Keywords

7.1 Introduction and Data

7.1.1 Overview

The North American Monsoon refers to the land-atmosphere interactions and atmospheric dynamics responsible for the contrast between the pronounced maximum in rainfall during June–September (Fig. 7.1) and dry conditions during October–May over the southwest portion of the continent of North America. Due to the large spatial expanse of the North American Monsoon and its extent across geographic and political borders, the North American Monsoon has been previously referred to as the Mexican Monsoon, the Arizona Monsoon, and the Southwest United States Monsoon. Here, the term “North American Monsoon” (NAM) is all-inclusive.

Fig. 7.1
figure 1

a Mean June–September 1948–2010 precipitation (mm) and b percent contribution of June–September precipitation to the annual total

Full size image

During June–September, long-term average precipitation totals in excess of 550 mm (Fig. 7.1a) are associated with the NAM along the spine of the Sierra Madre Occidental, extending between Nayarit and Sonora, and accounting for more than 70 % of the annual average precipitation (Fig. 7.1b). Lesser precipitation amounts, in the order of 200–350 mm, fall north of 30°N, in areas such as northwest Mexico, southern Arizona, and southern Utah (Fig. 7.1).

Due to the large spatial extent of the NAM region, the climatological onset and ending dates of the NAM greatly vary region-wide. Over Mexico, the onset of the NAM is during late May and early June (Vera et al. 2006). Over the southwest United States, the onset of the NAM is during early July (Fig. 7.2b), as areas such as Arizona and New Mexico observe an average onset date of July 3–14 (Vera et al. 2006). The maximum monthly average precipitation falls during July and August (Fig. 7.2b, c) and the termination of the NAM-related rains occurs regionally during September, first over the southwestern United States, and finally over northwest Mexico.

Fig. 7.2
figure 2

Mean 1948–2010 precipitation during a June, b July, c August and d September

Full size image

Multi-decadal climate variability during the twentieth century has been observed in terms of long-term changes in regional precipitation over southwestern North America (e.g., McCabe 2004; Seager et al. 2005; Weiss et al. 2009; Seager and Hoerling 2014). During June–September over the NAM region, a later onset and withdrawal of the NAM-related precipitation has been observed over the Southwest United States in recent decades (Grantz et al. 2007). Future climate projections using the Coupled Model Intercomparison Project Phase 5 (CMIP5) models, as part of the Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report (AR5), support the observed analyses of a later shift in the monsoon season in the presence of increased greenhouse gas concentrations in the atmosphere (Cook and Seager 2013). In this chapter, we review the recent literature examining the observed multi-decadal changes in the June–September rains associated with the NAM during the last half of the twentieth century, and possible projected changes of the NAM associated with increased greenhouse gas concentrations in the atmosphere in the future.

7.1.2 Data

Observed monthly precipitation was drawn from the Global Precipitation Climatology Centre (GPCC) precipitation dataset for 1948–2010 on a 0.5° × 0.5° fixed latitude-longitude grid (Becker et al. 2013). Observed atmospheric dynamics, shown in terms of stream function, pressure vertical velocity, and vertically integrated moisture flux, were drawn from the 6 hourly National Center for Environmental Prediction —National Center for Atmospheric Research (NCEP-NCAR) reanalysis for 1948–2010 on a 2.5° × 2.5° fixed latitude-longitude grid (Kalnay et al. 1996).

Possible changes of the June–September rains associated with the NAM for the duration of the current century were drawn from the CMIP5 RCP8.5 scenario using the methods of Cook and Seager (2013); they analyzed 41 ensemble simulations over 11 coupled global climate models from the RCP8.5 scenario for 2005–2099, and the historical scenario for 1980–2004. The climate of the RCP8.5 simulations was forced with strong concentrations of greenhouse gases in the atmosphere, while the historical simulations are forced by observed conditions within the instrumental period, such as incoming solar radiation and greenhouse gas concentrations. The models used and the number of ensembles included for each model is shown in Table 1 of Cook and Seager (2013).

7.2 Mean Monsoon Dynamics

The NAM system is usually considered to include the monsoon-like heavy summer rainfall centered in northwest Mexico and extending into the southwestern United States, as well as its regional continental linkages in circulation and precipitation. The basic aspects of the NAM system have been examined in several papers (Webster 1987; Douglas et al. 1993; Adams and Comrie 1997; Higgins et al. 1997; Barlow et al. 1998; Higgins et al. 1999; Higgins and Gochis 2007), and consist of a core area of precipitation with monsoonal characteristics—including a single precipitation season led by the highest surface temperatures, seasonal reversal of winds , deep tropical heating associated with an upper-level anticyclone, and associated changes to circulation over much of the United States (and to precipitation in the central United States in particular). Although the winds in the Gulf of California undergo a seasonal reversal, the change is not of a sufficient magnitude and scale to meet the criteria of Ramage (1971) for a monsoon, at least in existing large-scale data products. Recent work has suggested that the circulation response to the monsoon onset may, counter-intuitively, play a role in limiting its northward advance once it reaches the westerlies (Saini et al. 2013). In a larger-scale context, the NAM system is part of a regional monsoon system (Vera et al. 2006) which, in turn, is part of the global monsoon system (Trenberth et al. 2000; Wang and Ding 2008). Here, we briefly discuss the dynamical mechanisms responsible in forcing the June–September rainfall associated with the NAM before reviewing the recent and possible future changes of the NAM, the primary topic of this chapter.

The NAM-related reversal of seasonal wind responsible for the copious precipitation during June–September over southwest North America is forced by the surface pressure difference between the land surface and the adjacent Pacific Ocean. During the Northern Hemisphere summer, intense solar radiation heats the land surface faster than the ocean due to differences in the capacity of each body to store energy. As a result of the temperature contrast between the land and ocean, a surface low pressure area (often referred to as a “thermal low” (Fig. 7.3) develops along the northwest coast of Mexico, resulting in onshore flow and a flux of moisture toward land from the Gulf of California and the Gulf of Mexico (Fig. 7.4).

Fig. 7.3
figure 3

Mean 1948–2010 stream function during a June, b July, c August and d September at 500 hPa (shading) and 1000 hPa (contour) in units of m2 s−1

Full size image
Fig. 7.4
figure 4

Mean 1948–2010 vertically integrated moisture flux (vector; m s−1) and vertically integrated moisture flux divergence (s−1) during a June, b July, c August and d September

Full size image

The rainy season associated with the NAM begins in late May and early June over southern Mexico (Fig. 7.2a). The band of rain penetrates northward over the Sierra Madre Occidental during the months of June–August (Fig. 7.2a–c), associated with the northward progression of the thermal low pressure area (contours in Fig. 7.3a–d) caused by increased land-ocean differential heating, anticylonic circulation in the middle troposphere (shading in Fig. 7.3a–d), and reduced transient storm activity over the United States. The NAM-related precipitation begins over Northwest Mexico during June and over the southwest United States during early July. The vertical movement of moisture by convection supplies the moisture required for rainfall during June–September (Fig. 7.5). The northernmost extent of the NAM-related rains reaches Arizona and New Mexico, but rains over these areas are much lighter in comparison to the elevated parts of the Pacific coastal areas of Mexico.

Fig. 7.5
figure 5

Mean 1948–2010 pressure vertical velocity during a June, b July, c August and d September at 400 hPa in units of hPa s−1

Full size image

The moisture source responsible for NAM-related rainfall remains unclear. There are two hypotheses: one that the primary moisture source is from the Gulf of Mexico (Bryson and Lowry 1955; Sellers and Hill 1974) and the other that moisture originates from the Gulf of California. Schmitz and Mullen (1996) showed that moisture responsible for NAM-related precipitation is provided by both the Gulf of Mexico and the Gulf of California, as upper tropospheric moisture is supplied from the Gulf of Mexico and lower tropospheric moisture is supplied from the Gulf of California. As shown by vectors in Fig. 7.4, the horizontal transport of vertically-integrated moisture does not show a clear source when averaged monthly, especially during June and July. However, despite the weak flux of moisture around the rainiest areas associated with the NAM, there is strong horizontal moisture flux convergence during June–September over intense precipitation areas (shading in Fig. 7.4).

During late summer, the NAM-related precipitation weakens and regresses southward in an opposing pattern to the onset phase. The decrease of precipitation is associated with weakening middle tropospheric high pressure over Texas and New Mexico (Fig. 7.3d) and increased precipitation over adjacent areas. The NAM-related rainfall ends over the southwest United States during September.

7.3 Observed and Projected Changes

Significant precipitation increases during June–September have been observed over Arizona and New Mexico (Anderson et al. 2009); however, no significant precipitation trends have been observed over the rest of the NAM region (Anderson et al. 2009; Arriaga-Ramírez and Cavazos 2010). The regional variation of the long-term NAM-related precipitation trend has been attributed to changes in intensity and frequency of precipitation episodes, and to changes in the length of the NAM-related precipitation season (Englehart and Douglas 2006). Western Mexico precipitation during June–September is perhaps influenced differently from the rest of the core NAM region because of differences in moisture transports and preceding hydrological conditions (Gochis et al. 2007). In terms of the timing of NAM-related precipitation, Grantz et al. (2007) identified delays in the onset , peak, and cessation of precipitation over the southwestern United States, which implies a later shift in seasonality of the monsoon. Here, we investigate the linear trends in atmospheric pressure (in terms of stream function), moisture flux, and vertical motions for June–September 1948–2010 to examine the possible long-term forcing of NAM-related precipitation.

The linear trend in June–September 1948–2010 precipitation is shown in Fig. 7.6, and indicates significant precipitation increases over New Mexico and the core NAM region, and significant precipitation decreases over southwest Mexico. The significant precipitation changes over the southwestern United States and southwest Mexico are corroborated by Anderson et al. (2009) and Gochis et al. (2007), respectively.

Fig. 7.6
figure 6

Linear precipitation trend during June–September 1948–2010 contoured every 15 mm. Trends significant at p < 0.10 using a t-test are shaded

Full size image

The linear 1948–2010 trend in NAM-related June–September precipitation is forced by the regional teleconnection generated by anomalous cyclonic circulation (Fig. 7.7) centered at 15°N latitude over the eastern Pacific Ocean. The anomalous cyclonic circulation, a low pressure area, is responsible for the significant trend in ascending air over southern Mexico and the eastern tropical Pacific Ocean (Fig. 7.8), which does not appear to have a strong influence on the NAM-related precipitation during June–September.

Fig. 7.7
figure 7

Linear 1000 hPa trend during June–September 1948–2010 contoured every 1 × 106 m2 s−1. Trends significant at p < 0.10 using a t-test are shaded

Full size image
Fig. 7.8
figure 8

Linear 500 hPa pressure vertical velocity trend during June–September 1948–2010 contoured every 1 × 10−2 Pa s−1. Trends significant at p < 0.10 using a t-test are shaded

Full size image

The anomalous cyclonic circulation is clear in the 1948–2010 trend of vertically integrated moisture flux (vectors in Fig. 7.9). The circulation on the eastern side of the anomalous cyclone is responsible for the significant trend in vertically integrated moisture flux convergence (shading in Fig. 7.9) over the eastern tropical Pacific, northwestern Mexico, Gulf of California, and Baja California, which forces the increasing regional precipitation trends (Fig. 7.6) during June–September. Alhough moisture flux divergence is not shown in the 1948–2010 trend over southwestern Mexico for June–September, the diffluence of the trend in vertically integrated moisture flux helps to explain why a significant decreasing trend in precipitation is observed (Fig. 7.6). Off of the coast of southwestern Mexico, the curvature of the trend in vertically integrated moisture flux is tight and cyclonic (a southeast wind), but just to the east of the curvature is a southerly flow that results in diffluence and contributes to precipitation reductions.

Fig. 7.9
figure 9

Linear vertically integrated moisture flux trend (vector; m s−1) and moisture flux divergence trend (contour; 1 × 10−5 s−1) during June–September 1948–2010 contoured every 1 × 10−5 s−1. Trends of vertically-integrated moisture flux significant at p < 0.10 using a t-test are shaded

Full size image

Cook and Seager (2013) examined the RCP8.5 scenarios of 41 ensemble simulations from 11 coupled global climate models to determine whether the recent observed changes that support a later start and end to the NAM-related rains were reproduced in a climate with increased atmospheric greenhouse gas concentrations. Figure 7.10 shows the area-averaged precipitation time series from these same 41 simulations (see Cook and Seager 2013 for details) from 1980 to 2099 over the core NAM region of Cook and Seager (2013), defined as a box bound by the points 18°N–33°N, and 112°W–102°W (red boxes in Figs. 7.1 and 7.2). In Fig. 7.10, the historical simulations are shown for 1980–2004, while the RCP8.5 simulations are shown for 2005–2099.

Fig. 7.10
figure 10

Ensemble average CMIP5 precipitation projections from the RCP8.5 scenario averaged over the boxed regions in Figs. 7.1–7.2 (18°N–33°N; 112°W–102°W) in mm day−1. (According to Cook and Seager 2013). a June. b July. c September. d October

Full size image

The CMIP5 precipitation declines over the core NAM region for 1980–2099 are around 15 % during June and July, and the linear least squares regression decreases for each month are significant at p < 0.05 (Cook and Seager 2013). During June, the precipitation decreases are strong across the entire core NAM region as well as Baja California, Arizona, and New Mexico (Fig. 7.7 of Cook and Seager 2013). During July, the precipitation decreases are strongest along the Sierra Madre Occidental and along the shores of the Gulf of California, but are mixed elsewhere within the core NAM region (Fig. 7.7 of Cook and Seager 2013).

The CMIP5 precipitation increases over the core NAM region for 1980–2099 are around 14 and 6 % during September and October, respectively. The linear least squares regression increases during both September and October are significant at p < 0.05 (Cook and Seager 2013). During September and October, the precipitation increases are strongest over southeastern portions of the core NAM region, while precipitation changes are weak elsewhere (Fig. 7.7 of Cook and Seager 2013).

Reduced precipitation over the core NAM region during June and July, and enhanced precipitation over the core NAM region during September and October, imply a shift to a later monsoon in the presence of increased greenhouse gases within the RCP8.5 scenarios of the CMIP5 simulations. The decreases in rainfall during June and July in the RCP8.5 scenarios are caused by increases in the vertical stability of the atmosphere due to atmospheric warming and reduced available moisture at the surface (Cook and Seager 2013), despite an increase in the temperature contrast between the continental and oceanic areas (Torres-Alvarez et al. 2014). The increases in rainfall during September and October in the RCP8.5 scenarios are forced by increases in moisture convergence (Cook and Seager 2013).

Global climate models capture the large-scale circulation and seasonal cycle of the NAM (Liang et al. 2008; Gutzler 2009). However, due to the complexity of the NAM and its sensitivity to processes that are poorly resolved by global climate models, global climate models poorly simulate the finer-scale features of the NAM (Castro et al. 2007, Cerezo-Mota et al. 2011). The coupled climate system models have made no improvements in their simulations of the NAM between the CMIP Phase 3 (CMIP3) and CMIP Phase 5 (CMIP5) activities (Geil et al. 2013). Therefore, future projections of the NAM, using modeling approaches, must be interpreted with caution.