentieth century, the magnitude of ENSO’s influence on pluvial years within this study is inconclusive.

Fig. 15.

As in Fig. 14, but for the SGP, NGP, and HP.

Fig. 15.

As in Fig. 14, but for the SGP, NGP, and HP.

For the HP, September normally receives 3.50 cm, or 1.38 in. (an average amount of precipitation when compared with other months throughout the year; Fig. 15). For the STDEV dipole analysis, however, September occurred most frequently as a pluvial month (60% of the time). As previously mentioned, Ropelewski and Halpert (1986) found that ENSO was not a reliable discriminator of precipitation anomalies within the HP region from analyses spanning from 1875 to 1980. To compare with Ropelewski and Halpert (1986), past ENSO data were compared with HP pluvial years, as previously performed for the SGP region. It was found that of the 10 HP pluvial years noted in the STDEV analysis, 2 corresponded to years classified as ENSO, 4 corresponded to years classified as neutral, and 4 corresponded to years classified as La Niña. This further suggests that ENSO is not a reliable discriminator of precipitation anomalies within the HP, as previously stated by Ropelewski and Halpert (1986), and, in particular, of dipole anomalies.

For the NGP, May occurred most frequently as a pluvial month (approximately 62% of the time), and the overall period spanning April–July represented critical periods during which pluvial months occurred during STDEV transition periods. May has climatologically the second highest amount of precipitation within the NGP (Fig. 15). This suggests that receiving considerably above average rainfall in a month that already receives abundant rainfall may result in a large positive increase in precipitation from one year to the next. Furthermore, Barandiaran et al. (2013) found that synoptic features including upper-level lift associated with the jet stream and speed convergence at the exit region of the Great Plains low-level jet provided increased precipitation in the NGP during May.

As previously mentioned, a higher SD-to-SP dipole count implies a higher variability in annual rainfall and a lower SD-to-SP dipole count implies a lower variability in annual rainfall. This study demonstrated that the SGP and NGP had relatively high SD-to-SP dipole counts (nine for SGP and eight for NGP) while the HP had only four SD-to-SP dipoles, suggesting that the annual rainfall in the HP is characterized by fewer precipitation extremes in comparison with the SGP and NGP (Fig. 8). More specifically, the HP is characterized by low annual rainfall [~38 cm (15 in.) yr−1] and a reduced precipitation gradient, the NGP is characterized by reduced annual rainfall [~51 cm (20 in.) yr−1] and a large precipitation gradient, and the SGP is characterized by high annual rainfall [~76 cm (30 in.) yr−1] and a large precipitation gradient. As such, within the bounds of this study, it appears that large geographical areas with large precipitation gradients will likely yield a higher frequency in SD-to-SP dipoles and likewise an increase in annual rainfall variability relative to large geographical regions with small precipitation gradients. Furthermore, SD-to-SP dipoles were initiated from a significant drought year more frequently in the SGP and NGP relative to the HP (one out of every four significant drought years for the SGP and NGP, and one out of every six significant drought years for the HP; Fig. 9). This indicates that within the SGP and NGP the percent chance of a significant pluvial year immediately following a significant drought year is approximately 25%, with the HP less likely (16%) to have a significant pluvial year following a significant drought year.

5. Conclusions

Understanding how the Great Plains recovers from significant drought events is critical to the complex ecosystem of the region. The results of this study demonstrate that, in many cases, significant positive anomalies of precipitation (pluvials) occur immediately after or during significant drought events. For the Oklahoma CDs, it was found that a significant pluvial year immediately following a significant drought year occurred between 19% and 35% of the time for all significant drought years, with the gradient increasing from the southeastern CDs to the northwestern CDs. For the pluvial months within the pluvial years, June was a prominent pluvial month for the central and eastern Oklahoma CDs, suggesting that an extension of the wet season through June increases the likelihood of a significant pluvial year following a significant drought year. The STDEV dipole analysis revealed that considerably more STDEV dipoles existed in the latter half of the dataset (1955–2013) relative to the first half (1896–1954), suggesting that dramatic increases in precipitation from one year to the next within the Oklahoma CDs are increasing with time. For the larger regions within the Great Plains, the percent chance of a significant pluvial year following a significant drought year was approximately 25% of the time for the SGP and NGP and approximately 16% of the time for the HP. The STDEV dipole analysis revealed that the number of STDEV dipoles was consistent between the first and second half of the dataset for the NGP and HP but was increasing with time in the SGP. For the pluvial months within the pluvial years for the STDEV dipoles, each region had a unique, prominent pluvial month, with October occurring most frequently (70% of the time) within the SGP, September occurring most frequently (60% of the time) within the HP, and May occurring most frequently (62% of the time) within the NGP. Although the results within this study unveiled much about drought-to-pluvial dipoles, the fundamental environmental processes contributing to the abrupt change in precipitation on scales that range from local to global are poorly understood and require additional research.

Acknowledgments

The authors thank Gary McManus for his helpful suggestions concerning analysis regions. This work was supported, in part, by the NOAA Climate Program Office’s Sectoral Applications Research Program (SARP) Grant NA130AR4310122 and the Agriculture and Food Research Initiative Competitive Grant 2012-02355 from the USDA National Institute of Food and Agriculture.

REFERENCES

REFERENCES
Atlas
,
R.
,
N.
Wolfson
, and
J.
Terry
,
1993
:
The effect of SST and soil moisture anomalies on GLA model simulations of the 1988 U.S. summer drought
.
J. Climate
,
6
,
2034
2048
, doi:.
Barandiaran
,
D.
,
S. Y.
Wang
, and
K.
Hilburn
,
2013
:
Observed trends in the Great Plains low-level jet and associated precipitation changes in relation to recent droughts
.
Geophys. Res. Lett.
,
40
,
6247
6251
, doi:.
Basara
,
J. B.
,
C. M.
Peirano
,
J. E.
Tate
,
P. J.
Brown
,
J. D.
Hoey
, and
B. R.
Smith
,
2013
:
Drought and associated impacts in the Great Plains of the United States—A review
.
Int. J. Geosci.
,
4
,
72
81
, doi:.
Borchert
,
J. R.
,
1950
:
The climate of the central North American grassland
.
Ann. Assoc. Amer. Geogr.
,
40
,
1
39
, doi:.
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
, doi:.
Chang
,
F.-C.
, and
E. A.
Smith
,
2001
:
Hydrological and dynamical characteristics of summertime droughts over the U.S. Great Plains
.
J. Climate
,
14
,
2296
2316
, doi:.
Cook
,
B. I.
,
R. L.
Miller
, and
R.
Seager
,
2008
:
Dust and sea surface temperature forcing of the 1930s “Dust Bowl” drought
.
Geophys. Res. Lett.
,
35
, L08710, doi:.
Cook
,
B. I.
,
R.
Seager
, and
R. L.
Miller
,
2011
:
Atmospheric circulation anomalies during two persistent North American droughts: 1932–1939 and 1948–1957
.
Climate Dyn.
,
36
,
2339
2355
, doi:.
Dong
,
X.
, and Coauthors
,
2011
:
Investigation of the 2006 drought and 2007 flood extremes at the southern Great Plains through an integrative analysis of observations
.
J. Geophys. Res.
,
116
, D0324, doi:.
Findell
,
K. L.
, and
T. L.
Delworth
,
2010
:
Impacts of common sea surface temperature anomalies on global drought and pluvial frequency
.
J. Climate
,
23
,
485
503
, doi:.
Garbrecht
,
J. D.
, and
F. E.
Rossel
,
2002
:
Decade-scale precipitation increase in Great Plains at end of 20th century
.
J. Hydrol. Eng.
,
7
,
64
75
, doi:.
Ji
,
L.
, and
A. J.
Peters
,
2003
:
Assessing vegetation response to drought in the northern Great Plains using vegetation and drought indices
.
Remote Sens. Environ.
,
87
,
85
98
, doi:.
Karl
,
T. R.
,
J. M.
Melillo
, and
T. C.
Peterson
,
2009
: Global Climate Change Impacts in the United States. Cambridge University Press, 180 pp.
Kharin
,
V. V.
,
F. W.
Zwiers
,
X.
Zhang
, and
G. C.
Hegerl
,
2007
:
Changes in temperature and precipitation extremes in the IPCC ensemble of global coupled model simulations
.
J. Climate
,
20
,
1419
1444
, doi:.
Nalbantis
,
I.
, and
G.
Tsakiris
,
2009
:
Assessment of hydrological drought revisited
.
Water Resour. Manage.
,
23
,
881
897
, doi:.
Nigam
,
S.
,
B.
Guan
, and
A.
Ruiz-Barradas
,
2011
:
Key role of the Atlantic multidecadal oscillations in the 20th century drought and wet periods over the Great Plains
.
Geophys. Res. Lett.
,
38
, L16713, doi:.
O’Gorman
,
P. A.
, and
T.
Schneider
,
2009
:
The physical basis for increases in precipitation extremes in simulations of the 21st-century climate change
.
Proc. Natl. Acad. Sci. USA
,
106
,
14 773
14 777
, doi:.
Ropelewski
,
C. F.
, and
M. S.
Halpert
,
1986
:
North America precipitation and temperature patterns associated with the El Niño/Southern Oscillation (ENSO)
.
Mon. Wea. Rev.
,
114
,
2352
2362
, doi:.
Rossum
,
S.
, and
S.
Lavin
,
2000
:
Where are the Great Plains? A cartographic analysis
.
Prof. Geogr.
,
52
,
543
552
, doi:.
Schneider
,
J. M.
, and
D. L.
Ford
,
2013
:
An independent assessment of the monthly PRISM gridded precipitation product in central Oklahoma
.
Atmos. Climate Sci.
,
3
,
249
258
, doi:.
Schubert
,
S. D.
,
M. J.
Suarez
,
P. J.
Pegion
,
R. D.
Koster
, and
J. T.
Bachmeister
,
2004
:
Causes of long-term drought in the U.S. Great Plains
.
J. Climate
,
17
,
485
503
, doi:.
Schubert
,
S. D.
,
M. J.
Suarez
,
P. J.
Pegion
,
R. D.
Koster
, and
J. T.
Bachmeister
,
2008
:
Potential predictability of long-term drought and pluvial conditions in the U.S. Great Plains
.
J. Climate
,
21
,
802
816
, doi:.
Warrick
,
R. A.
,
1984
:
The possible impacts on wheat production of a recurrence of the 1930s drought in the U.S. Great Plains
.
Climatic Change
,
6
,
5
26
, doi:.