Abstract

Drier future conditions are projected for the arid southwest of North America, increasing the chances of the region experiencing severe and prolonged drought. To examine the mechanisms of decadal variability, 47 global climate model historical simulations performed for phase 5 of the Coupled Model Intercomparison Project (CMIP5) were assessed. On average, the CMIP5 models have higher climatological precipitation over the past century in southwestern North America than current instrumental or reanalysis products. The timing of the winter peak in climatological precipitation over California and Nevada is accurately represented. Models with resolutions coarser than 2° show a larger spread in the location and strength of the North American monsoon ridge and subsequent summer precipitation, in comparison with the higher-resolution models. Less than 20% of decadal variability in wintertime precipitation over California is associated with North Pacific sea surface temperature anomalies, a larger proportion than is associated with the tropical forcing but not sufficient for making decadal drought predictions. North American monsoon precipitation is strongly associated with local land temperatures on interannual-to-decadal time scales attributable to evaporative cooling and radiation changes driven by varying cloud cover. Soil moisture in Texas and Oklahoma in April is shown to be positively correlated with monsoon precipitation for the following summer, indicating a potential source of nonoceanic interseasonal persistence in southwestern North American hydroclimate. To make meaningful decadal predictions in the future, it is likely that forecasting will move away from sea surface temperature–driven anomaly patterns, and focus on land surface processes, which can allow persistence of precipitation anomalies via feedbacks.

1. Introduction

Southwestern North America is projected to undergo drying in the twenty-first century (Seager et al. 2007; Cayan et al. 2010; Christensen and Lettenmaier 2007), potentially increasing the frequency of occurrence and duration of drought-like conditions in the region. Climate records derived from tree core measurements (e.g., Stahle et al. 2007; Cook et al. 2004; Woodhouse et al. 2006) or vegetation growth in lake beds (e.g., Stine 1994) from the past millennia indicate that the southwestern United States is susceptible to severe impacts from multidecade droughts (“megadroughts”; Woodhouse and Overpeck 1998; Meehl and Hu 2006). Although shorter in duration, the Dust Bowl drought in the 1930s, named for the large dust storms caused by the degradation of soil stability resulting from agricultural use of the land (e.g., Schubert et al. 2004a), and the Southwest drought in the 1950s (e.g., Swetnam and Betancourt 1998) also had significant consequences lasting almost a decade. The potential economic and social cost of such intense and sustained events such as these motivates the need to understand the mechanisms for drought variability and persistence, critical for the eventual development of effective forecasting methods.

Decadal (5–10 yr) precipitation variability in southwestern North America is hypothesized to be partly attributable to Pacific or Atlantic Ocean sea surface temperature (SST) anomalies (e.g., Barlow et al. 2001; Hoerling and Kumar 2003; Hoerling et al. 2009; Schubert et al. 2004a, b; McCabe et al. 2004; Feng et al. 2011; Kushnir et al. 2010). Links of regional climate with tropical and extratropical SST anomalies are expected to lead to the potential to predict future drought conditions, as a result of the seasonal-to-multidecadal time scales present in oceanic modes (e.g., Hoerling et al. 2009). A number of modes, such as tropical, northern, and pan-Pacific Ocean SST anomalies, have been identified with varying impacts on regional climate at annual-to-multidecadal time scales (e.g., Barlow et al. 2001; Capotondi and Alexander 2010; Schubert et al. 2004b; Ting and Wang 1997). Specific drought or pluvial episodes during the past century have been previously shown to be associated with tropical Pacific Ocean SST anomalies via forced atmospheric models or composite analysis (Hoerling and Kumar 2003; Schubert et al. 2004a; Trenberth and Guillemot 1996). Cool tropical Pacific Ocean SST anomalies of La Niña periods are associated with positive height anomalies that divert the storm tracks farther north, resulting in reduced precipitation over California in winter (Schubert et al. 2008; Seager et al. 2010). North Pacific Ocean SST anomalies, forced by an atmospheric bridge from tropical heating anomalies (Lau 1997; Deser and Blackmon 1995; Alexander et al. 2002; Pierce 2002; Kumar et al. 2013), may modulate the tropical forcing on interannual-to-decadal time scales (Barlow et al. 2001; Gershunov and Barnett 1998; Gan and Wu 2013; Barsugli and Battisti 1998), influencing the moisture available to the storm tracks reaching western North America or the exact latitude of the polar jet (Cole et al. 2002; Liu et al. 2013; Sheppard et al. 2002; Gershunov and Barnett 1998; Seager et al. 2005; Yu et al. 2007; Yu and Zwiers 2007). Uncertainties remain with regard to the dynamics of how different modes of SST variability can influence interannual-to-multidecadal precipitation over western North America.

The longer persistence of time scales of oceanic anomalies compared to atmospheric fluctuations has drawn much of the focus of decadal predictability; however, the land surface is also an important source of decadal persistence. The interior of the contiguous United States is more likely to recycle local moisture and be influenced by the antecedent land conditions, as large-scale circulation driven by oceanic anomalies is less dominant farther inland over continental areas (Koster et al. 2000, 2004; Conil et al. 2007). The prevalence of seasonal-to-multidecadal drought in western North America suggests the importance of soil moisture in this region in terms of atmospheric feedbacks and the availability of water for local evaporation (Oglesby and Erickson 1989; Eltahir 1998). It has been theorized that the North American monsoon is influenced by the follow-on from winter and spring conditions, and that seasonal precipitation is strongly tied to land surface feedbacks in the region (Gutzler and Preston 1997; Gutzler 2000; Higgins et al. 1998; Higgins and Shi 2000; Matsui et al. 2003; Mo and Paegle 2000; Zhu et al. 2005; Notaro and Zarrin 2011; Small 2001; Entekhabi et al. 1992). Gutzler and Preston (1997) showed that suppressed monsoon precipitation during summer occurred in years with above normal snowpack in the southern Rocky Mountains, for 1961–90. This was thought to be attributable to abundant soil moisture reservoirs from snowmelt increasing the evaporation rate and thermal inertia of the western United States, maintaining low land surface temperatures and affecting the strength of the monsoon ridge and the subsequent precipitation (Gutzler and Preston 1997; Gutzler 2000; Lo and Clark 2002; Matsui et al. 2003). However, this relationship breaks down when the analysis is extended to the entire twentieth-century record (Gutzler 2000; Griffin et al. 2013). The respective roles of land feedbacks, oceanic forcing, and atmospheric internal variability in persisting long-term drought-like conditions therefore constitute an outstanding research question.

Recently, there has been increased interest in the capability of global climate models (GCMs) to simulate decadal conditions and the potential to predict near-term conditions on a global scale (e.g., Meehl et al. 2009; Branstator and Teng 2012; Keenlyside et al. 2008; Goddard et al. 2013; Jia and DelSole 2012). Mechanistic analysis of decadal variability in the coupled model framework requires robust simulations that faithfully simulate the conditions during the instrumental era, over which they can be validated for three-dimensional fields on monthly time scales. A comparison of the recently performed historical simulations from phase 5 of the Coupled Model Intercomparison Project (CMIP5) in section 3 evaluates basic model performance statistics in representing the dominant precipitation variability in southwestern North America. The large ensemble of simulations available through the CMIP5 archive also aids investigation of precipitation anomalies and decadal variability, which can lead to the long-term drought-like conditions seen in the proxy records (Ault et al. 2013a). Section 4 evaluates the consistency of the CMIP5 models in reproducing relationships between the Pacific Ocean and 5-winter average precipitation variability over California, and identifies the most important oceanic regions for driving decadal drought. Because of the relative importance of internal variability, the relationship of decadal variability of precipitation with SST anomalies is not sufficient to be of practical use in making predictions. Section 5 assesses the importance of land feedbacks to the North American summer monsoon precipitation variability, a potential source of nonoceanic intraseasonal persistence in the southwestern North American hydroclimate. Implications for the future of decadal predictability studies are discussed.

2. CMIP5 models and reanalysis products

Phase 5 of the Coupled Model Intercomparison Project (see, e.g., Taylor et al. 2012) involves a large number of international modeling centers providing a suite of experiments to address scientific questions and provide simulations to the Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report (AR5). Historical simulations are available from 20 different modeling centers extending over approximately the past 150 years, using appropriate forcings; the details of the model simulations and expanded names of all models are listed in Table 1. Many of the models have similar components, but different resolutions, land modules, biochemistry components, or physical parameterizations. Monthly mean fields for a single ensemble member are selected per model simulation.

Table 1.

Table of CMIP5 historical simulations. Stars indicate models in the higher-resolution category (<2°). Crosses indicate the models with soil moisture data available through the archive.

Table of CMIP5 historical simulations. Stars indicate models in the higher-resolution category (<2°). Crosses indicate the models with soil moisture data available through the archive.
Table of CMIP5 historical simulations. Stars indicate models in the higher-resolution category (<2°). Crosses indicate the models with soil moisture data available through the archive.

This analysis is focused on the oceanic and land surface temperature, midtropospheric geopotential height (as a representation of the large-scale circulation), and North American precipitation. Meridional and zonal wind fields are also examined for details of the mean circulation during the summer. The Goddard Institute for Space Studies (GISS) Surface Temperature Analysis (GISTEMP) Land–Ocean Temperature Index (LOTI) is available from 1880 (Hansen et al. 1981, 1999, 2001, 2010) and is used to assess the SST anomalies and the skin temperatures at the interface between soil and the atmosphere. Geopotential height, zonal and meridional winds in the midtroposphere, and precipitation measures are available from the 40-yr European Center for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) from 1958 (Uppala et al. 2005), and the National Oceanic and Atmospheric Administration (NOAA) Cooperative Institute for Research in Environmental Sciences (CIRES) Twentieth-Century Reanalysis, version 2 (20CR V2), from 1871 (Compo et al. 2011, 2006; Whitaker et al. 2004).

The data sources used here do suffer from limitations, both in terms of accuracy and temporal extent. The midtropospheric reanalysis fields are likely to suffer from a lack of robust input data prior to the mid-twentieth century. Monthly mean gauge precipitation records from the Global Precipitation Climatology Centre (GPCC; Rudolf et al. 2005, 1994, 2003; Rudolf and Schneider 2005; Beck et al. 2005; Schneider et al. 2014; Becker et al. 2013) from 1931 are used for southwestern North America. GPCC combines gauge data primarily from an international network of national weather services. Precipitation anomalies from around 67 200 stations worldwide are interpolated (more than 8 stations per 2.5° grid point, with durations of 10 or more years) and added to the climatology at that resolution. Reanalysis data based on satellite data, such as the 1979 onward Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP; Xie and Arkin 1997), are not suitable because of the short period of record. The focus here on precipitation, and therefore meteorological drought, is due to the more robust nature of gauge precipitation measurements for the past century than streamflow or soil moisture measurements, as well as the fact that precipitation is available in the archive for all of the CMIP5 historical simulations. Additionally, hydrological variables such as soil moisture or streamflow may not be similarly defined in each model.

3. Precipitation climatology for southwestern North America

Understanding the dynamics driving North American precipitation requires identifying an appropriate study region, as many distinct processes operate in different parts of the continent. One commonly used such region is the Great Basin in the southwestern United States (32°–42°N, 106°–118°W; Meehl and Hu 2006). However, even this region encompasses a range of interannual variability. Great Basin precipitation is dominated by mesoscale convective activity from the south during the summer monsoon season, and large-scale storms crossing the U.S. West Coast in winter. To identify the relationships between precipitation variability in these regions and SST or atmospheric circulation anomalies, and to choose more appropriate study regions, the accuracy of precipitation climatology during the winter and summer is first assessed. Around 50% of the annual precipitation in New Mexico and Arizona falls in a 3-month period during the North American summer monsoon, an important contribution to southwestern North America hydroclimate (Adams and Comrie 1997; Douglas et al. 1993; Hales 1972; Higgins et al. 1997; Vera et al. 2006). In California, however, precipitation is dominated by storm systems arriving from the ocean during the winter months. Therefore, the two regions that isolate the areas of highest interannual variability in southwestern North America are selected for closer consideration (see Fig. 1; henceforth “California region” and “monsoon region”). Each of these regions comprises only a few grid points, as the model resolutions range from 0.5° to 3.75° in the atmosphere, and results are regridded to a common spacing when required for comparison.

Fig. 1.

Map of North America, with U.S. state boundaries marked and the Great Basin (Meehl and Hu 2006) shaded. This includes the states of California, Nevada, Utah, Colorado, Arizona, and New Mexico. The rectangular boxes indicate the regions used in this study. The “California region” encompasses most of California and Nevada (32°–42°N, 114°–122°W), which experience a peak in precipitation in the winter months, and the “monsoon region” includes the northernmost extension of the North American summer monsoon (28°–38°N, 104°–110°W).

Fig. 1.

Map of North America, with U.S. state boundaries marked and the Great Basin (Meehl and Hu 2006) shaded. This includes the states of California, Nevada, Utah, Colorado, Arizona, and New Mexico. The rectangular boxes indicate the regions used in this study. The “California region” encompasses most of California and Nevada (32°–42°N, 114°–122°W), which experience a peak in precipitation in the winter months, and the “monsoon region” includes the northernmost extension of the North American summer monsoon (28°–38°N, 104°–110°W).

Figures 2a and 2b show the mean annual cycle of area averaged precipitation in these two regions. The monthly mean precipitation was averaged over three consecutive months for 12 seasonal means over the entire period of the dataset, which is around 150 years for the CMIP5 historical simulations, and fewer years for the reanalyses (see section 2). It can be seen from the black solid lines in Figs. 2a and 2b that the average of the CMIP5 historical simulations has higher mean precipitation than the reanalysis products. The spread between the individual model simulations, indicated by the gray and colored lines, shows the range of biases in the CMIP5 historical simulations. Differences in California precipitation intensity between the models are likely a result of difficulties in resolving coastal topology; however, all models and reanalysis data simulate the same seasonal cycle with a synchronous peak. Precipitation in the west, over California and Nevada (Fig. 2a) peaks in the winter [December–February (DJF)]. All CMIP5 historical simulations accurately represent this annual cycle, despite a large spread in the magnitude of the midwinter peak in precipitation between models. The model simulations that show higher precipitation on average include IPSL-CM5B-LR and MRI-CGCM3 in spring, and also FGOALS-g2 and GISS-E2 in early winter. The ERA-40 (black, dot dashed) shows consistently lower precipitation than the other reanalysis products (GPCC: black, dashed; 20CR V2: black, dotted) and the CMIP5 historical simulations across both regions, despite showing a similar shape to the time series.

Fig. 2.

Time series of (a),(b) mean and (c),(d) standard deviation of seasonal precipitation, averaged over the (top) California region and (bottom) North American monsoon region. The time series consist of twelve 3-month means. The gray and colored lines indicate each of the CMIP5 historical simulations, which are averaged over the entire period of the run (see Table 1). Individual models are identified with particularly high or low precipitation climatologies or variance. The black solid line is the average of all CMIP5 historical simulations. The observed and reanalysis time series, averaged over their respective time periods, are 70 yr of GPCC gauge data (black, dashed), 130 yr of 20CR V2 (black, dotted), and 40 yr of ERA-40 (black, dot dashed).

Fig. 2.

Time series of (a),(b) mean and (c),(d) standard deviation of seasonal precipitation, averaged over the (top) California region and (bottom) North American monsoon region. The time series consist of twelve 3-month means. The gray and colored lines indicate each of the CMIP5 historical simulations, which are averaged over the entire period of the run (see Table 1). Individual models are identified with particularly high or low precipitation climatologies or variance. The black solid line is the average of all CMIP5 historical simulations. The observed and reanalysis time series, averaged over their respective time periods, are 70 yr of GPCC gauge data (black, dashed), 130 yr of 20CR V2 (black, dotted), and 40 yr of ERA-40 (black, dot dashed).

Many of the CMIP5 historical simulations peak later in the summer than the reanalysis products for the monsoon region, and there is more variation in the shape of the mean precipitation time series (Fig. 2b) as compared to the more robust California region. The highest peaks in precipitation in July–September (JAS) correspond to the GFDL-ESM2G and MIROC4h models, while the Canadian Centre for Climate Modelling and Analysis (CCCma), IPSL, INM, and FGOALS historical simulations show a lack of precipitation in the monsoon region. Although this region does not capture the core Mexican component of the North American monsoon, it identifies a key region of interest to the southwestern U.S. climatology. Differences in the GPCC gauge data from other reanalyses and the models for this region highlight potential issues with assimilating pressure only at the surface or systematic biases in gauges, and the difficulty in accurately representing large-scale precipitation climatology based on measurements of small-scale thunderstorms of the monsoon. Although the climatological precipitation intensity tends to be greater, the CMIP5 historical simulation results have precipitation patterns in better agreement with the ERA-40 or 20CR V2 products, rather than the GPCC gauge data. This suggests the reanalysis precipitation fields are more reliant on the underlying model used in the assimilation, and therefore unlikely to be the best dataset for model validation.

The year-to-year precipitation variability is shown in Figs. 2c and 2d for the standard deviation of the area averaged precipitation, calculated for each 3-month period as in Figs. 2a and 2b. The California region shows high variability in the winter months, with the largest climatological values for the IPSL-CM5B-LR, MRI-CGCM3, and GISS historical simulations. The average variability of the CMIP5 model simulations in the monsoon region shows no change across the 12 seasons as a result of differences in the timing of maximum standard deviation among the simulations (Fig. 2d). For example, FGOALS-s2 peaks in October–December (OND), while INM-CM4.0 shows a minimum in precipitation variability in late summer. GFDL-ESM peaks in summer, with the highest year-to-year variability. The 20CR V2 and ERA-40 datasets also show a peak in the standard deviation in summer that is not seen in the GPCC gauge data, indicating differences between the reanalysis products for the representation of the current climate. The inconsistency of the CMIP5 historical simulations in this region suggests that the models do not accurately capture the spatial and temporal structure of the monsoon precipitation because of differences in the circulation. Therefore some details of the model biases need to be carefully assessed.

Since the CMIP5 models more accurately capture precipitation in the California region than in the monsoon region, details of the spatial structure of the monthly mean precipitation and the associated circulation during the summer months are next examined. Figure 3 compares the spatial patterns of the mean precipitation in the CMIP5 historical simulations during the early summer months, when the monsoon moisture flux moves north from Mexico into the southwestern United States. The fraction of models exceeding a threshold of 3 mm day−1 at each grid point for the monthly mean precipitation in June or July is indicated by the shading. The dashed and solid lines indicate the border of where the GPCC and 20CR V2 precipitation exceeds this same threshold. Note that similar results are seen when thresholds are chosen over a range from 1 to 5 mm day−1 (not shown). The CMIP5 historical simulations were separated into higher and lower atmospheric resolution categories, with the cutoff at around 2° resolution. Almost half of the model simulations fall into the higher-resolution category and are indicated by a star in the first column of Table 1. For June, around three-quarters of the CMIP5 historical simulations show more precipitation than the reanalysis products over northern Mexico, indicated by the northward expansion of the 3 mm day−1 threshold past the GPCC gauge data threshold at 22°N at the coast (Figs. 3a,c). The GPCC gauge data also show moisture flux propagation along the west and the east coast of Mexico in two separate northward extensions at 105° and 95°W, while the higher-resolution models tend to have precipitation confined to the west coast of Mexico, extending north into the United States, and the lower-resolution models have higher precipitation inland over northern Mexico. In July, the higher-resolution CMIP5 historical simulations show a spatial pattern that is more consistent with the GPCC gauge data and each other than the lower-resolution model simulations (Figs. 3b,d). The Sierra Madre Occidental mountain range bordering the west coast of Mexico that focuses the development of thunderstorms in this region (Douglas et al. 1993) is critical for capturing the spatial structure of the summer monsoon precipitation and therefore differing model representations of this feature are likely responsible for much of the intermodel disagreement in Fig. 3. The lower-resolution models are unlikely to capture the thunderstorms, evident in the broad region of high precipitation propagating directly northward, rather than northwest along the west coast of Mexico. The higher-resolution models better represent the topographical features required to capture the higher monsoon precipitation along the west coast of Mexico and have finer grid spacing, which can resolve the spatial structure in the northward propagating moisture flux.

Fig. 3.

Fraction of models that exceed 3 mm day−1 for average (a),(c) June and (b),(d) July precipitation. The thick dashed line indicates where the GPCC observations exceed the same threshold, and the fine dotted line indicates where the 20CR V2 exceeds the same threshold. High-resolution models are indicated by a star in Table 1, and generally have grid spacing finer than 2° resolution.

Fig. 3.

Fraction of models that exceed 3 mm day−1 for average (a),(c) June and (b),(d) July precipitation. The thick dashed line indicates where the GPCC observations exceed the same threshold, and the fine dotted line indicates where the 20CR V2 exceeds the same threshold. High-resolution models are indicated by a star in Table 1, and generally have grid spacing finer than 2° resolution.

Shortcomings in the representation of the summer precipitation may be associated with the timing and the spatial structure of the monsoon circulation. Therefore, the mean circulation at 500 hPa and the location of the subtropical ridge was also examined. Figure 4 shows the distribution of the mean center of anticyclonic circulation over southwestern North America at 500 hPa in the 47 CMIP5 historical simulations in June and July. The center of circulation is determined by the location of the minimum vorticity, which is calculated from the wind fields at 500 hPa, after confirming that a single center is located within the broader region of maximum geopotential height. The strengthening and northward shift of the monsoon ridge is apparent from June to July for most CMIP5 historical simulations and the ERA-40 and 20CR V2 results. There is more spread in the location of the ridge for the lower-resolution model simulations, which are more likely to have a southerly shifted weak circulation in June that tends to be zonally broader (not shown) and may be associated with the broader precipitation spatial structure seen in Fig. 3. The location of the center of circulation in the early summer is important for capturing the reversal of winds across the west coast of Mexico as the subtropical ridge moves north along with the intense solar heating. The clockwise circulation around the high pressure region results in offshore winds along the west coast of Mexico and southeasterly flow along the Gulf of California, channeling moisture into Arizona and the southwestern United States. Models that do not accurately capture this circulation are also likely to lack precipitation north of 30°N. For example, the IPSL and FGOALS historical simulations, which showed climatologically lower precipitation in the monsoon region in Fig. 2b, have a weak or delayed high pressure system at 500 hPa and correspond to the southernmost points in Fig. 4b. Alternatively, some of the higher-resolution CMIP5 historical simulations with more accurate monsoon ridge locations have a strong anticyclonic circulation, resulting in offshore winds across the Gulf of California that will not facilitate an influx of moisture from the Pacific Ocean into the region. The precipitation still tends to lie along the west coast of Mexico for the higher-resolution models, despite inaccuracies in the mean wind direction in the region, suggesting that the accurate topography can still enable local moisture to be deposited as precipitation along the Sierra Madre Occidental range.

Fig. 4.

Center of 500-hPa mean circulation for each CMIP5 model run in (a) June and (b) July. The lighter gray circles indicate the lower-resolution model simulations, and the darker gray circles the higher-resolution model simulations. ERA-40 and 20CR V2 are indicated by the cross and asterisk, respectively. The size of the symbol is proportional to the geopotential height and therefore strength of the monsoon ridge at that location at 500 hPa.

Fig. 4.

Center of 500-hPa mean circulation for each CMIP5 model run in (a) June and (b) July. The lighter gray circles indicate the lower-resolution model simulations, and the darker gray circles the higher-resolution model simulations. ERA-40 and 20CR V2 are indicated by the cross and asterisk, respectively. The size of the symbol is proportional to the geopotential height and therefore strength of the monsoon ridge at that location at 500 hPa.

The mean meridional and zonal winds at 500 hPa in July are compared for all models in Fig. 5 to examine the direction of moisture flow along the Gulf of California. While the CMIP5 historical simulations are typically too coarse in resolution to resolve the details of topology or mesoscale systems along the Gulf of California, Fig. 5 gives a measure of the winds of the larger-scale circulation discussed in Fig. 4. At the mouth of the Gulf of California (23°N, 107°W), indicated by the triangles, the model simulations all show the expected easterly flow (negative zonal wind), although the lower-resolution models show a larger spread in the strength of this wind. There is also a range of a northerly to southerly component to the wind direction at this location, suggestive of the range of the mean location of the monsoon ridge in July in the model simulations and also differences between the two sets of reanalysis data. At the top of the Gulf of California (31°N, 115°W), indicated by the circles in Fig. 5, the model simulations show southerly flow (positive meridional wind), but with a larger spread in the east–west direction. The lower-resolution model simulations are more likely to have a westerly, onshore component resulting from the southward shift of the monsoon ridge. This is consistent with the climatologically lower precipitation in the lower-resolution models in the southwestern United States. Shortcomings in summer precipitation in southwestern North America are therefore likely to be associated with the location or spatial structure of the monsoon circulation.

Fig. 5.

Meridional and zonal wind strength over the Gulf of California for each model run for the July climatology. The triangles indicate the strength of the components of the wind at the mouth of the Gulf of California at 23°N, 107°W. The circles indicate the strength of the components of the wind at the northernmost point of the Gulf at 31°N, 115°W. The darker gray points are the higher resolution model simulations, and the lighter gray points are the lower resolution model simulations. The open symbols are the ERA-40 and 20CR V2.

Fig. 5.

Meridional and zonal wind strength over the Gulf of California for each model run for the July climatology. The triangles indicate the strength of the components of the wind at the mouth of the Gulf of California at 23°N, 107°W. The circles indicate the strength of the components of the wind at the northernmost point of the Gulf at 31°N, 115°W. The darker gray points are the higher resolution model simulations, and the lighter gray points are the lower resolution model simulations. The open symbols are the ERA-40 and 20CR V2.

In comparison to the varying representation of summer precipitation climatology in the monsoon region, winter-dominated precipitation climatology in the California region is much more consistent between the CMIP5 historical simulations attributable to the models’ ability to capture the main dynamical processes. Extratropical cyclones over the Pacific Ocean move inland in the winter because of the southward migration of the jet stream. The location of the jet stream is strongly related to the atmospheric circulation across the northern Pacific Ocean and North America, which has been previously shown to be related to ocean temperature anomalies on annual time scales (Schubert et al. 2008; Seager et al. 2010). In spite of the limitations in the details of the simulations of southwestern North America precipitation, the robust annual cycle of precipitation and variability in California and Nevada for the CMIP5 historical simulations and reanalysis products therefore indicates sufficiently accurate representation of the current western U.S. hydroclimate, and the potential to identify large-to-regional relationships between the Pacific Ocean and interannual-to-decadal precipitation for this region.

4. Teleconnections from the Pacific Ocean

Multiyear pluvial or drought episodes in western United States during the past century have been identified as being associated with Pacific Ocean SST anomalies (Barlow et al. 2001; Schubert et al. 2004a,b; Capotondi and Alexander 2010; Ting and Wang 1997). The CMIP5 models are therefore assessed for their ability to capture the atmospheric circulation associated with precipitation anomalies over California, and teleconnections from Pacific Ocean SST anomalies, on interannual-to-decadal time scales. The DJF standardized precipitation anomalies for the California region, using a 5-yr running mean as a low-frequency filter, are regressed onto 5-yr running mean DJF SST or 500-hPa geopotential height anomalies. All fields are linearly detrended over the entire dataset in calculating the anomalies. The resulting regression patterns in Figs. 6a and 6b are therefore in units of degrees Celsius or meters per one standard deviation of precipitation in the California region. Regression parameters for each CMIP5 historical run are calculated for the full available time period, generally around 150 years, but only around 50 years for the MIROC4h and CanCM4 models and 100 years for FGOALS-g2 (Table 1). The CMIP5 models have improved ENSO teleconnections compared to the previous generation, with most models able to reproduce the seasonal geopotential departures related to tropical SST anomalies and associated teleconnections (Polade et al. 2013). Global climate models have been found to underestimate the decadal to multidecadal variability compared to proxy data (Ault et al. 2013a,b; Jamison and Kravtsov 2010), whereas atmospheric models forced with ocean data can capture the observed decadal variability over the instrumental record (Hoerling et al. 2011). A failure to produce low-frequency SST variability may contribute to differences seen in the teleconnections between the CMIP5 historical simulations and the observed pattern.

Fig. 6.

Regression of area averaged California region (shown by black rectangle) standardized precipitation anomaly onto (a) 500-hPa geopotential height anomalies (meters per standard deviation of California precipitation) and (b) SST anomalies, or surface temperature anomalies over land (°C per standard deviation of California precipitation), for 5-yr running means of DJF. The color contours indicate the average of all model simulation results, regridded to a coarse common grid spacing. The stippling indicates that more than two-thirds of the models agree on the sign of the regression. The black line contours are the results for GPCC gauge precipitation and ERA-40 geopotential height for 1958–2001 in (a) and GPCC gauge precipitation and GISTEMP anomalies for 1931–2001 in (b). Contour intervals are 4 m and 0.08°C in (a) and (b), respectively; the zero contour is thickened and negative values are dashed.

Fig. 6.

Regression of area averaged California region (shown by black rectangle) standardized precipitation anomaly onto (a) 500-hPa geopotential height anomalies (meters per standard deviation of California precipitation) and (b) SST anomalies, or surface temperature anomalies over land (°C per standard deviation of California precipitation), for 5-yr running means of DJF. The color contours indicate the average of all model simulation results, regridded to a coarse common grid spacing. The stippling indicates that more than two-thirds of the models agree on the sign of the regression. The black line contours are the results for GPCC gauge precipitation and ERA-40 geopotential height for 1958–2001 in (a) and GPCC gauge precipitation and GISTEMP anomalies for 1931–2001 in (b). Contour intervals are 4 m and 0.08°C in (a) and (b), respectively; the zero contour is thickened and negative values are dashed.

The relationship between anomalously low 5-yr average pressure at 500 hPa over the northeast Pacific Ocean and high precipitation along the California west coast, as seen in Fig. 6a, is robust amongst the CMIP5 historical simulations and the reanalysis pattern. Figure 6b shows a cool SST anomaly in the midlatitude northern Pacific Ocean associated with high California precipitation over a 5-yr average, that is similar in location to the Pacific decadal oscillation (PDO) SST pattern based on the leading empirical orthogonal function (EOF) pattern for detrended SST anomalies poleward of 20°N in the Pacific Ocean (Zhang et al. 1997). Equatorial Pacific Ocean SST anomalies also show a weak positive relationship with interannual-to-decadal California precipitation. Note that the stronger response of land surface temperature anomalies to increased precipitation in Fig. 6b is likely a result of the land being able to excite larger-amplitude fluctuations than the ocean. These 5-yr running mean patterns are consistent with successive years of a strengthening and expansion of the Aleutian low, along with intensification and southerly shift of the jet stream resulting in more precipitation from storms crossing the west coast of California in winter (e.g., Gershunov and Barnett 1998; Schubert et al. 2008; Seager et al. 2010). Conversely, a weaker Aleutian low is associated with a more northerly storm track, such that precipitation is diverted northward along the western U.S. coast and the Pacific Northwest sees more precipitation than the California region. Associated cooler equatorial Pacific Ocean SST anomalies also reduce the moisture available to the remaining more southerly storms, emphasizing the reduced long-term precipitation over California (Gershunov and Barnett 1998). On shorter time scales, the equator is expected to have a much more dominant relationship with California precipitation because of the influence of El Niño–Southern Oscillation (ENSO) on the Pacific–North American (PNA) teleconnection pattern. The CMIP5 historical simulations are consistent with previous studies that suggest that drought conditions in the western United States are associated with a negative SST anomaly in equatorial Pacific Ocean (e.g., Cole et al. 2002; Trenberth et al. 1988).

The Atlantic Ocean is also thought to have a strong connection to drought over the United States (McCabe et al. 2004; Feng et al. 2011). The CMIP5 historical simulations show no relationship between the Atlantic Ocean SST and 5-yr running mean California winter precipitation. The regression parameters over the Atlantic Ocean tend to be weak for each individual simulation (not shown), with differences in the sign. A model dependence of the relationship of the Atlantic Ocean was identified for the previous generation of CMIP models for more central U.S. precipitation over the Great Plains (Capotondi and Alexander 2010). Models are consistently found to have cool SST biases in the Atlantic Ocean (Liu et al. 2006; Richter et al. 2014) and less robust results for the Atlantic forcing than for the Pacific (Schubert et al. 2009; Mo et al. 2009). A major role of the Atlantic multidecadal oscillation (AMO) is to modulate ENSO’s effect on drought in western North America (Mo et al. 2009); therefore, the CMIP5 models may be lacking the strength of the connection in comparison to the direct tropical Pacific forcing and internal variability, or the nonlinearity of the influence may be emphasized in the models compared to the reanalysis result. This may also be influenced by a lack of decadal variability in the tropical Atlantic Ocean in the CMIP5 historical simulations. Further diagnosis is beyond the scope of this study but should be undertaken as part of further assessment of the CMIP5 models on the global scale.

As operational decadal forecasting becomes a more realistic endeavor, making accurate predictions requires dominant relationships to account for a substantial fraction of the low-frequency variability in regional precipitation. Figure 7 shows the fraction of variance of low-frequency precipitation averaged over the California region accounted for by anomalies at each grid point, calculated from the coefficient of determination, which assumes a linear relationship. The geopotential height anomalies over the northeast Pacific Ocean are responsible for around 60% of the variance in California precipitation, resulting from the strong relationship of incoming moisture with the latitude of the jet stream. SST anomalies in central-northern Pacific Ocean are associated with less than 20% of the variance in winter California precipitation in the average CMIP5 historical simulations, and less than 30% for the reanalysis. These values are similar to previous estimates of 35% of total interannual U.S. precipitation variability associated with North Pacific Ocean SSTs, determined from identifying modes of interannual-to-decadal variability in observations of the past century (McCabe et al. 2004; Wang and Ting 2000). The CMIP5 historical simulations are also consistent with smaller values for ENSO variability on these time scales compared to extratropical SSTs. The small fraction of explained variance will limit the predictability of precipitation associated with the low-frequency variability and persistence offered by the ocean. Furthermore, evidence that the North Pacific Ocean SST anomalies are generated by the same atmospheric perturbation influencing the precipitation anomaly over California suggests that they are not of practical use in making advanced predictions (e.g., Pierce 2002; Alexander et al. 2002; Kumar et al. 2013). Additional skill would require a feedback of the North Pacific Ocean SST anomalies on atmospheric circulation (Liu et al. 2013; Frankignoul and Sennéchael 2007).

Fig. 7.

Fraction of variance (r2) in precipitation over California region explained by (a) 500-hPa geopotential height and (b) SST anomalies at each grid point, for 5-yr running mean of DJF. Color contours are the average of all CMIP5 historical run results. Contour lines indicate the GPCC precipitation variance explained by 20CR V2 500-hPa geopotential height or GISTEMP anomalies. Note that the contour intervals differ between panels: 0.1 in (a) and 0.05 in (b).

Fig. 7.

Fraction of variance (r2) in precipitation over California region explained by (a) 500-hPa geopotential height and (b) SST anomalies at each grid point, for 5-yr running mean of DJF. Color contours are the average of all CMIP5 historical run results. Contour lines indicate the GPCC precipitation variance explained by 20CR V2 500-hPa geopotential height or GISTEMP anomalies. Note that the contour intervals differ between panels: 0.1 in (a) and 0.05 in (b).

To examine the lack of influence from Pacific SST anomalies, the SST and 500 hPa geopotential height anomalies were compared to the previously calculated regression patterns. For each model, the distribution of standardized precipitation anomalies was divided into three categories; above average (standardized precipitation >0.4), average (standardized precipitation between −0.4 and 0.4), and below average (standardized precipitation <−0.4), with approximately equal population in each category. For each year, the 5-yr running mean 500-hPa geopotential height or SST anomaly was compared to the corresponding regression patterns from Figs. 6a and 6b. The distribution of these pattern correlations for each model was also divided into three categories; positive (correlation >0.2), zero (correlation between −0.2 and 0.2), and negative (correlation <−0.2), with approximately equal population in each category. Figure 8 shows the average result of all CMIP5 simulations, as a percentage of the total number of values, while the error bars indicate a one standard deviation spread of the model results. When the precipitation is average, there are similar amounts of positive, zero, and negative pattern correlations for both the SST and 500-hPa geopotential height anomalies compared to the regression patterns, consistent with randomly shuffling the fields (bootstrap resampling 1000 times with replacement). When the precipitation is above average, there are more than 20% of the years where the SST and 500-hPa geopotential height anomaly patterns are positively correlated with the regression patterns, and less than 4% of the years show negative pattern correlations; both values are significantly different from the results of randomly shuffled fields (bootstrap resampling 1000 times with replacement). The 500-hPa geopotential height anomaly more often exhibits the inverse of the regression pattern for low precipitation values than for the SST anomalies. This is consistent with the atmosphere not always requiring the Pacific Ocean SST forcing to generate deviations in the circulation that result in anomalous precipitation over California in winter. It is likely that decadal drought is often generated by intrinsic internal atmospheric variability rather than forcing by oceanic heating anomalies. A similar result has been noted in other coupled model simulations (Hoerling et al. 2009; Capotondi and Alexander 2010; Hunt 2011; Stevenson et al. 2014, manuscript submitted to J. Climate) and statistical analyses (Cook et al. 2011). Persistence of initial anomalies driven by stochastic forcing by the atmosphere may occur in the presence of feedbacks with the land surface. This shifts the focus of future megadrought predictability studies toward the role of antecedent land conditions.

Fig. 8.

Histogram of the average of all model results calculated as a percentage of total years that fall into each category. For precipitation in three categories: high (precipitation standardized anomaly >0.4), average (precipitation standardized anomaly between −0.4 and 0.4), and low (precipitation standardized anomaly <−0.4). For pattern correlation of 500-hPa geopotential height or SST anomalies for each year with corresponding regression patterns in Figs. 6a,b in three categories: positive (pattern correlation >0.2), zero (pattern correlation between −0.2 and 0.2), and negative (pattern correlation <−0.2). Error bars indicate one standard deviation of the model spread. Significant results are indicated by a star above the bar (calculated from bootstrap resampling 1000 times).

Fig. 8.

Histogram of the average of all model results calculated as a percentage of total years that fall into each category. For precipitation in three categories: high (precipitation standardized anomaly >0.4), average (precipitation standardized anomaly between −0.4 and 0.4), and low (precipitation standardized anomaly <−0.4). For pattern correlation of 500-hPa geopotential height or SST anomalies for each year with corresponding regression patterns in Figs. 6a,b in three categories: positive (pattern correlation >0.2), zero (pattern correlation between −0.2 and 0.2), and negative (pattern correlation <−0.2). Error bars indicate one standard deviation of the model spread. Significant results are indicated by a star above the bar (calculated from bootstrap resampling 1000 times).

5. Land surface influence on the North American monsoon

The North American summer monsoon is a distinctly different precipitation regime than the winter storms over California. In the central United States, the dependence of precipitation on moisture recycling occurs predominantly during the summer season (Anderson et al. 2008). The hydroclimate of the interior region over New Mexico and northern Mexico is more likely to be influenced by land processes and soil moisture than SST teleconnections (Koster et al. 2004; Conil et al. 2007). Because of the increased likelihood of persistence of precipitation anomalies in regions where SST forcing is weak, interannual to multidecadal drought research in North America has often focused on the Great Plains (Anderson et al. 2008; Koster et al. 2000, 2004; Oglesby and Erickson 1989; Schubert et al. 2008; Mo et al. 1997; Mo and Juang 2003). To assess the role of land–atmosphere feedbacks in relation to SST teleconnections in the CMIP5 historical simulations, the North American monsoon region is examined for summer precipitation anomalies over 5 years. Standardized 5-yr running mean precipitation anomalies were regressed onto 5-yr running mean SST and land surface temperature or 500-hPa geopotential height anomalies for July–September (Figs. 9a,b). Using a summer seasonal mean shifted later in the calendar year than June–August (JJA) accounts for the delay of the thunderstorms reaching the northernmost extension of the monsoon in Colorado and avoids any discrepancy in monsoon onset dates between model simulations within June. Figure 9a indicates the low-frequency monsoon precipitation is associated with a positive 500-hPa geopotential height anomaly over the northern United States, and a negative anomaly over the monsoon region. This is consistent with a northward shift of the monsoon ridge resulting in higher precipitation. More than two-thirds of the models show similar results, despite shifts in the climatological circulation pattern. When considering only the high-resolution CMIP5 simulations to minimize the differences in climatological circulation, the average regression pattern is similar (not shown). The reanalysis shows an eastward displacement of the dipole pressure anomaly compared to the average of the model results.

Fig. 9.

Simultaneous regression of area averaged monsoon region (shown by black rectangle) standardized precipitation anomaly onto (a) 500-hPa geopotential height anomalies (meters per standard deviation of monsoon precipitation), (b) SST anomalies or surface temperature anomalies over land (°C per standard deviation of monsoon precipitation), and (c) temperature anomalies at 500-hPa level (°C per standard deviation of monsoon precipitation) for 5-yr running means of JAS. The color contours indicate the average of all model simulation results, regridded to a coarse common grid spacing. The stippling indicates that more than two-thirds of the models agree on the sign of the regression. The black line contours are the results for GPCC gauge precipitation and (a) ERA-40 geopotential height for 1958–2001 and (b) GISTEMP anomalies for 1931–2001. Contour intervals are 1 m and 0.6°C in (a) and (b), respectively; the zero contour is thickened and negative values are dashed.

Fig. 9.

Simultaneous regression of area averaged monsoon region (shown by black rectangle) standardized precipitation anomaly onto (a) 500-hPa geopotential height anomalies (meters per standard deviation of monsoon precipitation), (b) SST anomalies or surface temperature anomalies over land (°C per standard deviation of monsoon precipitation), and (c) temperature anomalies at 500-hPa level (°C per standard deviation of monsoon precipitation) for 5-yr running means of JAS. The color contours indicate the average of all model simulation results, regridded to a coarse common grid spacing. The stippling indicates that more than two-thirds of the models agree on the sign of the regression. The black line contours are the results for GPCC gauge precipitation and (a) ERA-40 geopotential height for 1958–2001 and (b) GISTEMP anomalies for 1931–2001. Contour intervals are 1 m and 0.6°C in (a) and (b), respectively; the zero contour is thickened and negative values are dashed.

The regression of monsoon precipitation onto SST and land surface temperature anomalies shows a strong local negative land temperature anomaly that is consistent among all of the CMIP5 models and the reanalysis (Fig. 9b). The regression parameters for the reanalysis also show an increase in monsoon precipitation for cooler local Pacific Ocean SST anomalies, which would enhance the land–ocean temperature contrast and strengthen the offshore winds along the west coast of Mexico. There is an order of magnitude difference between land temperature anomalies and SST anomalies in Fig. 9b, suggesting that the land response dominates beyond the difference between land and ocean in exciting temperature changes. There is a warm anomaly over southern Canada in the average of the regression patterns for the CMIP5 historical simulations, suggestive of the intense solar heating during the summer months, which helps set up a strong monsoon circulation. However, this is not supported by the reanalysis, which shows negative regression parameters into northern Canada for the shorter time period. This may also be attributable to the shifted location of the circulation anomaly relative to the average of the CMIP5 models in Fig. 9a. The warm anomaly over southern Canada is confirmed in the models by regressing precipitation onto the 500-hPa temperature anomaly, which shows a strengthening of the warm anomaly over northern North America and a cool anomaly over the Gulf of Mexico (Fig. 9c). It is apparent from this midtropospheric temperature gradient that the upper-level large-scale circulation plays an important role in driving the monsoon. The models results are not consistent for the actual monsoon region, with less than two-thirds of the models agreeing on the sign of the regression.

The negative temperature anomaly at the surface associated with high monsoon precipitation is likely attributable to a feedback where increased evaporative cooling occurs when the precipitation totals are higher and there is more available surface water (Matsui et al. 2003). The changes in the land surface also influence the water vapor abundance and cloud conditions, which further modify the surface energy, including reduced downwelling solar radiation because of increased clouds cover during thunderstorms. Composites of years where JAS monsoon standardized precipitation anomalies are greater than 1.5 standard deviations from the mean were calculated to investigate the timing and magnitude of the different land feedbacks. The changes in the monthly mean radiative, sensible, and latent heat fluxes leading up to the monsoon season were found to occur in the same month (around June) as the changes in land temperature and precipitation amounts. On average for the CMIP5 historical simulations, there is a 5% increase in the mean cloud cover in July–September associated with one standard deviation increase in precipitation in the monsoon region. This results in a decrease in downwelling shortwave radiation, and an increase in downwelling longwave radiation. Results for the dry monsoon composite show opposite sign anomalies of similar magnitude. Figure 10 shows the mean components of the surface energy flux balance for composites from 19 higher-resolution models with available energy fields, and the resulting direction of temperature change resulting from an imbalance of the anomalous radiative, thermal, and evaporative fluxes. The sum of the radiative components for the wet monsoon composite is a positive anomaly (i.e., an increase in energy into the surface) of around 3 W m−2. This does not completely offset the net turbulent heat flux, which is a negative latent plus sensible heat flux anomaly (magnitude around 3.5 W m−2 out of the surface) resulting from increased evaporation. Therefore the local surface temperature is expected to decrease during monsoon years with high precipitation.

Fig. 10.

Average of the energy flux balance components for 19 higher-resolution CMIP5 historical simulations for (a) composite of wet monsoon years and (b) composite of dry monsoon years, with error bars showing a one standard deviation spread in the results. The far right bar shows the residual ground heat flux plus temperature change. Positive values indicate an increase in energy into the surface.

Fig. 10.

Average of the energy flux balance components for 19 higher-resolution CMIP5 historical simulations for (a) composite of wet monsoon years and (b) composite of dry monsoon years, with error bars showing a one standard deviation spread in the results. The far right bar shows the residual ground heat flux plus temperature change. Positive values indicate an increase in energy into the surface.

The strong land–atmosphere feedback evident in the monsoon region and the coincident changes in temperature, evaporation, and precipitation make it difficult to accurately relate the monsoon strength to local antecedent soil moisture or temperature conditions. However, the timing, strength, and location of the ridge that influences the strength of the monsoon may be affected by the large-scale preexisting conditions. For example, the presence of above normal snowpack in the southern Rocky Mountains in spring has been shown to be intermittently associated with suppressed monsoon precipitation through increased evaporation and thermal inertia maintaining low surface temperatures (Gutzler and Preston 1997; Gutzler 2000; Mo and Paegle 2000; Griffin et al. 2013). The CMIP5 historical simulations and reanalysis products show no significant lagged correlation between winter or spring precipitation over the western United States and the following summer monsoon precipitation (Fig. 11). The lack of a correlation between antecedent conditions persisting from winter to the following summer over 150 years is consistent with long-term observations and tree-ring chronology (Griffin et al. 2013; Gutzler 2000; Zhu et al. 2005; Matsui et al. 2003). The lack of a connection between southern Rocky Mountain snowpack and monsoon precipitation may also be attributable to the northern extension of the later summer monsoon, which is more likely to be detached from antecedent conditions primed in the winter and spring. Additionally, the relevant snowpack, soil water, and streamflow dynamics may not be accurately captured in the CMIP5 land models in order to maintain low surface temperatures in years with high winter precipitation.

Fig. 11.

Regression of area averaged monsoon region (shown by black rectangle) standardized precipitation anomaly (kg m−2 per standard deviation of monsoon precipitation) for JAS onto previous April soil moisture anomalies. The color contours indicate the average of all 20 simulations, regridded to a coarse common grid spacing. The stippling indicates that more than two-thirds of the models agree on the sign of the regression.

Fig. 11.

Regression of area averaged monsoon region (shown by black rectangle) standardized precipitation anomaly (kg m−2 per standard deviation of monsoon precipitation) for JAS onto previous April soil moisture anomalies. The color contours indicate the average of all 20 simulations, regridded to a coarse common grid spacing. The stippling indicates that more than two-thirds of the models agree on the sign of the regression.

To examine the potential for antecedent conditions to influence the monsoon precipitation through local soil moisture rather than winter or spring precipitation, the top 2 m of the simulated soil moisture data was used. Not all modeling groups provided soil moisture data to the archive; models that contributed soil moisture data to Fig. 11 are indicated in Table 1 by a cross in the first column. While the top 20 cm of soil is exposed to high-frequency variability (time scales of less than a month), at a depth of 2 m soil moisture provides a natural integrator of precipitation and evaporation, and mostly filters out variability on time scales of less than 3 months (Wu and Dickinson 2004). The JAS monsoon standardized precipitation anomalies were regressed on to the integrated soil moisture anomalies for the months leading up to summer. Figure 11 shows the average of the regression parameters for the JAS standardized precipitation anomaly onto the previous April soil moisture. There are 20 CMIP5 historical simulations with available monthly soil moisture values, four of which are in the higher-resolution category for the atmosphere component. The regression parameters are not significant and model dependent for the region where a signal from snowpack in the southern Rocky Mountains would be expected. However, the average of the model results shown in Fig. 11 instead highlights a consistent relationship of high monsoon precipitation with abundant antecedent soil moisture in the southern Great Plains region, over Texas and Oklahoma, potentially bringing moisture to the monsoon region via winds from the Gulf of Mexico. The CanESM2, FGOALS-s2, and BCC-CSM1.1 historical simulations show very low regression parameters for the entire continent (not shown), which suggests a weak interaction between the land surface and precipitation in those models. A stronger relationship is evident in the higher-resolution models, such as ACCESS1.0, HadGEM2-CC, and HadGEM2-ES.

The low correlation between interannual precipitation variability and the Pacific Ocean, which was hypothesized to be an important source of decadal memory in the western United States, suggests that an alternate mechanism for persisting precipitation anomalies is required for effective forecasts of drought-like conditions in southwestern North America. This analysis suggests there is a broad consensus with CMIP5 models for local land surface feedbacks, and the connection between the antecedent conditions through soil moisture memory or land temperatures, as important sources of memory in the system. Therefore, understanding the dynamics of consecutive dry seasons will aid in predicting perturbations that will develop into established megadroughts in southwestern North America in the future.

6. Conclusions

The CMIP5 historical simulations were assessed for their representation of the current hydroclimate in southwestern North America, and used to investigate the connection with ocean- and land-based sources of decadal memory over the past 150 years. An outstanding research question in decadal predictions is the relative roles of ocean forcing, land feedbacks, and intrinsic variability of the atmosphere in generating and persisting long-term drought in arid regions such as southwestern North America. The potential economic and social cost of long-term drought-like conditions, and the benefit possible from effective forecasts, motivates the need to understand these mechanisms.

Analysis of decadal variability in the CMIP5 coupled modeling framework requires robust simulations that accurately represent the current climate and dynamical processes involved in the hydrological cycle. On average, the CMIP5 models have higher climatological precipitation in southwestern North America than reanalysis products. The annual cycle of precipitation over California and Nevada is accurately represented in the CMIP5 models, with a peak in boreal winter. Shortcomings in summer precipitation in southwestern North America are more severe than for winter precipitation and are associated with the location or spatial structure of the monsoon circulation. Higher-resolution models are more capable of the orographic precipitation along the west coast of Mexico because of a better representation of topography. Climatological winds along the Gulf of California in July are misrepresented in CMIP5 models, more severely limiting the moisture flux in Arizona and New Mexico in the lower-resolution models.

Previous studies have attributed drought in southwestern North America to Pacific or Atlantic SST anomalies (e.g., Meehl and Hu 2006). Here, regressions have shown that robust coincident anomaly patterns in the tropical and North Pacific Ocean and low-frequency (5 yr) winter California precipitation exist in the CMIP5 historical simulations, but are unable to explain more than 20% of the decadal variability, consistent with observations over the past century. The small fraction of explained variance will limit the predictability of precipitation associated with the decadal variability and persistence offered by the ocean. The remainder of the decadal variability is likely to be coupled with land–atmosphere feedbacks in regions where SST forcing is limited, or with anomalies generated by internal variability of the atmosphere.

The interannual signal of monsoon precipitation extending into the southwestern United States is strongly associated with local land temperatures attributable to evaporative cooling and radiation changes with varying cloud cover. This is consistent among the CMIP5 historical simulations, despite differences in the climatological monsoon circulation and the spatial patterns of resulting precipitation. The absence of a robust connection between low-frequency precipitation and SST anomalies in the Pacific Ocean emphasizes the role of antecedent land conditions as the source of memory in the system. Soil moisture integrated over the top 2 m in Oklahoma and Texas in April is shown to be positively correlated with the following summer monsoon precipitation in 17 out of the 20 model simulations with available data. This suggests that future investigations of decadal precipitation predictability will require more comprehensive consideration of land–atmosphere feedbacks involving soil moisture and aerosol (i.e., dust and soot) forcing.

The respective roles of land feedbacks, oceanic forcing, and atmospheric internal variability in persisting long-term drought-like conditions is an outstanding research question. These results show that the Pacific Ocean is not a dominant driver of interannual to decadal regional hydroclimate variability in the CMIP5 models. Therefore, in order to make meaningful decadal predictions in the future, it is likely that forecasting will move away from SST-driven anomaly patterns and focus on surface processes that can cause persistence of a precipitation anomaly.

Acknowledgments

This research is part of the Investigation of Decadal Climate Predictability and Hydroclimate Impacts (IDCPI) on the western United States, which is supported by the National Science Foundation, Division of Atmospheric and Geospace Sciences, Paleoclimate (NSF-Paleoclimate 1049104).

We acknowledge the World Climate Research Programme’s Working Group on Coupled Modelling, which is responsible for CMIP, and we thank the climate modeling groups (listed in Table 1 of this paper) for producing and making available their model output. For CMIP the U.S. Department of Energy’s Program for Climate Model Diagnosis and Intercomparison provides coordinating support and led development of software infrastructure in partnership with the Global Organization for Earth System Science Portals. (Hurrell et al. 2011; Taylor et al. 2012; Meehl et al. 2009; Hibbard et al. 2007).

GPCC Precipitation data (http://gpcc.dwd.de/) are provided by the NOAA/OAR/ ESRL PSD, Boulder, Colorado, USA, from their Web site (http://www.esrl.noaa.gov/psd/).

ERA-40 data used in this study have been obtained from the ECMWF Data Server.

The 20CR V2 data are provided by the NOAA/OAR/ESRL PSD, Boulder, Colorado, from their Web site (http://www.esrl.noaa.gov/psd/). Support for the Twentieth Century Reanalysis Project dataset is provided by the U.S. Department of Energy, Office of Science Innovative and Novel Computational Impact on Theory and Experiment (DOE INCITE) program, and Office of Biological and Environmental Research (BER), and by the National Oceanic and Atmospheric Administration Climate Program Office.

GISTEMP LOTI data are provided by NASA Goddard Institute for Space Studies, from their Web site (http://data.giss.nasa.gov/gistemp/).

The authors gratefully acknowledge the helpful comments from the anonymous reviewers.

REFERENCES

REFERENCES
Adams
,
D.
, and
A.
Comrie
,
1997
:
The North American monsoon
.
Bull. Amer. Meteor. Soc.
,
78
,
2197
2213
,
doi:10.1175/1520-0477(1997)078<2197:TNAM>2.0.CO;2
.
Alexander
,
M.
,
I.
Bladé
,
M.
Newman
,
J.
Lanzante
,
N.-C.
Lau
, and
J.
Scott
,
2002
:
The atmospheric bridge: The influence of ENSO teleconnections on air–sea interaction over the global oceans
.
J. Climate
,
15
,
2205
2231
,
doi:10.1175/1520-0442(2002)015<2205:TABTIO>2.0.CO;2
.
Anderson
,
B.
,
G.
Salvucci
,
A.
Ruane
,
J.
Roads
, and
M.
Kanamitsu
,
2008
:
A new metric for estimating the influence of evaporation on seasonal precipitation rates
.
J. Hydrometeor.
,
9
,
576
588
,
doi:10.1175/2007JHM968.1
.
Ault
,
T.
,
J.
Cole
,
J.
Overpeck
,
G.
Pederson
,
S.
St. George
,
B.
Otto-Bliesner
,
C.
Woodhouse
, and
C.
Deser
,
2013a
:
The continuum of hydroclimate variability in western North America during the last millennium
.
J. Climate
,
26
,
5863
5878
,
doi:10.1175/JCLI-D-11-00732.1
.
Ault
,
T.
,
C.
Deser
,
M.
Newman
, and
J.
Emile-Geay
,
2013b
:
Characterizing decadal to centennial variability in the equatorial Pacific during the last millennium
.
Geophys. Res. Lett.
,
40
,
3450
3456
,
doi:10.1002/grl.50647
.
Barlow
,
M.
,
S.
Nigam
, and
E.
Berbery
,
2001
:
ENSO, Pacific decadal variability, and U.S. summertime precipitation, drought, and stream flow
.
J. Climate
,
14
,
2105
2128
,
doi:10.1175/1520-0442(2001)014<2105:EPDVAU>2.0.CO;2
.
Barsugli
,
J.
, and
D.
Battisti
,
1998
:
The basic effects of atmospheric–ocean thermal coupling on midlatitude variability
.
J. Atmos. Sci.
,
55
,
477
493
,
doi:10.1175/1520-0469(1998)055<0477:TBEOAO>2.0.CO;2
.
Beck
,
C.
,
J.
Grieser
, and
B.
Rudolf
,
2005
:
A new monthly precipitation climatology for the global land areas for the period 1951 to 2000
.
Geophysical Research Abstracts
, Vol. 7, Abstract 07154.
Becker
,
A.
,
P.
Finger
,
A.
Meyer-Christoffer
,
B.
Rudolf
,
K.
Schamm
,
U.
Schneider
, and
M.
Ziese
,
2013
:
A description of the global land-surface precipitation data products of the Global Precipitation Climatology Centre with sample applications including centennial (trend) analysis from 1901–present
.
Earth Syst. Sci. Data
,
5
,
71
99
,
doi:10.5194/essd-5-71-2013
.
Bellouin
,
N.
,
O.
Boucher
,
J.
Haywood
,
C.
Johnson
,
A.
Jones
,
J.
Rae
, and
S.
Woodward
,
2007
: Improved representation of aerosols for HadGEM2. Hadley Centre Tech. Note 73, 42 pp.
Bentsen
,
M.
, and Coauthors
,
2012
:
The Norwegian Earth System Model, NorESM1-M—Part 1: Description and basic evaluation
.
Geosci. Model Dev. Discuss.
,
5
,
2843
2931
,
doi:10.5194/gmdd-5-2843-2012
.
Branstator
,
G.
, and
H.
Teng
,
2012
: Potential impact of initialization on decadal predictions as assessed for CMIP5 models. Geophys. Res. Lett.,39, L12703, doi:10.1029/2012GL051974.
Capotondi
,
A.
, and
M.
Alexander
,
2010
:
Relationship between precipitation in the Great Plains of the United States and global SSTs: Insights from the IPCC AR4 models
.
J. Climate
,
23
,
2941
2958
,
doi:10.1175/2009JCLI3291.1
.
Cayan
,
D.
,
T.
Das
,
D.
Pierce
,
T.
Barnett
,
M.
Tyree
, and
A.
Gershunov
,
2010
:
Future dryness in the southwest US and the hydrology of the early 21st century drought
.
Proc. Natl. Acad. Sci. USA
,
107
,
21 271
21 276
,
doi:10.1073/pnas.0912391107
.
Christensen
,
N.
, and
D.
Lettenmaier
,
2007
:
A multimodel ensemble approach to assessment of climate change impacts on the hydrology and water resources of the Colorado River Basin
.
Hydrol. Earth Syst. Sci.
,
11
,
1417
1434
,
doi:10.5194/hess-11-1417-2007
.
Cole
,
J.
,
J.
Overpeck
, and
E.
Cook
,
2002
:
Multiyear La Niña events and persistent drought in the contiguous United States
.
Geophys. Res. Lett.
,
29
(13),
doi:10.1029/2001GL013561
.
Collier
,
M.
, and Coauthors
,
2011
: The CSIRO-Mk3.6.0 atmosphere–ocean GCM: Participation in CMIP5 and data publication. Extended Abstracts, 19th Int. Congress on Modelling and Simulation, Perth, Australia, CSIRO, 2691–2697.
Collins
,
M.
,
S.
Tett
, and
C.
Cooper
,
2001
:
The internal climate variability of HadCM3, a version of the Hadley Centre coupled model without flux adjustments
.
Climate Dyn.
,
17
,
61
81
,
doi:10.1007/s003820000094
.
Collins
,
W. J.
, and Coauthors
,
2008
: Evaluation of the HadGEM2 model. Hadley Centre Tech. Note 74, 47 pp.
Collins
,
W. J.
, and Coauthors
,
2011
:
Development and evaluation of an Earth-system model—HadGEM2
.
Geosci. Model Dev. Discuss.
,
4
,
997
1062
,
doi:10.5194/gmdd-4-997-2011
.
Compo
,
G.
,
J.
Whitaker
, and
P.
Sardeshmukh
,
2006
:
Feasibility of a 100-year reanalysis using only surface pressure data
.
Bull. Amer. Meteor. Soc.
,
87
,
175
190
,
doi:10.1175/BAMS-87-2-175
.
Compo
,
G.
, and Coauthors
,
2011
:
The Twentieth Century Reanalysis Project
.
Quart. J. Roy. Meteor. Soc.
,
137
,
1
28
,
doi:10.1002/qj.776
.
Conil
,
S.
,
H.
Douville
, and
S.
Tyteca
,
2007
:
The relative influence of soil moisture and SST in climate predictability explored within ensembles of AMIP type experiments
.
Climate Dyn.
,
28
,
125
145
,
doi:10.1007/s00382-006-0172-2
.
Cook
,
B.
,
E.
Cook
,
K.
Anchukaitis
,
R.
Seager
, and
R.
Miller
,
2011
:
Forced and unforced variability of twentieth century North American droughts and pluvials
.
Climate Dyn.
,
37
,
1097
1110
,
doi:10.1007/s00382-010-0897-9
.
Cook
,
E.
,
C.
Woodhouse
,
C.
Eakin
,
D.
Meko
, and
D.
Stahle
,
2004
:
Long-term aridity changes in the western United States
.
Science
,
306
,
1015
1018
,
doi:10.1126/science.1102586
.
Deser
,
C.
, and
M.
Blackmon
,
1995
:
On the relationship between tropical and North Pacific sea surface temperature variations
.
J. Climate
,
8
,
1677
1680
,
doi:10.1175/1520-0442(1995)008<1677:OTRBTA>2.0.CO;2
.
Douglas
,
M.
,
R.
Maddox
,
K.
Howard
, and
S.
Reyes
,
1993
:
The Mexican monsoon
.
J. Climate
,
6
,
1665
1677
,
doi:10.1175/1520-0442(1993)006<1665:TMM>2.0.CO;2
.
Eltahir
,
E.
,
1998
:
A soil moisture–rainfall feedback mechanism: 1. Theory and observations
.
Water Resour. Res.
,
34
,
765
776
,
doi:10.1029/97WR03499
.
Entekhabi
,
D.
,
I.
Rodriguez-Iturbe
, and
R.
Bras
,
1992
:
Variability in large-scale water balance with land surface–atmosphere interaction
.
J. Climate
,
5
,
798
813
,
doi:10.1175/1520-0442(1992)005<0798:VILSWB>2.0.CO;2
.
Feng
,
S.
,
Q.
Hu
, and
R.
Oglesby
,
2011
:
Influence of Atlantic sea surface temperatures on persistent drought in North America
.
Climate Dyn.
,
37
,
569
586
,
doi:10.1007/s00382-010-0835-x
.
Frankignoul
,
C.
, and
N.
Sennéchael
,
2007
:
Observed influence of North Pacific SST anomalies on the atmospheric circulation
.
J. Climate
,
20
,
592
606
,
doi:10.1175/JCLI4021.1
.
Gan
,
B.
, and
L.
Wu
,
2013
:
Seasonal and long-term coupling between wintertime storm tracks and sea surface temperature in the North Pacific
.
J. Climate
,
26
,
6123
6136
,
doi:10.1175/JCLI-D-12-00724.1
.
Gershunov
,
G.
, and
T.
Barnett
,
1998
:
Interdecadal modulation of ENSO teleconnections
.
Bull. Amer. Meteor. Soc.
,
79
,
2715
2725
,
doi:10.1175/1520-0477(1998)079<2715:IMOET>2.0.CO;2
.
Goddard
,
L.
, and Coauthors
,
2013
:
A verification framework for interannual-to-decadal prediction experiments
.
Climate Dyn.
,
40
,
245
272
, doi:10.1007/s00382-012-1481-2.
Griffin
,
D.
, and Coauthors
,
2013
:
North American monsoon precipitation reconstructed from tree-ring latewood
.
Geophys. Res. Lett.
,
40
,
954
958
,
doi:10.1002/grl.50184
.
Gutzler
,
D.
,
2000
:
Covariability of spring snowpack and summer rainfall across the southwest United States
.
J. Climate
,
13
,
4018
4027
,
doi:10.1175/1520-0442(2000)013<4018:COSSAS>2.0.CO;2
.
Gutzler
,
D.
, and
J.
Preston
,
1997
:
Evidence for a relationship between spring snow cover in North America and summer rainfall in New Mexico
.
Geophys. Res. Lett.
,
24
,
2207
2210
,
doi:10.1029/97GL02099
.
Hales
,
J. J.
,
1972
:
Surges of maritime tropical air northward over the Gulf of California
.
Mon. Wea. Rev.
,
100
,
298
306
,
doi:10.1175/1520-0493(1972)100<0298:SOMTAN>2.3.CO;2
.
Hansen
,
J.
,
D.
Johnson
,
A.
Lacis
,
S.
Lebedeff
,
P.
Lee
,
D.
Rind
, and
G.
Russell
,
1981
:
Climate impact of increasing atmospheric carbon dioxide
.
Science
,
213
,
957
966
,
doi:10.1126/science.213.4511.957
.
Hansen
,
J.
,
R.
Ruedy
,
J.
Glascoe
, and
M.
Sato
,
1999
:
GISS analysis of surface temperature change
.
J. Geophys. Res.
,
104
,
30 997
31 022
,
doi:10.1029/1999JD900835
.
Hansen
,
J.
,
R.
Ruedy
,
M.
Sato
,
M.
Imhoff
,
W.
Lawrence
,
D.
Easterling
,
T.
Peterson
, and
T.
Karl
,
2001
:
A closer look at United States and global surface temperature change
.
J. Geophys. Res.
,
106
,
23 947
23 963
,
doi:10.1029/2001JD000354
.
Hansen
,
J.
,
R.
Ruedy
,
M.
Sato
, and
K.
Lo
,
2010
:
Global surface temperature change
.
Rev. Geophys.
,
48
,
RG4004
,
doi:10.1029/2010RG000345
.
Hibbard
,
K.
,
G.
Meehl
,
P.
Cox
, and
P.
Friedlingstein
,
2007
:
A strategy for climate change stabilization experiments
.
Eos, Trans. Amer. Geophys. Union
,
88
,
217
221
,
doi:10.1029/2007EO200002
.
Higgins
,
R.
, and
W.
Shi
,
2000
:
Dominant factors responsible for interannual variability of the summer monsoon in the southwestern United States
.
J. Climate
,
13
,
759
776
,
doi:10.1175/1520-0442(2000)013<0759:DFRFIV>2.0.CO;2
.
Higgins
,
R.
,
Y.
Yao
, and
X.
Wang
,
1997
:
Influence of the North American monsoon system on the U.S. summer precipitation regime
.
J. Climate
,
10
,
2600
2622
,
doi:10.1175/1520-0442(1997)010<2600:IOTNAM>2.0.CO;2
.
Higgins
,
R.
,
K.
Mo
, and
Y.
Yao
,
1998
:
Interannual variability of the U.S. summer precipitation regime with emphasis on the southwestern monsoon
.
J. Climate
,
11
,
2582
2606
,
doi:10.1175/1520-0442(1998)011<2582:IVOTUS>2.0.CO;2
.
Hoerling
,
M.
, and
A.
Kumar
,
2003
:
The perfect ocean for drought
.
Science
,
299
,
691
694
,
doi:10.1126/science.1079053
.
Hoerling
,
M.
,
X.-W.
Quan
, and
J.
Eischeid
,
2009
: Distinct causes for two principal U.S. droughts of the 20th century. Geophys. Res. Lett.,36, L19708, doi:10.1029/2009GL039860.
Hoerling
,
M.
, and Coauthors
,
2011
:
On North American decadal climate for 2011–20
.
J. Climate
,
24
,
4519
4528
,
doi:10.1175/2011JCLI4137.1
.
Hunt
,
B.
,
2011
:
Global characteristics of pluvial and dry multi-year episodes with emphasis on megadroughts
.
Int. J. Climatol.
,
31
,
1425
1439
,
doi:10.1002/joc.2166
.
Hurrell
,
J.
,
M.
Visbeck
, and
A.
Pirani
, Eds.,
2011
: WCRP Coupled Model Intercomparison Project Phase 5—CMIP5. CLIVAR Exchanges, No. 16, International CLIVAR Project Office, Southampton, United Kingdom, 52 pp. [Available online at http://www.clivar.org/publications/exchanges/Exchanges_56.pdf.]
Iversen
,
T.
, and Coauthors
,
2012
:
The Norwegian Earth System Model, NorESM1-M—Part 2: Climate response and scenario projections
.
Geosci. Model Dev. Discuss.
,
5
,
2933
2998
,
doi:10.5194/gmdd-5-2933-2012
.
Jamison
,
N.
, and
S.
Kravtsov
,
2010
:
Decadal variations of North Atlantic sea surface temperature in observations and CMIP3 simulations
.
J. Climate
,
23
,
4619
4636
,
doi:10.1175/2010JCLI3598.1
.
Jia
,
L.
, and
T.
DelSole
,
2012
: Multi-year predictability of temperature and precipitation in multiple climate models. Geophys. Res. Lett.,39, L17705, doi:10.1029/2012GL052778.
Johns
,
T.
, and Coauthors
,
2006
:
The new Hadley Centre climate model (HadGEM1): Evaluation of coupled simulations
.
J. Climate
,
19
,
1327
1353
,
doi:10.1175/JCLI3712.1
.
Jones
,
C.
, and Coauthors
,
2011
:
The HadGEM2-ES implementation of CMIP5 centennial simulations
.
Geosci. Model Dev.
,
4
,
543
570
,
doi:10.5194/gmd-4-543-2011
.
Keenlyside
,
N.
,
M.
Latif
,
J.
Jungclaus
,
L.
Kornblueh
, and
E.
Roeckner
,
2008
:
Advancing decadal-scale climate predictions in the North Atlantic sector
.
Nature
,
453
,
84
88
,
doi:10.1038/nature06921
.
Koster
,
R.
,
M.
Suarez
, and
M.
Heiser
,
2000
:
Variance and predictability of precipitation at seasonal-to-interannual timescales
.
J. Hydrometeor.
,
1
,
26
46
,
doi:10.1175/1525-7541(2000)001<0026:VAPOPA>2.0.CO;2
.
Koster
,
R.
, and Coauthors
,
2004
:
Regions of strong coupling between soil moisture and precipitation
.
Science
,
305
,
1138
1140
,
doi:10.1126/science.1100217
.
Kumar
,
A.
,
M.
Chen
,
M.
Hoerling
, and
J.
Eischeid
,
2013
:
Do extreme events require extreme forcings?
Geophys. Res. Lett.
,
40
,
3440
3445
,
doi:10.1002/grl.50657
.
Kushnir
,
Y.
,
R.
Seager
,
M.
Ting
,
N.
Naik
, and
J.
Nakamura
,
2010
:
Mechanisms of tropical Atlantic SST influence on North American precipitation variability
.
J. Climate
,
23
,
5610
5628
,
doi:10.1175/2010JCLI3172.1
.
Lau
,
N.-C.
,
1997
:
Interactions between global SST anomalies and the midlatitude atmospheric circulation
.
Bull. Amer. Meteor. Soc.
,
78
,
21
33
,
doi:10.1175/1520-0477(1997)078<0021:IBGSAA>2.0.CO;2
.
Liu
,
H.
,
C.
Wang
,
S.-K.
Lee
, and
D.
Enfield
,
2013
:
Atlantic warm pool variability in the CMIP5 simulations
.
J. Climate
,
26
,
5315
5336
,
doi:10.1175/JCLI-D-12-00556.1
.
Liu
,
Q.
,
N.
Wen
, and
Z.
Liu
,
2006
: An observational study of the impact of the North Pacific SST on the atmosphere. Geophys. Res. Lett.,33, L18611, doi:10.1029/2006GL026082.
Lo
,
F.
, and
M.
Clark
,
2002
:
Relationships between spring snow mass and summer precipitation in the southwestern United States associated with the North American monsoon system
.
J. Climate
,
15
,
1378
1385
,
doi:10.1175/1520-0442(2002)015<1378:RBSSMA>2.0.CO;2
.
Marsland
,
S.
,
H.
Haak
,
J.
Jungclaus
,
M.
Latif
, and
F.
Röske
,
2003
:
The Max-Planck-Institute global ocean/sea ice model with orthogonal curvilinear coordinates
.
Ocean Modell.
,
5
,
91
127
,
doi:10.1016/S1463-5003(02)00015-X
.
Martin
,
G.
, and Coauthors
,
2011
:
The HadGEM2 family of Met Office Unified Model climate configurations
.
Geosci. Model Dev.
,
4
,
723
757
,
doi:10.5194/gmd-4-723-2011
.
Matsui
,
T.
,
V.
Lakshmi
, and
E.
Small
,
2003
:
Links between snow cover, surface skin temperature, and rainfall variability in the North American monsoon system
.
J. Climate
,
16
,
1821
1829
,
doi:10.1175/1520-0442(2003)016<1821:LBSCSS>2.0.CO;2
.
McCabe
,
G.
,
M.
Palecki
, and
J.
Betancourt
,
2004
:
Pacific and Atlantic Ocean influences on multidecadal drought frequency in the United States
.
Proc. Natl. Acad. Sci. USA
,
101
,
4136
4141
,
doi:10.1073/pnas.0306738101
.
Meehl
,
G.
, and
A.
Hu
,
2006
:
Megadroughts in the Indian monsoon region and southwest North America and a mechanism for associated multidecadal Pacific sea surface temperature anomalies
.
J. Climate
,
19
,
1605
1623
,
doi:10.1175/JCLI3675.1
.
Meehl
,
G.
, and Coauthors
,
2009
:
Decadal prediction: Can it be skillful?
Bull. Amer. Meteor. Soc.
,
90
,
1467
1485
,
doi:10.1175/2009BAMS2778.1
.
Mo
,
K.
, and
J.
Paegle
,
2000
:
Influence of sea surface temperature anomalies on the precipitation regimes over the southwest United States
.
J. Climate
,
13
,
3588
3598
,
doi:10.1175/1520-0442(2000)013<3588:IOSSTA>2.0.CO;2
.
Mo
,
K.
, and
H.-M.
Juang
,
2003
:
Relationships between soil moisture and summer precipitation over the Great Plains and the Southwest
.
J. Geophys. Res.
,
108
, 8610, doi:10.1029/2002JD002952.
Mo
,
K.
,
J.
Paegle
, and
R.
Higgins
,
1997
:
Atmospheric processes associated with summer floods and droughts in the central United States
.
J. Climate
,
10
,
3028
3046
,
doi:10.1175/1520-0442(1997)010<3028:APAWSF>2.0.CO;2
.
Mo
,
K.
,
J.-K.
Schemm
, and
S.-H.
Yoo
,
2009
:
Influence of ENSO and the Atlantic multidecadal oscillation on drought over the United States
.
J. Climate
,
22
,
5962
5982
,
doi:10.1175/2009JCLI2966.1
.
Notaro
,
M.
, and
A.
Zarrin
,
2011
: Sensitivity of the North American monsoon to antecedent Rocky Mountain snowpack. Geophys. Res. Lett.,38, L17403, doi:10.1029/2011GL048803.
Oglesby
,
R.
, and
D.
Erickson
,
1989
:
Soil moisture and the persistence of North American drought
.
J. Climate
,
2
,
1362
1380
,
doi:10.1175/1520-0442(1989)002<1362:SMATPO>2.0.CO;2
.
Pierce
,
D.
,
2002
:
The role of sea surface temperatures in interactions between ENSO and the North Pacific Oscillation
.
J. Climate
,
15
,
1295
1308
,
doi:10.1175/1520-0442(2002)015<1295:TROSST>2.0.CO;2
.
Polade
,
S.
,
A.
Gershunov
,
D.
Cayan
,
M.
Dettinger
, and
D.
Pierce
,
2013
:
Natural climate variability and teleconnections to precipitation over the Pacific–North American region in CMIP3 and CMIP5 models
.
Geophys. Res. Lett.
,
40
, 2296–2301, doi:10.1002/grl.50491.
Raddatz
,
T.
, and Coauthors
,
2007
:
Will the tropical land biosphere dominate the climate–carbon cycle feedback during the twenty-first century?
Climate Dyn.
,
29
,
565
574
,
doi:10.1007/s00382-007-0247-8
.
Richter
,
I.
,
S.-P.
Xie
,
S.
Behera
,
T.
Doi
, and
Y.
Masumoto
,
2014
:
Equatorial Atlantic variability and its relation to mean state biases in CMIP5
.
Climate Dyn.
, 42, 171–188, doi:10.1007/s00382-012-1624-5.
Rotstayn
,
L.
,
M.
Collier
,
M.
Dix
,
Y.
Feng
,
H.
Gordon
,
S.
O’Farrell
,
I.
Smith
, and
J.
Syktus
,
2010
:
Improved simulation of Australian climate and ENSO-related rainfall variability in a global climate model with an interactive aerosol treatment
.
Int. J. Climatol.
,
30
,
1067
1088
, doi:10.1002/joc.1952.
Rudolf
,
B.
, and
U.
Schneider
,
2005
: Calculation of gridded precipitation data for the global land-surface using in-situ gauge observations. Proc. Second Workshop of the Int. Precipitation Working Group, Monterey, CA, IPWG, 231–247.
Rudolf
,
B.
,
H.
Hauschild
,
W.
Rueth
, and
U.
Schneider
,
1994
: Terrestrial precipitation analysis: Operational method and required density of point measurements. Global Precipitations and Climate Change, M. Desbois and F. Désalmand, Eds., Springer, 173–186,
doi:10.1007/978-3-642-79268-7_10
.
Rudolf
,
B.
,
T.
Fuchs
,
U.
Schneider
, and
A.
Meyer-Christoffer
,
2003
: Introduction of the Global Precipitation Climatology Centre (GPCC). GPCC Rep., 16 pp. [Available online at gpcc.dwd.de.]
Rudolf
,
B.
,
C.
Beck
,
J.
Grieser
, and
U.
Schneider
,
2005
: Global precipitation analysis products. Global Precipitation Climatology Centre (GPCC), DWD Rep., 8 pp.
Sakamoto
,
T.
, and Coauthors
,
2012
:
MIROC4h—A new high-resolution atmosphere–ocean coupled general circulation model
.
J. Meteor. Soc. Japan
,
90
,
325
359
,
doi:10.2151/jmsj.2012-301
.
Schneider
,
U.
,
A.
Becker
,
P.
Finger
,
A.
Meyer-Christoffer
,
M.
Ziese
, and
B.
Rudolf
,
2014
: GPCC’s new land surface precipitation climatology based on quality-controlled in situ data and its role in quantifying the global water cycle. Theor. Appl. Climatol.,115, 15–40,
doi:10.1007/s00704-013-0860-x
.
Schubert
,
S.
,
M.
Suarez
,
P.
Pegion
,
R.
Koster
, and
J.
Bacmeister
,
2004a
:
On the cause of the 1930s Dust Bowl
.
Science
,
303
,
1855
1859
,
doi:10.1126/science.1095048
.
Schubert
,
S.
,
M.
Suarez
,
P.
Pegion
,
R.
Koster
, and
J.
Bacmeister
,
2004b
:
Causes of long-term drought in the U.S. Great Plains
.
J. Climate
,
17
,
485
503
,
doi:10.1175/1520-0442(2004)017<0485:COLDIT>2.0.CO;2
.
Schubert
,
S.
,
M.
Suarez
,
P.
Pegion
,
R.
Koster
, and
J.
Bacmeister
,
2008
:
Potential predictability of long-term drought and pluvial conditions in the U.S. Great Plains
.
J. Climate
,
21
,
802
816
,
doi:10.1175/2007JCLI1741.1
.
Schubert
,
S.
, and Coauthors
,
2009
:
A U.S. CLIVAR project to assess and compare the responses of global climate models to drought-related SST forcing patterns: Overview and results
.
J. Climate
,
22
,
5251
5272
,
doi:10.1175/2009JCLI3060.1
.
Seager
,
R.
,
Y.
Kushnir
,
C.
Herweijer
,
N.
Naik
, and
J.
Velez
,
2005
:
Modeling of tropical forcing of persistent droughts and pluvials over western North America: 1856–2000
.
J. Climate
,
18
,
4065
4088
,
doi:10.1175/JCLI3522.1
.
Seager
,
R.
, and Coauthors
,
2007
:
Model projections of an imminent transition to a more arid climate in southwestern North America
.
Science
,
316
,
1181
1184
,
doi:10.1126/science.1139601
.
Seager
,
R.
,
N.
Naik
,
M.
Ting
,
M.
Cane
,
N.
Harnik
, and
Y.
Kushnir
,
2010
:
Adjustment of the atmospheric circulation to tropical Pacific SST anomalies: Variability of transient eddy propagation in the Pacific–North America sector
.
Quart. J. Roy. Meteor. Soc.
,
136
,
277
296
.
Sheppard
,
P.
,
A.
Comrie
,
G.
Packin
,
K.
Angersbach
, and
M.
Hughes
,
2002
:
The climate of the US Southwest
.
Climate Res.
,
21
,
219
238
,
doi:10.3354/cr021219
.
Small
,
E.
,
2001
:
The influence of soil moisture anomalies on variability of the North American monsoon system
.
Geophys. Res. Lett.
,
28
,
139
142
,
doi:10.1029/2000GL011652
.
Smith
,
D.
,
S.
Cusack
,
A.
Colman
,
C.
Folland
,
G.
Harris
, and
J.
Murphy
,
2007
:
Improved surface temperature prediction for the coming decade from a global climate model
.
Science
,
317
,
796
799
,
doi:10.1126/science.1139540
.
Smith
,
D.
,
R.
Eade
,
N.
Dunstone
,
D.
Fereday
,
J.
Murphy
,
H.
Pohlmann
, and
A.
Scaife
,
2010
:
Skilful multi-year predictions of Atlantic hurricane frequency
.
Nat. Geosci.
,
3
,
846
849
,
doi:10.1038/ngeo1004
.
Stahle
,
D.
,
F.
Fye
,
E.
Cook
, and
R.
Griffin
,
2007
:
Tree-ring reconstructed megadroughts over North America since A.D. 1300
.
Climatic Change
,
83
,
133
149
,
doi:10.1007/s10584-006-9171-x
.
Stine
,
S.
,
1994
:
Extreme and persistent drought in California and Patagonia during mediaeval time
.
Nature
,
369
,
546
549
,
doi:10.1038/369546a0
.
Swetnam
,
T.
, and
J.
Betancourt
,
1998
:
Mesoscale disturbance and ecological response to decadal climatic variability in the American Southwest
.
J. Climate
,
11
,
3128
3147
,
doi:10.1175/1520-0442(1998)011<3128:MDAERT>2.0.CO;2
.
Syktus
,
J.
, and Coauthors
,
2011
: The CSIRO-QCCCE contribution to CMIP5 using the CSIRO Mk3.6 climate model. Extended Abstracts, 19th Int. Congress on Modelling and Simulation, Perth, Australia, Modelling and Simulation Society of Australia and New Zealand, 2782–2788.
Tatebe
,
H.
, and Coauthors
,
2012
:
The initialization of the MIROC climate models with hydrographic data assimilation for decadal prediction
.
J. Meteor. Soc. Japan
,
90
,
275
294
,
doi:10.2151/jmsj.2012-A14
.
Taylor
,
K.
,
R.
Stouffer
, and
G.
Meehl
,
2012
:
An overview of CMIP5 and the experiment design
.
Bull. Amer. Meteor. Soc.
,
93
,
485
498
,
doi:10.1175/BAMS-D-11-00094.1
.
Ting
,
M.
, and
H.
Wang
,
1997
:
Summertime U.S. precipitation variability and its relation to Pacific sea surface temperature
.
J. Climate
,
10
,
1853
1873
,
doi:10.1175/1520-0442(1997)010<1853:SUSPVA>2.0.CO;2
.
Trenberth
,
K.
, and
C.
Guillemot
,
1996
:
Physical processes involved in the 1988 drought and 1993 floods in North America
.
J. Climate
,
9
,
1288
1298
,
doi:10.1175/1520-0442(1996)009<1288:PPIITD>2.0.CO;2
.
Trenberth
,
K.
,
G.
Branstator
, and
P.
Arkin
,
1988
:
Origins of the 1988 North American drought
.
Science
,
242
,
1640
1645
,
doi:10.1126/science.242.4886.1640
.
Uppala
,
S.
, and Coauthors
,
2005
:
The ERA-40 Re-Analysis
.
Quart. J. Roy. Meteor. Soc.
,
131
,
2961
3012
,
doi:10.1256/qj.04.176
.
Vera
,
C.
, and Coauthors
,
2006
:
Towards a unified view of the American monsoon systems
.
J. Climate
,
19
,
4977
5000
,
doi:10.1175/JCLI3896.1
.
Voldoire
,
A.
, and Coauthors
,
2013
:
The CNRM-CM5.1 global climate model: Description and basic evaluation
.
Climate Dyn.
,
40
,
2091
2121
,
doi:10.1007/s00382-011-1259-y
.
Volodin
,
E.
,
N.
Dianskii
, and
A.
Gusev
,
2010
:
Simulating present-day climate with the INMCM4.0 coupled model of the atmospheric and oceanic general circulations
.
Izv. Atmos. Ocean. Phys.
,
46
,
414
431
,
doi:10.1134/S000143381004002X
.
Wang
,
H.
, and
M.
Ting
,
2000
:
Covariabilities of winter U.S precipitation and Pacific sea surface temperatures
.
J. Climate
,
13
,
3711
3719
,
doi:10.1175/1520-0442(2000)013<3711:COWUSP>2.0.CO;2
.
Watanabe
,
M.
, and Coauthors
,
2010
:
Improved climate simulation by MIROC5: Mean states, variability, and climate sensitivity
.
J. Climate
,
23
,
6312
6335
,
doi:10.1175/2010JCLI3679.1
.
Watanabe
,
M.
, and Coauthors
,
2011
:
MIROC-ESM: Model description and basic results of CMIP5-20c3m experiments
.
Geosci. Model Dev.
,
4
,
845
872
,
doi:10.5194/gmd-4-845-2011
.
Whitaker
,
J.
,
G.
Compo
,
X.
Wei
, and
T.
Hamill
,
2004
:
Reanalysis without radiosondes using ensemble data assimilation
.
Mon. Wea. Rev.
,
132
,
1190
1200
,
doi:10.1175/1520-0493(2004)132<1190:RWRUED>2.0.CO;2
.
Woodhouse
,
C.
, and
J.
Overpeck
,
1998
:
2000 years of drought variability in the central United States
.
Bull. Amer. Meteor. Soc.
,
79
,
2693
2714
,
doi:10.1175/1520-0477(1998)079<2693:YODVIT>2.0.CO;2
.
Woodhouse
,
C.
,
S.
Gray
, and
D.
Meko
,
2006
: Updated streamflow reconstructions for the Upper Colorado River basin. Water Resour. Res.,42, W05415, doi:10.1029/2005WR004455.
Wu
,
W.
, and
R.
Dickinson
,
2004
:
Time scales of layered soil moisture memory in the context of land–atmosphere interaction
.
J. Climate
,
17
,
2752
2764
,
doi:10.1175/1520-0442(2004)017<2752:TSOLSM>2.0.CO;2
.
Xie
,
P.
, and
P.
Arkin
,
1997
:
Global precipitation: A 17-year monthly analysis based on gauge observations, satellite measurements, and numerical model outputs
.
Bull. Amer. Meteor. Soc.
,
78
,
2539
2558
,
doi:10.1175/1520-0477(1997)078<2539:GPAYMA>2.0.CO;2
.
Xin
,
X.
,
T.
Wu
,
J.
Li
, Z. Wang, W. Li, and F. Wu,
2013
:
How well does BCC-CSM1.1 reproduce the 20th century climate change over China?
Atmos. Oceanic Sci. Lett.
,
6
,
21
26
.
Yu
,
B.
, and
F.
Zwiers
,
2007
:
The impact of combined ENSO and PDO on the PNA climate: A 1,000-year climate modeling study
.
Climate Dyn.
,
29
,
837
851
,
doi:10.1007/s00382-007-0267-4
.
Yu
,
B.
,
A.
Shabbar
, and
F.
Zwiers
,
2007
:
Enhanced PNA-like climate response to Pacific interannual and decadal variability
.
J. Climate
,
20
,
5285
5300
,
doi:10.1175/2007JCLI1480.1
.
Yukimoto
,
S.
, and Coauthors
,
2011
: Meteorological Research Institute–Earth System Model version 1 (MRI-ESM1)—Model description. MRI Tech. Rep. 64, 70 pp. [Available online at http://www.mri-jma.go.jp/Publish/Technical/DATA/VOL_64/index_en.html.]
Zhang
,
Y.
,
J.
Wallace
, and
D.
Battisti
,
1997
:
ENSO-like interdecadal variability: 1900–93
.
J. Climate
,
10
,
1004
1020
,
doi:10.1175/1520-0442(1997)010<1004:ELIV>2.0.CO;2
.
Zhu
,
C.
,
D.
Lettenmaier
, and
T.
Cavazos
,
2005
:
Role of antecedent land surface conditions on North American monsoon rainfall variability
.
J. Climate
,
18
,
3104
3121
,
doi:10.1175/JCLI3387.1
.