Abstract

Annual average and maximum spells with no precipitation in the southern United States are analyzed. In this study, dry spells are defined as consecutive days with no measurable rainfall. The study area includes 70 weather stations in Texas, Oklahoma, Louisiana, Arkansas, Mississippi, and Tennessee. Interarrival times between daily precipitation records for each station provide the data for this analysis. All 70 stations were analyzed from 1950 to 2013. Six stations that each have data for more than 100 years were analyzed for the period from 1908 to 2013. Approximately 25% of stations in the region show significant negative trends through time, indicating that dry spells have become shorter through time at these locations. The strongest geographical indicator for the number of consecutive dry days across this region was longitude. Dry spells tend to have had longer durations at the westernmost stations because of natural climatological controls.

1. Introduction

Concern over extreme weather events in a dynamic climate is increasing (Peterson et al. 2013; Schiermeier 2011; Changnon et al. 1997). The south-central United States is one area of the country that experiences these extremes in various forms, including hurricanes, flooding, tornadoes, severe thunderstorms, droughts, and more (Roberts 2010; Coumou and Rahmstorf 2012). Furthermore, within the United States, the southeastern quadrant of the country, which includes the south-central region, has experienced more billion-dollar weather disasters than any other region of the country (Powell and Keim 2015; NOAA 2013).

This paper focuses on dry spells in the south-central United States, which is a topic that is not often researched, with only a few published papers set in various regions. For example, Groisman and Knight (2008) examined prolonged dry episodes across the conterminous United States during the warm season. They concluded that the mean duration of dry episodes in the eastern United States increased significantly and that the return period of a 1-month-or-longer episode was reduced from 15 years to 6–7 years. McCabe et al. (2010) conducted a similar study that examined daily precipitation in the southwestern United States through an analysis of interarrival times between each precipitation event. They found that a large portion of the variability in dry-event length is attributable to variability of El Niño–Southern Oscillation, especially for water years and cool seasons. An analysis of these interarrival times of rainfall is important for understanding overall ecological impacts of dry spells, as well as fire potential. For example, Lafon and Quiring (2012) show that intra-annual precipitation variability influences fire occurrence more strongly than does total annual precipitation. Last, Liu et al. (2015) examined dry spells in northeastern China and found that the number of dry days is increasing in the summer half of the year and decreasing in the winter half.

Dry spells are clearly related to issues of drought, but droughts are longer-term phenomena with more insidious impacts because of their slow onset, hard-to-define start times, extended durations, and widespread effects on society. Dry spells may not have the same overall impacts as droughts, but they obviously contribute to drought as well as having an impact on agricultural irrigation schedules (Usman and Reason 2004), assist in driving up ambient air temperatures (Powell and Keim 2015) and thereby driving up energy demands, increase or decrease airborne pollen (Ong et al. 1997) and other airborne pollutants, and increase fire dangers (Beverly and Martell 2005; Lafon and Quiring 2012), among other effects. Dry spells are also much more frequent than droughts, and changes in their frequency and intensity are worth examining on behalf of those who might be affected by multiple days without precipitation in a given year. Therefore, this study examines dry spells across the south-central United States.

2. Scope and data

Similar to what was done in Roberts (2010), the south-central region of the United States is the area of focus for this study, which includes the states of Texas, Oklahoma, Louisiana, Arkansas, Mississippi, and Tennessee (Fig. 1). The region displays a wide range of precipitation climates, from the wet and humid regions of the mountains of eastern Tennessee and coastal Louisiana to the desert of western Texas (see Trewartha 1981). Annual precipitation totals range from >1600 mm down to a low of near 200 mm, with a very steep precipitation gradient starting in eastern Texas and eastern Oklahoma and moving westward. Within the region, rainfall is produced by several mechanisms, including thunderstorms, mesoscale convective complexes, frontal systems, and hurricanes.

Fig. 1.

Study area including the six states within the SRCC and SCIPP regions. Asterisks denote stations with records extending back to 1908. Carets denote stations with data ending in 2009. Boxes are placed around the stations used in the long-term analysis.

Fig. 1.

Study area including the six states within the SRCC and SCIPP regions. Asterisks denote stations with records extending back to 1908. Carets denote stations with data ending in 2009. Boxes are placed around the stations used in the long-term analysis.

This region makes up the geographical area overseen by the National Oceanic and Atmospheric Administration (NOAA) Southern Regional Climate Center (SRCC) and the NOAA “Regional Integrated Sciences and Assessments” program that is called the Southern Climate Impacts Planning Program (SCIPP). The analysis examines 70 stations across the region that were selected on the basis of the quality of available data, thereby providing the best spatiotemporal portrayal of both the average and annual maximum dry spells.

In addition, the region, especially in Texas and Oklahoma, has a history of severe aridity and devastating droughts. These droughts include, but are not limited to, the Texas drought in the 1950s (Andreadis et al. 2005), the drought of 1988 (Namias 1991), the drought of 1999–2000 (Sauchyn et al. 2003), the severe drought of 2006 in Louisiana (Leberg et al. 2007), and the most recent drought of 2011 across Texas and Oklahoma (Nielsen-Gammon 2012).

The objectives of this study are to

  1. create time series for the temporal variability of the annual average and annual maximum dry spells at selected stations in the study area,

  2. test for trends in the annual average and annual maximum dry spells at these stations, and

  3. examine the spatial variability of the annual average and annual maximum dry spells across the study area.

Similar to the approach used in Roberts (2010), weather stations used in this analysis are from the National Weather Service Cooperative Observer Program (COOP). The COOP network was initiated in the 1890s to formulate meteorological observations across the United States, mainly for agricultural purposes (Daly et al. 2007). Using the COOP network of weather stations, maps of the region, and analysis of the station history, 70 stations were selected (Roberts 2010). Stations were chosen on the basis of the three conditions offered by Quiring (2009): 1) the station must possess a record of daily precipitation data that begins at least on, or prior to, 1 January 1950, 2) the station must not be within an 80-km proximity to another selected station, and 3) the station must have a data record that is at least 95% complete (5% or fewer missing data). The percentages of missing data were calculated by dividing the number of days with missing records by the total number of days used. In our analysis, missing data end a dry run, which is why it was paramount to include stations with only limited amounts of missing data. The number of days with missing data is 293, 116, 577, 1024, 459, and 589 at Morgan City, Louisiana; Luling, Texas; Brownsville, Tennessee; Ardmore, Oklahoma; Calico Rock, Arkansas; and Natchez, Mississippi; respectively.

For each of the 70 stations, datasets were created of the interarrival periods between days with rainfall, defined as the number of consecutive dry days between rain events, for the period 1950–2013. Measurable rainfall ends a dry run. No other weather-related elements or factors are involved in this study other than runs of consecutive days with less than measurable precipitation. As such, traces of rainfall were considered to be dry days in this analysis. For 24 stations, data exist back to 1908. Six stations (one per state) show comparative results for precipitation records using data from 1908 to the present. To keep consistency in the remaining analysis, 1950 was selected as the first year for all stations. Figure 1 shows stations with records that begin in 1908 (indicated by an asterisk after the station name). All stations include data up to 2013 except for three stations in Louisiana that ended data collection in 2009 (indicated by a caret after the station name). Texas includes 25 stations, Oklahoma includes 7, Louisiana includes 16, Arkansas includes 6, and both Mississippi and Tennessee have 8 stations.

The length of each run of dry days for each year is found using the “R” computer software (Hubbard et al. 2004). All dry-run lengths (in days) over the course of a year are averaged to get an annual average length of the dry spells. Additional manual efforts were used to identify the single annual maximum dry spell for each year.

3. Methods and results

The average and maximum interannual variability are modeled using time series plots at the six stations. Linear-regression lines are included to indicate trends. The final portion of the analysis examines the spatial variability of all 70 stations in the study area. Given the relatively small size of the region, more spatial correlation is expected at stations that are located closer together than at those farther away. The average and maximum dry spells are plotted from west to east to visualize the spatial variability.

a. Visualizing daily precipitation

Temporal variability of dry spells is initially examined by analyzing daily precipitation data using heat maps (i.e., a graphical representation of data in which the individual values contained in a matrix are represented as colors) at one station per state with a record from 1908 to 2013 (Wilkinson and Friendly 2009). Similar to the approach in Roberts (2010), the six stations selected are Luling, Ardmore, Morgan City, Calico Rock, Natchez, and Brownsville. Figure 2 shows precipitation days from 1 January 1908 to 31 December 2013 for these six stations that have more than 100 years of data. Day 1 at the bottom of each graph is 1 January, and 31 December is at the top of the graph. We exclude 29 February (leap year) in the maps for graphic consistency. Years are sequentially noted across the x axis. This figure provides a visual of 38 690 consecutive days of rain/no rain for each station. Blue indicates a precipitation day, gray indicates a dry day, and white indicates a day with missing data. The blocks of gray—or dry spells—shown in these graphs are the driving force behind this study (Roberts 2010). Displaying the daily precipitation in this manner provides a way to visualize long runs of dry days that exist at these stations throughout their entire record. It also demonstrates the differing precipitation-day climates that exist across this very diverse region (Roberts 2010).

Fig. 2.

Daily precipitation days from 1908 to 2013 for (a) Morgan City, (b) Luling, (c) Brownsville, (d) Ardmore, (e) Calico Rock, and (f) Natchez. Blue represents precipitation days, gray represents dry days, and white showcases missing data.

Fig. 2.

Daily precipitation days from 1908 to 2013 for (a) Morgan City, (b) Luling, (c) Brownsville, (d) Ardmore, (e) Calico Rock, and (f) Natchez. Blue represents precipitation days, gray represents dry days, and white showcases missing data.

Figure 2a shows the rain day/dry day climatological behavior for Morgan City (29.68°N, 91.18°W). A prevalent signal beginning in late summer–late September identifies the time of year during which convective thunderstorms are more likely (Roberts 2010). This explains the cluster of precipitation days that seems to appear nearly every year during that time. Figure 2b shows the daily-precipitation climate for Luling (29.67°N, 97.66°W). The summer convective pattern illustrated at Morgan City is not evident here. This is likely because Morgan City is closer to the moisture-laden air near the Gulf of Mexico. Figure 2c depicts data for Brownsville (35.59°N, 89.26°W). There is a pattern of many precipitation days earlier in the year and more dry days in the latter part of the year. Figure 2d shows daily precipitation for Ardmore (34.17°N, 97.13°W), where spring has clusters of precipitation days. Figure 2e shows the rain-day climate for Calico Rock (36.11°N, 92.16°W). The precipitation pattern is fairly uniform throughout the year. Figure 2f shows daily precipitation for Natchez (31.59°N, 91.34°W), where precipitation days peak during the early part of the year and then again in summer (Roberts 2010).

b. Annual average dry spells

The Spearman rank test (SRT) is used to identify trends across the region. The SRT is used on the average and maximum dry spells for all stations beginning in 1950. This method has been used in the literature for data of this type (Hanson et al. 1989; Yin 1993; Keim et al. 1995; Keim 1997, 1999; Groisman et al. 2005). The SRT is used to determine the relationship between the average (and maximum) annual length of consecutive dry days and the year in which they take place. If a dry spell ran over 31 December for one year and into the next, it was broken up into two dry spells, the first ending on 31 December and the next one beginning on 1 January for the following year. This method was utilized to keep the study consistent as an annual analysis (Roberts 2010). To find a valid trend, the test must find a statistically significant association between the data and year. The correlation coefficients ρ produced by the test determine whether a station’s trend is negative or positive, and all trends with significance p ≤ 0.10 are highlighted. All 70 stations are shown in Table 1, using data from 1950 to 2013.

Table 1.

SRT trend information for all 70 stations for annual average dry spells from 1950 to 2013. Stations are listed from west to east across the study area, wrapping by column. Asterisks denote significance at the 90% level.

SRT trend information for all 70 stations for annual average dry spells from 1950 to 2013. Stations are listed from west to east across the study area, wrapping by column. Asterisks denote significance at the 90% level.
SRT trend information for all 70 stations for annual average dry spells from 1950 to 2013. Stations are listed from west to east across the study area, wrapping by column. Asterisks denote significance at the 90% level.

Twenty-one of the 70 stations have significant trends over this time period, and the vast majority have been negative trends, indicating that interannual times have become shorter with time. The period of 1950–2013 began with a regional drought that lasted several years, and therefore the negative trends may be expected. The stations with negative significant trends are somewhat clustered in Louisiana, but all states have multiple sites with a negative trend. The four significant positive trends are somewhat scattered and occur in four different states. The spatial distribution of positive and negative trends in the annual average dry spells over time can be seen in Fig. 3.

Fig. 3.

Spatial distribution of positive-trend and negative-trend relationships in annual average dry spells over time.

Fig. 3.

Spatial distribution of positive-trend and negative-trend relationships in annual average dry spells over time.

Figure 4 shows a spatial-variability plot for the mean annual average dry spells (Roberts 2010). The stations are listed by longitude from west to east. The plot clearly shows an overall decrease in the interarrival time length from west to east across the study area. Moving from west to east across the region, the averages level off beginning with stations in east Texas and Oklahoma, extending eastward. Figure 4 shows each station’s rainfall interarrival time averaged from 1950 to 2013. It does not take into account the interannual variability within each station’s annual average dry-spell lengths. A linear regression shows that the average dry-spell length decreases by 0.319 day (p value = 0.000) for every 1° move eastward. The value for correlation coefficient squared (R2) indicates that roughly 81% of the variability in average dry-spell length can be attributed to geography (moving from the drier southwestern United States to the wetter southeastern United States).

Fig. 4.

Spatial variability of annual average dry-spell-length averages (1950–2013). Stations are listed from west to east.

Fig. 4.

Spatial variability of annual average dry-spell-length averages (1950–2013). Stations are listed from west to east.

The annual average dry-spell lengths through time for each of the six long-term stations are shown in Fig. 5. Ordinary least squares regression lines are plotted on each time series to visualize trends. A downward-sloping (upward sloping) line indicates that the annual average dry-spell lengths have become shorter (longer) in length over time. Table 2 shows the correlation coefficients and corresponding significance for the six stations for the years 1908–2013. As shown, Ardmore has a significant long-term positive trend, which was also found in the short-term dataset. Significant negative trends were found at Natchez and Morgan City, both of which were sites with insignificant results from 1950 to 2013, but these two sites with long-term negative trends fall within the area with the primary cluster of negative trends during the shorter dataset. This result suggests that no bias is added to our model when using the starting point of 1950, regardless of the high-magnitude drought that was occurring in Texas.

Fig. 5.

Interannual variability in annual average dry spells from 1908 to 2013 for (a) Ardmore, (b) Calico Rock, (c) Luling, (d) Brownsville, (e) Natchez, and (f) Morgan City. The regression line is included to indicate the direction of the trend.

Fig. 5.

Interannual variability in annual average dry spells from 1908 to 2013 for (a) Ardmore, (b) Calico Rock, (c) Luling, (d) Brownsville, (e) Natchez, and (f) Morgan City. The regression line is included to indicate the direction of the trend.

Table 2.

SRT coefficients ρ and corresponding significance p at the 90th percentile for one station per state using data from 1908 to 2013. Asterisks denote significance. ADSmax shows the maximum average dry-spell length (days), and ADSmean shows the mean average dry-spell length (days).

SRT coefficients ρ and corresponding significance p at the 90th percentile for one station per state using data from 1908 to 2013. Asterisks denote significance. ADSmax shows the maximum average dry-spell length (days), and ADSmean shows the mean average dry-spell length (days).
SRT coefficients ρ and corresponding significance p at the 90th percentile for one station per state using data from 1908 to 2013. Asterisks denote significance. ADSmax shows the maximum average dry-spell length (days), and ADSmean shows the mean average dry-spell length (days).

c. Annual maximum dry spells

The annual maximum dry spells are also examined for each of the stations. Trends are analyzed using the SRT. Results at all 70 stations are shown in Table 3, using data from 1950 to 2013. Ten of the 70 stations have significant trends, and eight of these are negative, indicating that the annual extreme dry spell has become shorter in duration over time. Six of these eight sites are in Louisiana, concentrated in the south. In western Texas, however, two sites (Balmorhea and Midland) both have positive trends, indicating longer extreme annual dry spells, which is consistent with recent droughts in Texas. The spatial distribution of positive and negative trends in the annual maximum dry spells over time can be seen in Fig. 6.

Table 3.

As in Table 1, but for maximum annual dry spells from 1950 to 2013.

As in Table 1, but for maximum annual dry spells from 1950 to 2013.
As in Table 1, but for maximum annual dry spells from 1950 to 2013.
Fig. 6.

As in Fig. 3, but for annual maximum dry spells.

Fig. 6.

As in Fig. 3, but for annual maximum dry spells.

Figure 7 shows a spatial-variability plot for the average annual maximum dry-spell lengths. Again, there is a declining slope from the westernmost stations to the easternmost stations until the Louisiana–Texas and Oklahoma–Arkansas borders. The values become stable at approximately 20 days around the west-to-east midpoint of the study area at approximately 95°W longitude. Linear regression shows that the average maximum dry-spell length decreases by 1.74 days (p value = 0.000) for every 1° increase in longitude (i.e., moving eastward). The R2 value indicates that roughly 79% of the variability in the average maximum dry-spell length can be attributed to longitude, which serves as a surrogate for moisture availability. Spatial variability suggests that longitude plays a significant role in the magnitude of the average and the maximum dry spells at each station. Each station’s longest annual maximum on record is shown in Fig. 8, and it follows a west-to-east decline in values that is similar to that seen in Figs. 4 and 7 above. The longer annual maximum dry spells are in southwestern Texas, with the longest for the region at 194 days having occurred in Balmorhea in 2011. The shorter annual maximum dry spells are in eastern Tennessee and in southwestern Arkansas, with several locations having their longest dry spell since 1950 at 30–40 days. The shortest maximum dry spell is only 29 days recorded at Kingsport, Tennessee, in 1953.

Fig. 7.

As in Fig. 4, but for annual maximum dry-spell averages.

Fig. 7.

As in Fig. 4, but for annual maximum dry-spell averages.

Fig. 8.

The single longest annual maximum dry spell (days) in each station’s precipitation record from 1950 to 2013.

Fig. 8.

The single longest annual maximum dry spell (days) in each station’s precipitation record from 1950 to 2013.

Figure 9 shows the interannual variability in the maximum dry-spell lengths for each of the six stations with data back to 1908. Ordinary least squares regression lines are plotted on each time series to visualize trends. Table 4 shows the correlation coefficients and significance for the six long-term stations. None of them have positive trends, but Natchez and Morgan City both have significant long-term negative trends in annual maximum dry spells, indicating a reduction in the severity of the most extreme dry spells.

Fig. 9.

As in Fig. 5, but for annual maximum dry spells for (a) Ardmore, (b) Calico Rock, (c) Luling, (d) Brownsville, (e) Natchez, and (f) Morgan City.

Fig. 9.

As in Fig. 5, but for annual maximum dry spells for (a) Ardmore, (b) Calico Rock, (c) Luling, (d) Brownsville, (e) Natchez, and (f) Morgan City.

Table 4.

As in Table 2, but MDSmax shows the highest maximum dry-spell length (days) and MDSmean shows the mean maximum dry-spell length (days).

As in Table 2, but MDSmax shows the highest maximum dry-spell length (days) and MDSmean shows the mean maximum dry-spell length (days).
As in Table 2, but MDSmax shows the highest maximum dry-spell length (days) and MDSmean shows the mean maximum dry-spell length (days).

4. Discussion and concluding remarks

This study examined 70 stations throughout the southern United States and focused on 6 of these stations for a temporally extended analysis of the average and maximum dry-spell lengths. Trend information was provided, as well as information on spatial variability across the study area. With 18 of the 70 stations in this study resulting in significant negative trends at the 90% level, the evidence suggests that the annual average dry spells have decreased in length with time since 1950 at those stations. Correspondingly, only four stations had significant positive trends. This is true, as well, with the eight stations that indicate a significant negative trend (90% level) with the annual maximum dry spells since 1950.

Only two stations had significant positive trends in the annual maximum dry spells. Both of these stations were located in western Texas. The positive trends in western Texas may be related to a shift in the Great Plains low-level jet, which has shifted northward in recent years. This shift has increased rainfall in the northern Great Plains and decreased rain in the southern Great Plains, including Texas and Oklahoma (Barandiaran et al. 2013). Overall, for 25% of the stations in this region, the annual average and maximum dry spells have significantly decreased in length over time, which indicates that these stations are experiencing more rain days in later parts of the record. For the remainder of the region, the decreasing pattern may be related to shifting positions of the Bermuda high complex. A westward shift in the Bermuda high may have increased instability and precipitation across most of the south-central study region (Li et al. 2012; Powell and Keim 2015).

Some interesting results are worth noting after comparing the two trend studies. The only station that shows a significant positive trend in both the annual maximum and average dry-spell lengths is Balmorhea. This result suggests that dry spells have increased in western Texas, which is already the driest area within our study region. The only stations that show a significant negative trend in both the maximum and average annual dry spell lengths are Union City, Oklahoma; Leesville, Louisiana; Jennings, Louisiana; and Donaldsonville, Louisiana. At these stations, dry spells have been significantly decreasing, indicating an increase in overall precipitation days.

The results in McCabe et al. (2010) and Lafon and Quiring (2012) suggest that decreasing dry-length durations may lessen fire occurrence. If the trends identified here continue, these results help to support the idea of decreasing dry-spell length being related to fire occurrence that is found in McCabe et al. (2010) and Lafon and Quiring (2012). A potential future research study would incorporate information about dry spells corresponding to growing seasons, as in that conducted by Groisman and Knight (2008). Dry spells may be more significant during warm seasons as a result of soil moisture being depleted much more rapidly than during cool seasons. It might be beneficial to consider quantifying dry spells during growing seasons to model the effects on various vegetation. Furthermore, parsing the data by seasons might also yield additional information into the dynamics driving some of the temporal patterns. In addition, seasonal analyses sometimes show that a trend in one season may be masking an opposite trend in another, as would have been the case in Liu et al. (2015) had they not examined warm (summer) and cool (winter) seasons.

There is currently great concern over future changes in precipitation patterns of the south-central United States, with increasing summer aridity being projected for the region by some climate models over the next 100 years (Wetherald and Manabe 1995; Keim et al. 2011). While the precipitation climate of the region is very complex, as demonstrated by Trewartha (1981), these results for the south-central United States indicate a regionwide reduction in the length of dry spells, suggesting wetter conditions. This result is somewhat in opposition to what climate models are suggesting for the future, as noted by Keim et al. (2011).

Acknowledgments

The authors acknowledge the Southern Climate Impacts Planning Program at Louisiana State University.

REFERENCES

REFERENCES
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
1000
, 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:.
Beverly
,
J. L.
, and
D. L.
Martell
,
2005
:
Characterizing extreme fire and weather events in the Boreal Shield ecozone of Ontario
.
Agric. For. Meteor.
,
133
,
5
16
, doi:.
Changnon
,
S.
,
D.
Changnon
,
E.
Fosse
,
D.
Hoganson
,
R.
Roth
Sr.
, and
J.
Totsch
,
1997
:
Effects of recent weather extremes on the insurance industry: Major implications for the atmospheric sciences
.
Bull. Amer. Meteor. Soc.
,
78
,
425
435
, doi:.
Coumou
,
D.
, and
S.
Rahmstorf
,
2012
:
A decade of weather extremes
.
Nat. Climate Change
,
2
,
491
496
, doi:.
Daly
,
C.
,
W. P.
Gibson
,
G. H.
Taylor
,
M. K.
Doggett
, and
J. I.
Smith
,
2007
:
Observer bias in daily precipitation measurements at United States Cooperative Network stations
.
Bull. Amer. Meteor. Soc.
,
88
,
899
912
, doi:.
Groisman
,
P. Ya.
, and
R. W.
Knight
,
2008
:
Prolonged dry episodes over the conterminous United States: New tendencies emerging during the last 40 years
.
J. Climate
,
21
,
1850
1862
, doi:.
Groisman
,
P. Ya.
,
R. W.
Knight
,
D. R.
Easterling
,
T. R.
Karl
,
G. C.
Hegerl
, and
V. N.
Razuvaev
,
2005
:
Trends in intense precipitation in the climate record
.
J. Climate
,
18
,
1326
1350
, doi:.
Hanson
,
K.
,
G. A.
Maul
, and
T. R.
Karl
,
1989
:
Are atmospheric “greenhouse” effects apparent in the climate record of the contiguous United States (1895–1987)?
Geophys. Res. Lett.
,
16
,
49
52
, doi:.
Hubbard
,
K. G.
,
A. T.
Degaetano
, and
K. D.
Robbins
,
2004
:
A modern applied climate information system
.
Bull. Amer. Meteor. Soc.
,
85
,
811
812
, doi:.
Keim
,
B. D.
,
1997
:
Preliminary analysis of the temporal patterns of heavy rainfall across the southeastern United States
.
Prof. Geogr.
,
49
,
94
104
, doi:.
Keim
,
B. D.
,
1999
:
Precipitation annual maxima as a measure of change in extreme rainfall magnitudes in the southeastern United States over the past century
.
Southeast. Geogr.
,
39
,
235
245
, doi:.
Keim
,
B. D.
,
G. E.
Faiers
,
R. A.
Muller
,
J. M.
Grymes
III
, and
R. V.
Rohli
,
1995
:
Long-term trends of precipitation and runoff in Louisiana, USA
.
Int. J. Climatol.
,
15
,
531
541
, doi:.
Keim
,
B. D.
,
R.
Fontenot
,
C.
Tebaldi
, and
D.
Shankman
,
2011
:
Hydroclimatology of the U.S. Gulf Coast under global climate change scenarios
.
Phys. Geogr.
,
32
,
561
582
, doi:.
Lafon
,
C. W.
, and
S. M.
Quiring
,
2012
:
Relationships of fire and precipitation regimes in temperate forests of the eastern United States
.
Earth Interact.
,
16
,
1
15
, doi:.
Leberg
,
P. L.
,
M. C.
Green
,
B. A.
Adams
,
K. M.
Purcell
, and
M. C.
Luent
,
2007
:
Response of waterbird colonies in southern Louisiana to recent drought and hurricanes
.
Anim. Conserv.
,
10
,
502
508
, doi:.
Li
,
W.
,
L.
Li
,
M.
Ting
, and
Y.
Liu
,
2012
:
Intensification of Northern Hemisphere subtropical highs in a warming climate
.
Nat. Geosci.
,
5
,
830
834
, doi:.
Liu
,
X. D.
,
B. H.
Liu
,
H.
Henderson
,
M.
Xu
, and
D. W.
Zhou
,
2015
:
Observed changes in dry day frequency and prolonged dry episodes in northeast China
.
Int. J. Climatol.
,
35
,
196
214
, doi:.
McCabe
,
G. J.
,
D. R.
Legates
, and
H. F.
Lins
,
2010
:
Variability and trends in dry day frequency and dry event length in the southwestern United States
.
J. Geophys. Res.
,
115
,
D07108
, doi:.
Namias
,
J.
,
1991
:
Spring and summer 1988 drought over the contiguous United States—Causes and predictions
.
J. Climate
,
4
,
54
65
, doi:.
Nielsen-Gammon
,
J. W.
,
2012
:
The 2011 Texas drought
.
Texas Water J.
,
3
,
59
95
. [Available online at https://journals.tdl.org/twj/index.php/twj/article/view/6463.]
NOAA
,
2013
: NOAA Extended Reconstructed Sea Surface Temperature (SST) v3b. NOAA/Office of Oceanic and Atmospheric Research/Earth System Research Laboratory Physical Sciences Division. [Available online at http://www.esrl.noaa.gov/psd/data/gridded/data.noaa.ersst.html.]
Ong
,
E. K.
,
P. E.
Taylor
, and
R. B.
Knox
,
1997
:
Forecasting the onset of the grass pollen season in Melborne (Australia)
.
Aerobiologia
,
13
,
43
48
, doi:.
Peterson
,
T. C.
, and Coauthors
,
2013
:
Monitoring and understanding changes in heat waves, cold waves, floods, and droughts in the United States: State of knowledge
.
Bull. Amer. Meteor. Soc.
,
94
,
821
834
, doi:.
Powell
,
E. J.
, and
B. D.
Keim
,
2015
:
Trends in daily temperature and precipitation extremes for the southeastern United States: 1948–2012
.
J. Climate
,
28
,
1592
1612
, doi:.
Quiring
,
S. M.
,
2009
:
Developing objective operational definitions for monitoring drought
.
J. Appl. Meteor. Climatol.
,
48
,
1217
1229
, doi:.
Roberts
,
M.
,
2010
: Dry event trends and frequencies in the south central United States. M.S. thesis, Dept. of Geography and Anthropology, Louisiana State University, Baton Rouge, LA, 114 pp. [Available online at http://etd.lsu.edu/docs/available/etd-07082010-021415/unrestricted/RobertsThesis.pdf.]
Sauchyn
,
D. J.
,
J.
Stroich
, and
A.
Beriault
,
2003
:
A paleoclimatic context for the drought of 1999-2001 in the northern Great Plains of North America
.
Geogr. J.
,
169
,
158
167
, doi:.
Schiermeier
,
Q.
,
2011
:
Extreme measures: Can violent hurricanes, floods and droughts be pinned on climate change? Scientists are beginning to say yes
.
Nature
,
477
,
148
149
, doi:.
Trewartha
,
G. T.
,
1981
: The Earth’s Problem Climates. University of Wisconsin Press, 372 pp.
Usman
,
M. T.
, and
C. J. C.
Reason
,
2004
:
Dry spell frequencies and their variability over southern Africa
.
Climate Res.
,
26
,
199
211
, doi:.
Wetherald
,
R. T.
, and
S.
Manabe
,
1995
:
The mechanisms of summer dryness induced by greenhouse warming
.
J. Climate
,
8
,
3096
3108
, doi:.
Wilkinson
,
L.
, and
M.
Friendly
,
2009
:
The history of the cluster heat map
.
Amer. Stat.
,
63
,
179
184
, doi:.
Yin
,
Z. Y.
,
1993
:
Spatial patterns of temporal trends in moisture conditions in the southeastern United States
.
Geogr. Ann.
,
75A
,
1
11
, doi:.