• Arkin, P. A., 1982: The relationship between interannual variability in the 200-mb tropical wind field and the southern oscillation. Mon. Wea. Rev.,110, 1393–1404.

  • Bonner, W. D., 1968: Climatology of the low-level jet. Mon. Wea. Rev.,96, 833–850.

  • Carleton, A. M., 1986: Synoptic-dynamic character of “bursts” and “breaks” in the southwest U.S. summer precipitation singularity. J. Climatol.,6, 605–623.

  • Charba, J. P., A. W. Harrell III, and A. C. Lackner III, 1992: A monthly precipitation amount climatology derived from published atlas maps: Development of a digital database. TDL Office Note 92-7, 20 pp. [Available from TDL Office, NOAA, U.S. Department of Commerce, 1325 East–West Hwy., Silver Spring, MD 20910.].

  • Cressman, G. P., 1959: An operational objective analysis system. Mon. Wea. Rev.,87, 367–374.

  • Dhar, O. N., P. R. Rakhecha, and B. N. Mandal, 1980: Does the early or late onset of the monsoon provide any clue to the subsequent rainfall during the monsoon season? Mon. Wea. Rev.,108, 1069–1072.

  • Douglas, M. W., R. A. Maddox, K. Howard, and S. Reyes, 1993: The Mexican Monsoon. J. Climate,6, 1665–1677.

  • Glahn, H. R., T. L. Chambers, W. S. Richardson, and H. P. Perrotti, 1985: Objective map analysis for the local AFOS MOS Program. NOAA Tech. Memo. NWS TDL 75, 34 pp. [Available from NOAA, U.S. Department of Commerce. 1325 East–West Hwy., Silver Spring, MD 20910.].

  • Helfand, H. M., and S. D. Schubert, 1995: Climatology of the Great Plains low-level jet and its contribution to the continental moisture budget of the United States. J. Climate,8, 784–806.

  • Higgins, R. W., J. E. Janowiak, and Y. Yao, 1996: A Gridded Hourly Precipitation Data Base for the United States (1963–1993). NCEP/Climate Prediction Center Atlas 1, 47 pp.

  • ——, Y. Yao, E. S. Yarosh, J. E. Janowiak, and K. C. Mo, 1997a: Influence of the Great Plains low-level jet on summertime precipitation and moisture transport over the central United States. J. Climate,10, 481–507.

  • ——, ——, and X. Wang, 1997b: Influence of the North American Monsoon System on the United States summer precipitation regime. J. Climate,10, 2600–2622.

  • ——, K. C. Mo, and Y. Yao, 1998: Interannual variability of the United States summer precipitation regime with emphasis on the southwestern monsoon. J. Climate,11, 2582–2606.

  • Janowiak, J. E., A. Gruber, C. R. Kondragunta, R. E. Livezey, and G. J. Huffman, 1998: A comparison of the NCEP–NCAR Reanalysis Precipitation and the GPCP Raingauge–Satellite combined dataset with observational error considerations. J. Climate,11, 2960–2979.

  • Johnson, A. M., 1976: The climate of Peru, Bolivia and Ecuador. World Survey of Climatology. Vol. 12, Climates of Central and South America, W. Schwerdtfeger and H. E. Landsberg, Eds., Elsevier, 147–218.

  • Kalnay, E., and Coauthors, 1996: The NCEP–NCAR Reanalysis Project. Bull. Amer. Meteor. Soc.,77, 437–471.

  • Kiladis, G. N., and H. Van Loon, 1988: The Southern Oscillation. Part VII: Meteorological anomalies over the Indian and Pacific sectors associated with extremes of the oscillation. Mon. Wea. Rev.,116, 120–136.

  • Kousky, V. E., and C. F. Ropelewski, 1997: The Tropospheric Seasonally Varying Mean Climate over the Western Hemisphere (1979–1995). NCEP/Climate Prediction Center Atlas 3, 135 pp.

  • Livezey, R. E., 1995: Field intercomparison. An Analysis of Climate Variability Applications of Statistical Techniques, H. Von Storch and A. Navara, Eds., Springer-Verlag, 159–176.

  • Meehl, G. A., 1994: Influence of the land surface in the Asian summer monsoon: External conditions versus internal feedbacks. J. Climate,7, 1033–1049.

  • ——, and W. A. Washington, 1988: A comparison of soil moisture sensitivity in two global climate models. J. Atmos. Sci.,45, 1476–1492.

  • Mo, K. C., J. N. Paegle, and R. W. Higgins, 1997: Atmospheric processes associated with summer floods and droughts in the central United States. J. Climate,10, 3028–3046.

  • Mock, C. J., 1996: Climatic controls and spatial variations of precipitation in the western United States. J. Climate,9, 1111–1125.

  • Okabe, I. T., 1995: The North American Monsoon. Ph.D. dissertion, University of British Columbia, 146 pp.

  • Parrish, D. F., and J. C. Derber, 1992: The National Meteorological Center’s spectral statistical interpolation analysis system. Mon. Wea. Rev.,120, 1747–1763.

  • Ramage, C. S., 1971: Monsoon Meteorology. Academic Press, 391 pp.

  • Ropelewski, C. F., and M. S. Halpert, 1986: North American precipitation and temperature patterns associated with El Niño–Southern Oscillation (ENSO). Mon. Wea. Rev.,114, 2352–2362.

  • ——, and ——, 1996: Quantifying southern oscillation–precipitation relationships. J. Climate,9, 1043–1059.

  • Smith, T. M., R. W. Reynolds, R. E. Livezey, and D. C. Stokes, 1996:Reconstruction of historical sea surface temperatures using empirical orthogonal functions. J. Climate,9, 1403–1420.

  • Stensrud, D. J., R. L. Gall, S. L. Mullen, and K. W. Howard, 1995:Model climatology of the Mexican monsoon. J. Climate,8, 1775–1794.

  • Tang, M., and E. R. Reiter, 1984: Plateau monsoons of the Northern Hemisphere: A comparison between North America and Tibet. Mon. Wea. Rev.,112, 617–637.

  • Wallace, J. M., 1975: Diurnal variations in precipitation and thunderstorm frequency over the conterminous United States. Mon. Wea. Rev.,103, 406–419.

  • Webster, P. J., 1987: The elementary monsoon. Monsoons, J. S. Fein and P. L. Stephens, Eds., John Wiley and Sons, 3–32.

  • ——, and S. Yang, 1992: Monsoon and ENSO: Selectively interactive systems. Quart. J. Roy. Meteor. Soc.,118, 877–926.

  • View in gallery

    Typical station distribution for hourly reporting stations in the United States and daily reporting stations in Mexico used in the US–MEXICO merged precipitation dataset. Gridlines represent the (33 × 26) grid to which the station data have been analyzed. The topography of the region is also included, courtesy of the United States National Geophysical Data Center. The resolution of the data is 5 min (0.083°). The topography data are available online at http://www:ngdc.noaa.gov/mgg/mggd.html. Shading intervals are at 500, 1000, 2000, and 3000 m.

  • View in gallery

    Histograms of the mean (1963–88) daily and 5-day running mean precipitation (units: mm day−1) during May–August at grid points in the southwestern United States and Mexico from the merged US–MEXICO precipitation database. Grid points used for the AZNM, NWMEX, and SWMEX precipitation indices (defined in section 2b) are indicated by the “#,” “*,” and “+” symbols, respectively. The grid points at which the data are valid are located at the center of each box. The topography of the region is also included, as described in Fig. 1.

  • View in gallery

    Mean (1963–88) daily and 5-day running mean area averaged precipitation (units: mm day−1) for the (a) AZNM, (b) NWMEX, and (c) SWMEX regions.

  • View in gallery

    Evolution of the composite mean (1963–88) daily precipitation (mm day−1) for the (a) AZNM, (b) NWMEX, and (c) SWMEX regions relative to monsoon onset.

  • View in gallery

    (a) Mean (1963–88) seasonal precipitation (units: mm) for July–September from the US–MEXICO merged analysis. The contours are 50, 100, 200, 400, 600, 800, and 1000 mm and values greater than 100 mm are shaded. (b) Contribution of the precipitation during July–September to the annual total, expressed in percent, from the US–MEXICO merged precipitation analysis. The contour interval is 5% and values greater than 40% are shaded.

  • View in gallery

    Analysis of the contribution of the mean (1963–88) monthly precipitation to the annual mean (units: percent) over the conterminous United States and Mexico for each month from the US–MEXICO merged analysis. The contour interval is 5% and areas with values exceeding 15% are shaded.

  • View in gallery

    Mean (1968–88) monthly 925-hPa vector wind (m s−1), 200-hPa streamlines, and US–MEXICO precipitation (shading) for (a) May, (b) June, (c) July, and (d) August. Circulation data are from the NCEP–NCAR Reanalysis. A topography mask has been applied to the 925-hPa winds. Precipitation amounts are in mm day−1 and values greater than 1 mm day−1 are shaded. The characteristic vector length is 10 m s−1.

  • View in gallery

    Mean (1968–88) monthly 925-hPa vector wind (m s−1), 200-hPa streamlines, and US–MEXICO precipitation (shading) represented as a difference between consecutive months for (a) May–April, (b) June–May, (c) July–June, and (d) August–July. A topography mask has been applied to the 925-hPa winds. The characteristic vector length is 2 m s−1 and precipitation differences >0.5 mm day−1 (<−0.5 mm day−1) are shaded dark (light).

  • View in gallery

    Histograms of mean (1963–94) monthly precipitation (mm day −1) from selected grid points over the conterminous United States and Mexico: (a) (34°N, 110°W) in Arizona, (b) (30°N, 97.5°W) in Texas, (c) (36°N, 95°W) in Oklahoma, (d) (46°N, 110°W) in Montana, (e) (28°N, 110°W) in Sonora, (f) (20°N, 105°W) in Jalisco, and (g) (16°N, 97.5°W) in Oaxaca.

  • View in gallery

    Standard deviation of (a) the seasonal (June–September) mean precipitation (mm) and (b) the seasonal (June–September) percent departure from normal (units: percent) based on 26 yr (1963–88) of data. In (a) the contours are 50, 100, and 200 mm and values greater than 50 mm are shaded. In (b) the contour interval is 10% and values greater than 20% are shaded. In each case, areas where the seasonal precipitation is less than 30 mm are masked.

  • View in gallery

    Seasonal (June–September) percent departure from normal precipitation (%) for (a) AZNM, (b) NWMEX, and (c) SWMEX based on 26 yr (1963–88) of data. Anomalies are computed with respect to the apropriate area mean for the 26-yr period. On panel (c) the years with mature cold or warm episode conditions are indicated by a C or a W, respectively.

  • View in gallery

    Mean (1963–88) calendar date of onset for the summer monsoon at each grid point used in the AZNM, NWMEX, and SWMEX precipitation indices.

  • View in gallery

    Time series of the date of onset for the summer monsoon in (a) AZNM, (b) NWMEX, and (c) SWMEX for the period 1963–88. The mean onset date for each region is indicated by a solid line and one standard deviation is indicated by dashed lines. All dates are Julian dates in the calendar year (e.g., day 182 = 1 July, etc.).

  • View in gallery

    Composite evolution of monthly precipitation (mm day−1) over (a) AZNM, (b) NWMEX, and (c) SWMEX for early (dashed line), late (dash-dot line), and all (1963–88) monsoons (thick solid line).

  • View in gallery

    Time series of accumulated precipitation anomalies [departures from mean (1963–88) daily values] in SWMEX (dash-dot line), NWMEX (dashed line), and AZNM (solid line) for the 60-day period after monsoon onset in (a) SWMEX, (b) NWMEX, and (c) AZNM.

  • View in gallery

    Correlation between rainfall for the 60-day period after monsoon onset in (a) SWMEX, (b) NWMEX, and (c) AZNM and rainfall over Mexico and the United States. The contour interval is 10%, the zero contour is omitted for clarity. Regions where the correlation is locally significant at the 95% confidence level are shaded.

  • View in gallery

    Composite evolution of monthly precipitation (mm day−1) over (a) AZNM, (b) NWMEX, and (c) SWMEX for wet (dashed line), dry (dash-dot line), and all (1963–88) monsoons (thick solid line). Composite evolution of monthly precipitation (mm day−1) over (d) AZNM, (e) NWMEX, and (f) SWMEX for La Niña (dashed line) and El Niño (dash-dot) years, and all (1963–88) years (thick solid line).

  • View in gallery

    Maps of the composite seasonal (JJAS) precipitation anomalies (mm day−1) for (a) wet monsoons in AZNM, (b) dry monsoons in AZNM, (c) wet monsoons in NWMEX, (d) dry monsoons in NWMEX, (e) wet monsoons in SWMEX, and (f) dry monsoons in SWMEX. In each case the anomalies are departures from the mean (1963–88) seasonal values. The contour interval is 0.25 mm day−1. Dark (light) shading indicates areas where seasonal precipitation anomalies are greater than 10% of the mean and positive (negative). Areas where the seasonal precipitation is less than 30 mm are masked.

  • View in gallery

    Same as Fig. 18, except for (a) El Niño and (b) La Niña.

  • View in gallery

    Composite seasonal (JJAS) 200-hPa wind anomalies, 200-hPa streamfunction anomalies, and US–MEXICO precipitation anomalies for (a) wet monsoons in AZNM, (b) dry monsoons in AZNM, (c) wet monsoons in NWMEX, (d) dry monsoons in NWMEX, (e) wet monsoons in SWMEX, and (f) dry monsoons in SWMEX. All fields are represented as departures from mean (1968–88) seasonal values. The contour interval is 1×106m2 s−1 and the standard vetor length is 10 m s−1. Precipitation anomalies >0.25 mm day−1 (<−0.25 mm day−1) are shaded dark (light).

  • View in gallery

    Same as Fig. 20, except for (a) El Niño and (b) La Niña.

  • View in gallery

    Composite seasonal (JJAS) sea surface temperature anomalies [departures from mean (1963–88) seasonal values] represented as the difference between wet and dry monsoons in (a) AZNM, (b) NWMEX, and (c) SWMEX, and (d) as the difference between El Niño and (b) La Niña. In each panel the dark (light) shading indicates positive (negative) anomalies significant at the 90% confidence level. The contour interval is 0.2°C.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 324 313 19
PDF Downloads 134 129 15

Interannual Variability of the North American Warm Season Precipitation Regime

View More View Less
  • 1 Climate Prediction Center, NOAA/NWS/NCEP, Washington, D.C.
  • | 2 Research and Data Systems Corporation, Greenbelt, Maryland
  • | 3 Department of Atmospheric Sciences, Creighton University, Omaha, Nebraska
© Get Permissions
Full access

Abstract

Interannual variability of the North American warm season–precipitation regime is examined in three regions of the United States and Mexico: Arizona–New Mexico, northwest Mexico, and southwest Mexico. Daily observed precipitation over the United States and Mexico for a 26-yr (1963–88) period and various fields from the National Centers for Environmental Prediction–National Center for Atmospheric Research Reanalysis are used to compare and contrast hydrologic conditions and atmospheric circulation features associated with early, late, wet, and dry monsoons in each region. Relationships between anomalous monsoon behavior and the El Niño–Southern Oscillation phenomenon are examined. Some factors associated with the atmosphere’s lower boundary conditions that might influence the interannual variability of the warm season precipitation regime are discussed.

The mean seasonal evolution of the North American monsoon system is characterized by the regular northward progression of heavy precipitation from southern Mexico by early June to the southwestern United States by early July. While the seasonal normal rainfall and its variability are largest in southwest Mexico, the mean seasonal percent departure from normal is largest in Arizona–New Mexico. Wet (dry) monsoons in southwest Mexico tend to occur during La Niña (El Niño). This association is attributed, in part, to the impact of local sea surface temperature anomalies on the land–sea thermal contrast, hence the strength of the monsoon. There is also a weak association between dry monsoons in Arizona–New Mexico (northwest Mexico) and La Niña (El Niño). Wet summer monsoons in Arizona–New Mexico tend to follow winters characterized by dry conditions in the southwestern United States and vice versa.

Although the onset and duration of the monsoon are quite regular in each region, the precise date of onset in a given region is highly variable and likely to be unrelated to the date of onset in the other regions. Early monsoons in Arizona–New Mexico tend to have heavy seasonal rainfall while late monsoons in northwest Mexico tend to have deficient seasonal rainfall. The onset date in southwest Mexico is not related to seasonal rainfall. However, interannual fluctuations in rainfall over the entire monsoon region for the 2-month period after onset in southwest Mexico are highly correlated, suggesting that knowledge of the starting date in southwest Mexico may be useful for analyzing, understanding, and possibly predicting these fluctuations.

Corresponding author address: Dr. R. W. Higgins, Climate Prediction Center, NWS/NCEP W/NP52, 5200 Auth Road, Room 605, Washington, DC 20233-9910.

Email: wd52wh@sgi85.wwb.noaa.gov

Abstract

Interannual variability of the North American warm season–precipitation regime is examined in three regions of the United States and Mexico: Arizona–New Mexico, northwest Mexico, and southwest Mexico. Daily observed precipitation over the United States and Mexico for a 26-yr (1963–88) period and various fields from the National Centers for Environmental Prediction–National Center for Atmospheric Research Reanalysis are used to compare and contrast hydrologic conditions and atmospheric circulation features associated with early, late, wet, and dry monsoons in each region. Relationships between anomalous monsoon behavior and the El Niño–Southern Oscillation phenomenon are examined. Some factors associated with the atmosphere’s lower boundary conditions that might influence the interannual variability of the warm season precipitation regime are discussed.

The mean seasonal evolution of the North American monsoon system is characterized by the regular northward progression of heavy precipitation from southern Mexico by early June to the southwestern United States by early July. While the seasonal normal rainfall and its variability are largest in southwest Mexico, the mean seasonal percent departure from normal is largest in Arizona–New Mexico. Wet (dry) monsoons in southwest Mexico tend to occur during La Niña (El Niño). This association is attributed, in part, to the impact of local sea surface temperature anomalies on the land–sea thermal contrast, hence the strength of the monsoon. There is also a weak association between dry monsoons in Arizona–New Mexico (northwest Mexico) and La Niña (El Niño). Wet summer monsoons in Arizona–New Mexico tend to follow winters characterized by dry conditions in the southwestern United States and vice versa.

Although the onset and duration of the monsoon are quite regular in each region, the precise date of onset in a given region is highly variable and likely to be unrelated to the date of onset in the other regions. Early monsoons in Arizona–New Mexico tend to have heavy seasonal rainfall while late monsoons in northwest Mexico tend to have deficient seasonal rainfall. The onset date in southwest Mexico is not related to seasonal rainfall. However, interannual fluctuations in rainfall over the entire monsoon region for the 2-month period after onset in southwest Mexico are highly correlated, suggesting that knowledge of the starting date in southwest Mexico may be useful for analyzing, understanding, and possibly predicting these fluctuations.

Corresponding author address: Dr. R. W. Higgins, Climate Prediction Center, NWS/NCEP W/NP52, 5200 Auth Road, Room 605, Washington, DC 20233-9910.

Email: wd52wh@sgi85.wwb.noaa.gov

1. Introduction

Monsoon circulation systems, which develop over low latitude continental regions in response to thermal contrast between the continent and adjacent oceanic regions, are a major component of continental warm season precipitation regimes (e.g., see the review by Webster 1987). These systems are characterized by seasonal reversals of the circulation and precipitation regimes (e.g., Ramage 1971) with many links to weather and climate fluctuations (e.g., Kiladis and van Loon 1988; Webster and Yang 1992). Much of North America is characterized by such a monsoon system (hereafter referred to as the North American monsoon system or NAMS). This system provides a useful framework for describing and diagnosing warm season climate controls and the nature and causes of year-to-year variability (e.g., Higgins et al. 1997b). This system displays many similarities (as well as differences) with its Asian counterpart (e.g., Tang and Reiter 1984). While the NAMS is less impressive than its Asian sister on a global scale, it still has a tremendous impact on local climate. It is a specific goal of this study to improve our understanding of the interannual variability of the warm season precipitation regime over North America as it relates to the evolution of the NAMS.

Of significance for the understanding of the warm season precipitation regime of North America is the fact that the NAMS affects much of the United States and Mexico (e.g., Higgins et al. 1997b). Over the United States there is evidence of a continental-scale mode in the warm season precipitation pattern consisting of an out of phase relationship between the Southwest and the Great Plains and an in phase relationship between the Southwest and the East Coast (e.g., Higgins et al. 1997b;Mo et al. 1997). Okabe (1995) has shown that phase reversals in this pattern are related to the development and decay of the monsoon. A detailed description of the life cycle of the NAMS in terms of development, mature and decay phases, and a literature review of the major elements of the NAMS are given in Higgins et al. (1997b).

Knowledge of the start or “onset” of the NAMS is of considerable importance since, for example, it represents one key to the timing of the planting of crops. The onset is usually sudden, with the weather in the monsoon region changing abruptly from relatively hot, dry conditions to cool, rainy ones. The interannual variability of the monsoon modulates the annual cycle to occasionally produce years with flood or drought. While the interannual variability of the NAMS is important, the difference between the monsoon flood and drought years is smaller than the difference between the weather in the preonset and the postonset period (e.g., Webster 1987). In spite of the highly periodic nature of the NAMS, there are also large variations in the circulation and rainfall within the monsoon season, often referred to as “active” and “break” periods. The intensity of the seasonal mean monsoon is influenced by the nature of variability within the monsoon season. Previous attempts to relate rainfall anomalies for the monsoon season to the date of onset of the Indian monsoon (e.g., Dhar et al. 1980) have generally shown little relationship indicating that the intraseasonal variability of monsoon rainfall is quite large.

We extend our earlier work by diagnosing the interannual variability of the North American warm season precipitation in three regions of Mexico and the United States: Arizona–New Mexico (AZNM), northwest Mexico (NWMEX), and southwest Mexico (SWMEX). The variability of monsoon rainfall in each region is studied using observed daily precipitation over the United States and Mexico for a 26-yr (1963–88) period. Composites of observed precipitation and various fields from the National Centers for Environmental Prediction (NCEP)–National Center for Atmospheric Research (NCAR) Reanalysis are used to compare and contrast hydrologic conditions and atmospheric circulation features for early, late, wet, and dry monsoons in each region; criteria used to identify these monsoon characteristics are discussed in detail in sections 2b, 5, and 6. In addition, relationships between El Niño–Southern Oscillation (ENSO) and these monsoon characteristics are explored.

To advance the seasonal prediction of warm season precipitation over North America requires a better understanding of the physical processes that govern the time-dependent behavior of the monsoon system. To this end, this study also identifies some factors that influence the interannual variability of the NAMS. In the case of the Asian summer monsoon, it is well known that variations in land–sea temperature contrast exert a strong control on the strength of the monsoon circulation, hence the start of the monsoon. A number of factors have been linked to this contrast, including external conditions, such as the impact of snowcover from the previous winter on albedo and slowly varying boundary forcing, such as soil moisture or sea surface temperature (e.g., Meehl 1994). The relationships highlighted here will be used to investigate the mechanisms of interannual variability of the NAMS in follow on studies.

Section 2 describes the datasets and the methodology. Section 3 reviews key features of the warm season precipitation regime. The interannual variability of warm season precipitation over North America is discussed in section 4. Hydrologic conditions and atmospheric circulation features associated with early, late, wet, and dry monsoons are discussed in sections 5 and 6. A summary and discussion are given in section 7.

2. Data analysis

a. Datasets

In order to study the interannual variability of warm season precipitation we employ a set of gridded daily precipitation analyses over the conterminous United States and Mexico. The analyses over the United States were developed from hourly observations for approximately 2500 stations obtained from the National Weather Service-Techniques Development Laboratory (Higgins et al. 1996). The analyses over Mexico were developed from long-term daily observations for 161 stations archived at the National Climatic Data Center, Asheville, North Carolina. The combined United States and Mexico dataset covers the period 1 January 1963 through 31 December 1988. The analyses were gridded to a horizontal resolution of 2° lat × 2.5° long (Fig. 1) using a Cressman (1959) scheme with modifications (Glahn et al. 1985; Charba et al. 1992). For convenience, we will refer to these analyses as the US–MEXICO precipitation dataset. We note that in this study the term “rainfall” is equivalent to measurable precipitation.

The primary dataset used to study atmospheric circulation features is the NCEP–NCAR Reanalysis (currently underway at NCEP in cooperation with NCAR; Kalnay et al. 1996). The reanalysis project has provided more than 50 years (1948–98+) of global gridded fields produced with a fixed state-of-the-art analysis system and large input database (including data available after the operational cutoff time). The NCEP–NCAR assimilation system consists of the NCEP Medium-Range Forecast (MRF) spectral model and the operational NCEP Spectral Statistical Interpolation (SSI; Parrish and Derber 1992) with the latest improvements (Kalnay et al. 1996). The assimilation is performed at a horizontal resolution of T62 and 28 sigma levels in the vertical with seven levels below 850 hPa. In this study we utilize the reanalysis winds, which are instantaneous fields available every 6 h.

In section 6 we explore relationships between wet (dry) monsoons and ENSO using sea surface temperature (SST) data obtained from the historical reconstruction of Smith et al. (1996) and gridded to a horizontal resolution of 2° lat × 2° long for the period 1950–95.

b. Identifying the onset date

The onset of the Mexican Monsoon (Douglas et al. 1993; Stensrud et al. 1995) is characterized by heavy rainfall over southern Mexico, which quickly spreads northward along the western slopes of the Sierra Madre Occidental and into Arizona and New Mexico by early July. Histograms of the mean (1963–94) daily rainfall (and the 5-day running mean) during summer at each grid point over Mexico and the southwestern United States (Fig. 2) show the timing of the northward progression of the monsoon. Further examination of Fig. 2 shows that the mean daily precipitation amounts also decrease rapidly toward the north. The southern and southwestern coasts of Mexico display a peak in rainfall during June followed by a relative minimum in July–August and a secondary peak in September (not shown, but see section 3 Fig. 9g). The midsummer relaxation in precipitation is not observed further north (also see section 3 Fig. 9f). The northern edge of the monsoon extends into Arizona and New Mexico (e.g., Douglas et al. 1993), but the rainfall is much lighter and more directly influenced by midlatitude effects. Based on differences in the characteristics of warm season precipitation and in the onset date of the monsoon (also see section 3 Figs. 5, 10, and 12) in Arizona–New Mexico, northwest Mexico, and southwest Mexico, we selected three regions (indicated by the “#,” “*,” and “+” symbols on Fig. 2) to study the interannual variability of the warm season precipitation regime; hereafter we refer to these regions as AZNM, NWMEX, and SWMEX, respectively.

In section 5 we will use the date of onset of the monsoon in each region to classify monsoons as early or late. The procedure for identifying the onset of the summertime rains in each region is as follows. In each region a precipitation index is obtained by averaging daily accumulations of observed precipitation at each grid point of the appropriate region (the shaded regions on Fig. 2). Care is used in choosing the grid points for each precipitation index; for example, Arizona exhibits a “pure” monsoon signal, that is, a sudden onset of monsoon rains, while eastern New Mexico has a more gradual increase due to mixed influences of the monsoon, the dryline, and the Great Plains low-level jet (see Fig. 4 of Higgins et al. 1997b). The mean daily (and 5-day running mean) area averaged precipitation for the three regions (Fig. 3) clearly show the northward progression of the monsoon.

The onset date in each region is determined using the resulting time series and a threshold crossing procedure. Because the rainfall amounts in each region are different (see Fig. 2), it is necessary to use different threshold criteria to define monsoon onset. The magnitude and duration criteria used are 0.5 mm day−1 and 3 days for AZNM, 1.0 mm day−1 and 5 days for NWMEX, and 2.0 mm day−1 and 5 days for SWMEX. In each region, the start of the monsoon occurs when the selection criteria are first satisfied after 1 May. Composite evolution fields for 1963–88 are obtained by averaging over all of the monsoons relative to the day when the precipitation in a given region first satisfies the threshold criteria; this day is designated as the onset day, or day 0. Note that by realigning the time series in this way we are not performing a simple average based on calendar day. The composite evolution of each precipitation index (Fig. 4) shows the onset of the monsoon rains. The average calendar date of onset (day 0 on each panel) is 7 June, 17 June, and 7 July in SWMEX, NWMEX, and AZNM, respectively. It is important to note that the compositing scheme makes the monsoon onset appear to be abrupt because it is keyed to synoptic as well as climate variability, as evidenced by the overshoot on each panel of Fig. 4. However, our choice of threshold criteria minimizes the overshoot immediately after onset. Statistics for the onset date of the monsoon and a classification of early and late monsoons in each region are discussed in section 5.

c. Significance of correlations

To assess the significance of correlations presented in sections 4–6, we followed the approach in Janowiak et al. (1998). The effective time between independent samples (Livezey 1995) was first computed according to
i1520-0442-12-3-653-eq1
where T is the effective number of years between independent samples, i is the lag number, N is the total number of lags, n is the sample length, and ρa and ρb are lagged autocorrelations for time series a and b. There are a total of 26-yearly values in the study period, and autocorrelations up to lag 13 (i.e., one-half of all samples in time) were computed (i.e., n = 26 and N = 13). Each time series was then “thinned” according to the value of T and the correlation coefficient was computed from only the independent data. The critical value of the correlation coefficient was evaluated statistically using a t test (with a null hypothesis of zero correlation);in each case statistical significance was assessed relative to the 95% confidence level.

3. Key features of the summer precipitation regime

Rainfall associated with the NAMS is clearly evident in Fig. 5a, which shows seasonal mean (1963–88) precipitation for July–September, usually the three rainiest summer months. The largest values of seasonal mean rainfall (Fig. 5a), in excess of 800 mm, occur along the southeast coast of Mexico in the vicinity of the Bay of Campeche. For this region, orography appears to play an important role in determining the seasonal mean rainfall. Heavy precipitation is also observed to the west of the Sierra Madre Occidental along the west coast of Mexico. Rainfall in excess of 200 mm is found in the United States from the Great Plains to the East Coast, with an area exceeding 400 mm over Florida. Examination of the mean monthly rainfall for June, July, August, and September (not shown) reveals that, in general, July and August are comparable but larger than that for June or September. Along the west coast of Mexico and in Arizona–New Mexico, the heaviest rainfall occurs during the month of August.

The extension of the monsoon rainfall into the southwestern United States is evident in Fig. 5b, which shows the ratio (expressed in percent) of rain falling during the 3-month period July–September to the annual mean precipitation. The highest values (exceeding 60%) are found along the west coast of Mexico; similar results were found by Douglas et al. (1993). The maximum values extend northward along the axis of the Sierra Madre Occidental and then northeastward across the Rio Grande valley in New Mexico and into the high plains of southeastern Colorado and western Kansas. As found by Douglas et al. (1993), southwest New Mexico appears to be the region most affected by the monsoon in the United States.

The contribution of the summer monsoon rainfall to the annual total does not reveal the month-to-month variations in rainfall. To put the summer monsoon precipitation in context within the annual cycle, the contributions to the annual total precipitation for each month of the year are displayed in Fig. 6. The bulk of the annual rainfall over much of Mexico occurs during the 4-month period (June–September). Other striking features include 1) the rapid increase in values along the west coast of Mexico and over southeastern Arizona and New Mexico from June to July; 2) the sharp gradient in values over northern Baja California during July and August; 3) the rapid decrease in contributions along the west coast of Mexico from September to October; and 4) the increased values in September over northeast Mexico, which are likely due to increases in the frequency of land falling tropical storms.

Throughout the warm season the low-level flow over the southern United States and Mexico is strongly influenced by the subtropical anticyclones (vectors on Fig. 7), with brisk southerlies over the southern Great Plains (reflecting the Great Plains low-level jet) and northwesterlies west of Baja California (reflecting the Baja jet);areas with no vectors indicate where the surface is above 925 hPa. The 200-hPa circulation center (streamlines in Fig. 7) is located over the western and southern United States during July and August and is likely related to enhanced atmospheric heating over the elevated terrain of the western United States. The resulting middle- and upper-tropospheric “monsoon high” is analogous to the Tibetan High over Asia (e.g., Tang and Reiter 1984) and the warm season Bolivian High over South America (e.g., Johnson 1976). During this 4-month period the large-scale 200-hPa flow (streamlines on Fig. 7) is characterized by the northward migration of the monsoon anticyclone along the west coast of Mexico to a position over northwestern Mexico by July. Increases in precipitation over the southwestern United States coincide with the arrival of the monsoon anticyclone in July (e.g., Okabe 1995; Higgins et al. 1997b). The precipitable water (not shown) indicates abundant moisture over the tropical eastern Pacific, Gulf of California, Baja California, western Mexico, and the eastern half of the United States (see Higgins et al. 1997b).

Climatological aspects of the onset of the warm season precipitation regime over Mexico and the United States can be viewed from maps of the mean rainfall and circulation difference between consecutive months (Fig. 8). The April–May period is characterized by a transition from the cold season circulation regime to the warm season one (Fig. 8a). This is accompanied by a decrease in upper-level westerlies over the continent, by an increase in precipitation over southern Mexico in response to the developing NAMS, and by an increase in precipitation over the central and southern Great Plains in response to increases in the amplitude of the diurnal cycle of precipitation (e.g., Wallace 1975; Higgins et al. 1996) and in the frequency of occurrence of the Great Plains low-level jet (e.g., Bonner 1968; Helfand and Schubert 1995; Higgins et al. 1997a). From May to June precipitation increases over most of Mexico; the largest increases in rainfall over the continent for any consecutive two-month period occur over southeastern Mexico during this period. Over the western United States there are notable increases in height (as reflected in the 200-hPa winds) and an increased southerly component in the low-level (925 hPa) flow off the west coast of Mexico, consistent with the increased precipitation there. The largest monthly variation in rainfall for the southwestern United States occurs between June and July (Fig. 8c) when increases exceeding 1 mm day−1 are found over much of southeastern Arizona and southwestern New Mexico. During this period the precipitation regime is characterized by an out-of-phase relationship between precipitation over southwestern North America and the U.S. Great Plains/Northern Tier and an in-phase relationship between precipitation over southwestern North America and in the Southeast. Changes in the upper-tropospheric wind and divergence fields (mean vertical motion) are broadly consistent with the evolution of this precipitation pattern (e.g., Higgins et al. 1997b). Previous studies have linked the onset of summer rains over northern Mexico and the southwestern United States to a decrease of rainfall over the Great Plains (e.g., Higgins et al. 1997b; Mock 1996; Tang and Reiter 1984; Douglas et al. 1993) and to an increase of rainfall along the East Coast (Tang and Reiter 1984). From July to August, there are no significant changes in the large-scale precipitation pattern over the conterminous United States, consistent with the fact that the NAMS is in its mature phase. There is a tendency for the monsoon anticyclone to begin its southward trek as indicated by the upper-level anticyclonic circulation over west central Mexico and the broad cyclonic circulation over the northwestern United States in the difference map (Fig. 8d).

Histograms of the mean monthly precipitation at various locations around the conterminous United States and Mexico reveal other aspects of regional relationships in precipitation (Fig. 9). Over Arizona (Fig. 9a) the maximum precipitation occurs in August during the peak of the monsoon. Over Texas (Fig. 9b) and Oklahoma (Fig. 9c) there are two peaks (May and September) with a relative minimum in rainfall during July and August. Similar behavior is found over Montana (Fig. 9d) though the September maximum is much weaker. This out-of-phase relationship is consistent with changes in the large-scale circulation (as discussed above). Rainfall over Sonora (northwest Mexico) (Fig. 9e) and Jalisco (west central Mexico) (Fig. 9f) exhibits a single peak during July–August in concert with the precipitation maximum in Arizona. Of the three Mexican states shown, Oaxaca (southern Mexico) (Fig. 9g) is closest to convection associated with the intertropical convergence zone (ITCZ); this region displays a typical double peak in summer precipitation (June and September) with a relative minimum during July and August. Additional features of the tropospheric mean climate during the warm season are discussed in Kousky and Ropelewski (1997).

4. Interannual variability of warm season precipitation

Maps of the standard deviation of seasonal mean rainfall and the standard deviation of the seasonal percent departure from normal based on 26 yr (1963–88) of data are shown in Figs. 10a,b, respectively. The standard deviation of seasonal mean rainfall (Fig. 10a) is large over the southern half of Mexico, where the seasonal normal precipitation is large. The standard deviation of the seasonal percent departure (Fig. 10b) is large in the southwestern United States where the amount of rainfall is small but where large changes can occur from year to year.

The seasonal (June–September) percent departure from normal rainfall for each year in AZNM, NWMEX, and SWMEX based on 26 yr of data is shown in Fig. 11. The correlations among these time series (after accounting for the effective time between independent yearly samples as described in section 2c) are 0.54, 0.02, and 0.26 for (AZNM, NWMEX), (AZNM, SWMEX) and (NWMEX, SWMEX), respectively; only the first coefficient is statistically significant at the 95% level. While AZNM and NWMEX are not statistically independent by this measure, differences in the mean seasonal precipitation (Fig. 5a), in the standard deviation of the mean seasonal precipitation (Fig. 10a), and in the onset date of the monsoon (Fig. 12) seem to justify our choice of regions. In section 6 we will show that monsoons with heavy or deficient rainfall over one of the regions are not necessarily (and often are not) accompanied by anomalies of the same sign (or magnitude) in one of the other regions, further supporting this point.

Research has shown that variations in seasonal precipitation in some parts of North America are linked to the ENSO phenomenon (e.g., Ropelewski and Halpert 1986, 1996). In this study we examine this concept using the precipitation database and Southern Oscillation Index (SOI) data, which is commonly used as an indicator of the state of ENSO. Five-month running mean SOI data were used to identify summer (June–September) seasons that experienced mature “cold” and “warm” episode conditions during the 1963–88 period. Summer seasons with mature warm episode conditions occurred in 1965, 1969, 1972, 1977, 1982, and 1987. Summer seasons with mature cold episode conditions occurred in 1964, 1971, 1973, 1975, and 1988. Warm and cold episodes are indicated by “W” and “C,” respectively, on Fig. 11c. Correlations between the SOI and the seasonal percent departure from normal in AZNM, NWMEX, and SWMEX for the 26-yr period are −0.15, 0.18, and 0.48, respectively; when the 1982–83 event is removed the correlations are roughly the same. The correlation between the SOI and the seasonal percent departure in SWMEX is statistically significant at the 95% level, implying that wet (dry) summer monsoons in SWMEX tend to be associated with La Niña (El Niño). This is consistent with evidence presented in section 6b that local SST influences on the land–sea thermal contrast are probably an important factor for monsoon strength in SWMEX.

The composite seasonal percent departure for the El Niño (La Niña) years is 0.6%, −10.3%, and −7.2% (−8.5%, −2.2%, and 6.5%) for AZNM, NWMEX, and SWMEX, respectively. Thus, in the mean, El Niño events are associated with deficient monsoons in NWMEX and SWMEX while La Niña events are associated with heavy monsoons in SWMEX and deficient monsoons in AZNM. Note that there is a reversal in the departures from SWMEX to AZNM, suggesting that local boundary forcing and other midlatitude factors may dominate SST effects in AZNM during ENSO (see section 6). Table 1 shows that during mature warm episode conditions, negative seasonal departures are observed in four of six cases (three of six cases) in NWMEX (SWMEX); one additional weak negative departure is observed in each region. During mature cold episode conditions, negative seasonal departures are observed in four of five cases in AZNM.

5. Characteristics of early and late monsoons

a. Interannual variability of the onset date

The daily precipitation indices for AZNM, NWMEX, and SWMEX (see section 2b) were used to determine the starting date of the summer monsoon for each year during the period 1963–88. Statistics associated with the onset date in each region are given in Table 2. The table shows that the mean and median dates for the start of the summer monsoon are very close in each region. The largest variability is found in NWMEX, where the time span between the earliest and latest start dates is 51 days. In AZNM the range between the earliest and latest start dates is closer to one month. The mean onset date for each grid point used in the precipitation indices is shown in Fig. 12. Correlations among the time series of the onset date (Fig. 13) are 0.18, 0.33, and 0.22 for (AZNM, NWMEX), (AZNM, SWMEX), and (NWMEX, SWMEX), respectively. These correlations are not statistically significant suggesting that the onset date of the monsoon in each region is more or less independent. In other words, early (late) onset in one region does not necessarily imply early (late) onset in the other regions. The low correlation between time series of the onset date in each region (Figs. 13a–c) is somewhat surprising since, in a mean sense, the monsoon progresses northward in a rapid, orderly fashion (Fig. 12).

The possible association between the dates of onset of the monsoon in a given region and the seasonal (June–September) precipitation anomalies in that region can be determined by correlating the time series of rainfall departure (Fig. 11) with the time series of onset date (Fig. 13). Despite the fact that the date of onset can fluctuate by more than a month, we find the interesting but counterintuitive result that the rainfall anomalies for the monsoon season in SWMEX are not related to the date of onset in SWMEX (the correlation between the time series is −0.03). This implies that the month-to-month variability of monsoon rainfall in this region is quite large. Recall that this part of Mexico normally experiences a relative minimum in monsoon precipitation during July and August. The anomalies for the monsoon season in NWMEX are also not significantly correlated with the onset date (the correlation between the time series is −0.32). However, the correlation between rainfall anomalies for the monsoon season in AZNM and the onset date is 0.52, which is statistically significant at the 95% level. Thus, early (late) monsoons in AZNM tend to have heavy (deficient) seasonal precipitation. Correlations between the June SOI and the onset date in each region are not statistically significant at the 95% level (−0.26, 0.27, and −0.13 for AZNM, NWMEX, and SWMEX, respectively). When this calculation is repeated using the seasonal June–September SOI, similar results are obtained.

b. Classification

If we use the time series in Fig. 13 to classify early (late) monsoons as those whose onset date is at least one standard deviation below (above) the mean onset date, then we find
i1520-0442-12-3-653-eq2
Very few of the years are in common between the regions, reinforcing the point that there is little relationship between the onset date of the monsoon in each region. This classification produces a reasonable separation between the average calendar date of early (late) monsoons in each region; 22 May (21 June), 28 May (2 July), and 23 June (17 July) for SWMEX, NWMEX, and AZNM, respectively. The time span between the average date of onset of early and late monsoons is largest in NWMEX (35 days), consistent with our previous findings on the variability of the onset date (see Table 2). The composite seasonal (JJAS) rainfall anomalies (percent departure from normal) for early (late) monsoons are 21.5%, 0.8%, and 0.2% (−3.0%, −10.1%, and 0.3%) in AZNM, NWMEX, and SWMEX, respectively. Thus, the strongest relationships are between early monsoons in AZNM and heavy rainfall and between late monsoons in NWMEX and deficient rainfall; comparison of the years of early monsoons in AZNM (listed above) to the years of wet monsoons in AZNM (see section 6a) shows five out of seven years in common. Reasons for the reverse relationship between early (late) monsoons and heavy (deficient) rainfall in these regions compared to SWMEX are likely related to differences in the importance of SST anomalies, soil-moisture anomalies, and internal dynamics in each region; these differences will be explored in our follow-on studies.

The composite evolution of monthly precipitation for early, late, and all (1963–88) monsoons in AZNM, NWMEX, and SWMEX is shown in Figs. 14a, b, and c, respectively. In both early and late years the onset of the monsoon rains is clearly evident in each region, just as it is in the composite based on all of the years. Early monsoons in AZNM are characterized by several months of above normal precipitation, starting in June and extending through the end of the year (Fig. 14a), consistent with large seasonal anomalies (see section 6b). Late monsoons in NWMEX have several months of below normal precipitation, though the temporal coherence is not as high as in AZNM. Late monsoons in AZNM, early monsoons in NWMEX, and both early and late monsoons in SWMEX have more intraseasonal fluctuations, which accounts for the weaker seasonal signal.

c. Monsoon monitoring and potential predictability

While the onset of the monsoon in one region is more or less independent of the onset in another region, a different picture emerges if the onset date in a particular region is used to accumulate rainfall in all of the regions. Figure 15a shows time series of precipitation anomalies [departures from the mean (1963–88) daily precipitation in each region for the 60-day period after monsoon onset in SWMEX]. Correlations between the three time series in Fig. 15a are all statistically significant at the 95% level [0.80, 0.47, and 0.70 for (AZNM, NWMEX), (AZNM, SWMEX), and (NWMEX, SWMEX), respectively]. In addition, the time series of the onset date in SWMEX (Fig. 13c) is significantly correlated with each time series in Fig. 15a (e.g., the correlation between SWMEX onset dates and SWMEX precipitation anomalies is 0.75) implying that early (late) monsoons in SWMEX favor less (more) rainfall over all three regions for two months after onset. A possible physical explanation is that early monsoons in SWMEX tend to occur when the land–sea thermal contrast is weaker than at normal onset time (due to weaker heating over land), which might account for the lighter precipitation, and vice versa. We note that the correlations are lower when precipitation is accumulated for longer periods after onset in SWMEX (e.g., 90–120 days) and that the seasonal (JJAS) rainfall is uncorrelated with the onset date as reported in section 5a.

When this calculation is repeated using onset dates in NWMEX (Fig. 15b), the correlations are somewhat lower [0.75, 0.26, and 0.53 for (AZNM, NWMEX), (AZNM SWMEX), and (NWMEX, SWMEX), respectively]; the first and third coefficients are statistically significant at the 95% level. When it is repeated using onset dates in AZNM (Fig. 15c) the correlations are much lower [0.31, −0.32, and 0.15 for (AZNM, NWMEX), (AZNM, SWMEX), and (NWMEX, SWMEX), respectively] and not statistically significant.

The correlation between rainfall in SWMEX and rainfall over Mexico and the conterminous United States for the 60-day period after monsoon onset in SWMEX (Fig. 16a) is significant at the 95% level along the west coast of Mexico and over southern Arizona. The most significant out of phase relationship is with precipitation over the northern and central Great Plains. These correlations do not imply, however, that the objective procedure is ineffective in choosing the onset date, because wet (dry) conditions for a 60-day period after onset in SWMEX do not guarantee a wet (dry) monsoon in the other regions, which usually have later onset dates. Thus, knowing the onset date of the monsoon in SWMEX may be useful for characterizing precipitation not only in SWMEX, NWMEX, and AZNM but also over other portions of the continental United States, at least for the 2-month period after onset. Significant correlations with monsoon rainfall in NWMEX and AZNM (Figs. 16b,c) are more regionally confined than they are with monsoon rainfall in SWMEX, suggesting that knowledge of the onset date in these regions is likely to be less useful for characterizing rainfall elsewhere.

In summary, these results suggest that the onset date of the monsoon in SWMEX is potentially useful for characterizing variations in warm season precipitation over the monsoon region after onset. Maps such as those in Fig. 16 could be constructed for any period after onset and then used to characterize rainfall for that period. There are a large number of factors that produce variability within a monsoon season, including synoptic-scale disturbances (lows, tropical storms), bursts and breaks (e.g., Carleton 1986), monsoon troughs, quasi-periodic oscillations, and midlatitude effects. Each one of these factors needs to be isolated and investigated within the context of the NAMS before we can fully appreciate how the onset of the monsoon is related to intraseasonal fluctuations in precipitation.

6. Characteristics of wet and dry monsoons

a. Classification

The results shown in Fig. 11 may be used to classify individual monsoons as wet or dry (relative to normal) in each region; the terms wet (dry) refer to the amount of precipitation during the monsoon season relative to normal and are synonymous with the terms heavy (deficient). If we choose years when seasonal anomalies (percent departure from normal) are greater than or equal to 0.5 (less than or equal to −0.5) standard deviations, then
i1520-0442-12-3-653-eq3
It can be seen that each region has a similar number of wet and dry monsoons. The composite seasonal (JJAS) percent departure from normal for wet (dry) monsoons in each region is 34.3%, 14.7%, and 10.6% (−24.5%, −16.5%, −14.6%) for AZNM, NWMEX, and SWMEX, respectively.

It is of interest to note that during wet (dry) monsoons, the individual months also show heavy (deficient) rain. Table 3 gives the rainfall departure from normal for the season and for the individual months of June, July, August, and September during wet and dry monsoons in each region. For most of the years, at least three of the four months have departures of the same sign. This indicates that, in spite of large month-to-month variability, particular seasons of heavy and deficient rain have significant temporal and spatial coherence. The temporal coherence is higher for the deficient monsoons in NWMEX and SWMEX, consistent with the fact that the deficient events show a stronger relationship with El Niño.

b. Precipitation and tropospheric circulation

The composite evolution of monthly precipitation for wet, dry, and all (1963–88) monsoons in AZNM, NWMEX, and SWMEX is shown in Figs. 17a, b, and c, respectively. In both wet and dry years the onset of the monsoon is clearly evident in each region, just as it is in the composites based on all of the years. Wet (dry) monsoons are characterized by several consecutive months of above (below) normal rainfall in each region (i.e., high temporal coherence in a composite sense) generally starting in June. Figure 17a also shows that wet (dry) summer monsoons in AZNM are preceded by dry (wet) conditions during the preceding winter. Higgins et al. (1998) showed that wet (dry) summer monsoons in AZNM tend to follow winters characterized by dry (wet) conditions in the southwestern United States and wet (dry) conditions in the Pacific Northwest. They attributed this association, in part, to the “memory” imparted to the atmosphere by the accompanying Pacific sea surface temperature anomalies (SSTA) in the preceding seasons.

The composite evolution of monthly precipitation in each region for El Niño and La Niña (Figs. 17d–f) years shows how the monthly departures contribute to the seasonal departures reported in section 4. The similarity between the SWMEX wet (dry) and La Niña (El Niño) composites during the warm season is evident, though the difference between wet and dry is almost twice the difference between cold and warm. During the late fall–early winter each region is relatively wet (dry) during El Niño (La Niña), though the signal decreases toward the south presumably because of the latitudinal dependence of the extended (retracted) Pacific jet on precipitation.

Maps of the composite seasonal (JJAS) precipitation anomalies [departures from the mean (1963–88) seasonal values] for wet and dry events are shown in Fig. 18; similar composites for El Niño and La Niña events are shown in Fig. 19. In each case, seasonal departures from normal exceeding 10% of the mean are shaded. The maps in Figs. 18 and 19 were compared using pattern correlation. It was found that the best agreement was between SWMEX wet and La Niña (r = 0.67) and between SWMEX dry and El Niño (r = 0.71). The SWMEX dry, NWMEX dry, and AZNM wet composites all show some evidence of a Great Basin signal (seasonal percent departures in the 10%–25% range), which Ropelewski and Halpert (1986) identified during the warm season of El Niño events. We note that when the 1982–83 event is removed there is little change in the qualitative nature of the composites and quantitative changes are minimal.

During wet and dry monsoons the large-scale upper-level flow shows dramatic seasonal JJAS departures from normal (Fig. 20) consistent with departures in the continental-scale precipitation pattern (Fig. 18). Wet (dry) monsoons in AZNM (Figs. 20a,b) feature an enhanced (suppressed) monsoon anticyclone consistent with the precipitation anomalies over the southwestern United States. As shown by Higgins et al. (1998), wet (dry) monsoons in this region are associated with a suppressed (enhanced) local Hadley circulation during the spring and summer consistent with the patterns of tropical precipitation anomalies in the vicinity of the ITCZ. Higgins et al. (1998) also showed that anomalies in ITCZ precipitation and in the local Hadley circulation are most pronounced during the spring preceding the monsoon and that these changes are accompanied by consistent and coherent changes in the SST and the subsurface thermal structure in the vicinity of the eastern Pacific cold tongue.

Composites of the seasonal (JJAS) 200-hPa wind and streamfunction anomalies [departures from the mean (1968–88) seasonal values] for wet monsoons in SWMEX (Fig. 20e) and for La Niña (Fig. 21b) both feature a cyclonic couplet straddling the equator (e.g., Arkin 1982) and 200-hPa westerly (925-hPa easterly) anomalies along the equator, typical of mature cold episode conditions (Fig. 21a). Of course the anomalies are stronger in the La Niña composite, but all of the basic features are present in each case. Alternately, dry monsoons in SWMEX feature an anticyclonic couplet straddling the equator and 200-hPa easterly (925-hPa westerly) anomalies along the equator (Fig. 20f), typical of mature warm episode conditions. The SWMEX wet (dry) patterns also have easterly (westerly) departures of the winds in the Northern Hemisphere subtropics consistent with a retracted (extended) North Pacific jet. We note that the NWMEX dry pattern (Fig. 20d) is quite similar to the SWMEX dry pattern, consistent with the possible association between El Niño and deficient monsoons in NWMEX.

The association between ENSO and anomalous monsoons in SWMEX is attributed, in part, to the impact of local sea surface temperature anomalies on land–sea thermal contrast, hence the strength of the monsoon. Composites of seasonal (June–September) SSTA (departures from mean 1963–88 seasonal values) for wet and dry monsoons in each region show a high degree of antisymmetry for each pair of composites. For this reason, Figs. 22a–c show composites of the seasonal SSTA represented as the difference between wet and dry monsoons in AZNM, NWMEX, and SWMEX, respectively. Comparison of these figures to the composite seasonal SSTA represented as the difference between La Niña and El Niño (Fig. 22d) shows a good degree of similarity between SWMEX and ENSO. Both composites imply that, for a given boundary forcing over land, negative SSTA would tend to enhance the land–sea thermal contrast, hence the strength of the monsoon and vice versa. We note that about half of the years with mature El Niño (La Niña) events (see the classification in section 4) are also years with dry (wet) monsoons in SWMEX. In AZNM and NWMEX the comparison between the monsoon composites and ENSO is not as good and, in particular, along the west coast of North America the SSTA are in the opposite sense. In these regions other factors (such as local feedbacks due to slowly varying land surface forcing or internal dynamics) might dominate. Again we note that the SSTA pattern is preserved when we remove the 1982–83 event from the composites, though there are slight changes in amplitude.

7. Summary and discussion

The North American warm season is characterized by a monsoon system that provides a useful framework for describing and diagnosing the interannual variability of the warm season precipitation regime. Daily observed precipitation over the United States and Mexico and various fields from the NCEP–NCAR Reanalysis were used to compare and contrast hydrologic conditions and atmospheric circulation features associated with early, late, wet, and dry monsoons in three regions (AZNM, NWMEX, and SWMEX) that exhibit many different characteristics of monsoon precipitation.

While the seasonal normal rainfall and its variability are largest in SWMEX, the mean seasonal percent departure from normal is largest in AZNM. There is considerable intraseasonal variability during the monsoon as the standard deviation of monthly mean departures is roughly twice that of the seasonal departures.

Wet (dry) monsoons in each region are characterized by several consecutive months of above (below) normal rainfall. Wet (dry) summer monsoons in AZNM tend to follow winters characterized by dry (wet) conditions in the southwestern United States and wet (dry) conditions in the Pacific Northwest. The strongest association between the monsoon and ENSO was found in southwest Mexico, where wet (dry) monsoons often occur during La Niña (El Niño). This association was attributed, in part, to the impact of local sea surface temperature anomalies on land–sea thermal contrast, hence the strength of the monsoon. Roughly half of the wet and dry monsoons in SWMEX occurred during ENSO events, implying that other factors are also important for determining monsoon strength in this region.

While the mean seasonal evolution of the North American monsoon system (NAMS) is characterized by the regular northward progression of heavy precipitation from southern Mexico by early June to the southwestern United States by early July, the precise date of onset in a given region is highly variable and likely to be unrelated to the date of onset in the other regions. The date of onset is correlated with seasonal rainfall anomalies such that early monsoons in AZNM tend to be heavy and late monsoons in NWMEX tend to be deficient. It is notable that the classifications of early and wet monsoons in AZNM have five events in common. No relationship exists between the SWMEX onset date and seasonal rainfall anomalies, which is somewhat counterintuitive given the association with ENSO. However, interannual fluctuations in the rainfall that occurs over the entire monsoon region during the two month period after onset in SWMEX are highly correlated, suggesting that knowledge of the starting date in SWMEX may be useful for analyzing, understanding, and possibly predicting these fluctuations.

Clearly, it will be useful to improve our ability to determine monsoon onset in advance. This requires that we determine the dominant factors influencing the interannual variability of monsoon onset. A convenient framework for understanding and describing potential mechanisms of interannual variability of the monsoon is to isolate factors associated with the atmospheres internal dynamics and with its lower boundary conditions. The combined effects of dynamical instabilities (e.g., synoptic-scale disturbances), nonlinear interactions among various scales of motion, thermal and orographic forcing, and tropical–extratropical interactions are all examples of internal dynamical processes that may produce interannual variability in the NAMS. There is a growing body of modeling and observational evidence that the slowly varying boundary conditions of SST, soil moisture, sea ice, and snow at the earth’s surface can influence the interannual variability of the atmospheric circulation. Over land, internal feedbacks involving soil moisture can provide a significant influence on monsoon rainfall (e.g., Meehl and Washington 1988;Meehl 1994) but these feedbacks are part of a much more complicated loop involving many other factors. Although these processes operate together, it is also necessary to isolate them in order to determine their relative contributions to the development, maintenance and decay of warm season hydrologic anomaly regimes. It is hoped that our analyses provide a useful foundation for follow-on studies of the mechanisms of interannual variability of the warm season precipitation regime over North America. Our analyses also highlight the critical need for adequate in situ precipitation data and upper-air analyses at sufficiently high temporal resolution and in a form convenient for analysis. In general, finer resolution models and long-term observations will be required to understand regional-scale impacts of the monsoon. Ideally, this could be accomplished via a multiyear regional (North American) reanalysis similar in philosophy to the multiyear NCEP–NCAR global reanalysis.

Acknowledgments

We wish to thank the NCEP–NCAR Reanalysis Team for the Reanalysis data and David Easterling of the National Climatic Data Center, Asheville, for the Mexican precipitation data. The authors are also indebted to John Janowiak, Gerry Bell, Vern Kousky, Gene Rasmusson, Chet Ropelewski, Ants Leetmaa, and Kingtse Mo for insightful discussions. This work was partially supported by the NOAA Office of Global Programs under the Pan American Climate Studies (PACS) project and the GEWEX Continental-Scale International Project (GCIP) and by Interagency Agreement No. S-41367-F under the authority of NASA/GSFC.

REFERENCES

  • Arkin, P. A., 1982: The relationship between interannual variability in the 200-mb tropical wind field and the southern oscillation. Mon. Wea. Rev.,110, 1393–1404.

  • Bonner, W. D., 1968: Climatology of the low-level jet. Mon. Wea. Rev.,96, 833–850.

  • Carleton, A. M., 1986: Synoptic-dynamic character of “bursts” and “breaks” in the southwest U.S. summer precipitation singularity. J. Climatol.,6, 605–623.

  • Charba, J. P., A. W. Harrell III, and A. C. Lackner III, 1992: A monthly precipitation amount climatology derived from published atlas maps: Development of a digital database. TDL Office Note 92-7, 20 pp. [Available from TDL Office, NOAA, U.S. Department of Commerce, 1325 East–West Hwy., Silver Spring, MD 20910.].

  • Cressman, G. P., 1959: An operational objective analysis system. Mon. Wea. Rev.,87, 367–374.

  • Dhar, O. N., P. R. Rakhecha, and B. N. Mandal, 1980: Does the early or late onset of the monsoon provide any clue to the subsequent rainfall during the monsoon season? Mon. Wea. Rev.,108, 1069–1072.

  • Douglas, M. W., R. A. Maddox, K. Howard, and S. Reyes, 1993: The Mexican Monsoon. J. Climate,6, 1665–1677.

  • Glahn, H. R., T. L. Chambers, W. S. Richardson, and H. P. Perrotti, 1985: Objective map analysis for the local AFOS MOS Program. NOAA Tech. Memo. NWS TDL 75, 34 pp. [Available from NOAA, U.S. Department of Commerce. 1325 East–West Hwy., Silver Spring, MD 20910.].

  • Helfand, H. M., and S. D. Schubert, 1995: Climatology of the Great Plains low-level jet and its contribution to the continental moisture budget of the United States. J. Climate,8, 784–806.

  • Higgins, R. W., J. E. Janowiak, and Y. Yao, 1996: A Gridded Hourly Precipitation Data Base for the United States (1963–1993). NCEP/Climate Prediction Center Atlas 1, 47 pp.

  • ——, Y. Yao, E. S. Yarosh, J. E. Janowiak, and K. C. Mo, 1997a: Influence of the Great Plains low-level jet on summertime precipitation and moisture transport over the central United States. J. Climate,10, 481–507.

  • ——, ——, and X. Wang, 1997b: Influence of the North American Monsoon System on the United States summer precipitation regime. J. Climate,10, 2600–2622.

  • ——, K. C. Mo, and Y. Yao, 1998: Interannual variability of the United States summer precipitation regime with emphasis on the southwestern monsoon. J. Climate,11, 2582–2606.

  • Janowiak, J. E., A. Gruber, C. R. Kondragunta, R. E. Livezey, and G. J. Huffman, 1998: A comparison of the NCEP–NCAR Reanalysis Precipitation and the GPCP Raingauge–Satellite combined dataset with observational error considerations. J. Climate,11, 2960–2979.

  • Johnson, A. M., 1976: The climate of Peru, Bolivia and Ecuador. World Survey of Climatology. Vol. 12, Climates of Central and South America, W. Schwerdtfeger and H. E. Landsberg, Eds., Elsevier, 147–218.

  • Kalnay, E., and Coauthors, 1996: The NCEP–NCAR Reanalysis Project. Bull. Amer. Meteor. Soc.,77, 437–471.

  • Kiladis, G. N., and H. Van Loon, 1988: The Southern Oscillation. Part VII: Meteorological anomalies over the Indian and Pacific sectors associated with extremes of the oscillation. Mon. Wea. Rev.,116, 120–136.

  • Kousky, V. E., and C. F. Ropelewski, 1997: The Tropospheric Seasonally Varying Mean Climate over the Western Hemisphere (1979–1995). NCEP/Climate Prediction Center Atlas 3, 135 pp.

  • Livezey, R. E., 1995: Field intercomparison. An Analysis of Climate Variability Applications of Statistical Techniques, H. Von Storch and A. Navara, Eds., Springer-Verlag, 159–176.

  • Meehl, G. A., 1994: Influence of the land surface in the Asian summer monsoon: External conditions versus internal feedbacks. J. Climate,7, 1033–1049.

  • ——, and W. A. Washington, 1988: A comparison of soil moisture sensitivity in two global climate models. J. Atmos. Sci.,45, 1476–1492.

  • Mo, K. C., J. N. Paegle, and R. W. Higgins, 1997: Atmospheric processes associated with summer floods and droughts in the central United States. J. Climate,10, 3028–3046.

  • Mock, C. J., 1996: Climatic controls and spatial variations of precipitation in the western United States. J. Climate,9, 1111–1125.

  • Okabe, I. T., 1995: The North American Monsoon. Ph.D. dissertion, University of British Columbia, 146 pp.

  • Parrish, D. F., and J. C. Derber, 1992: The National Meteorological Center’s spectral statistical interpolation analysis system. Mon. Wea. Rev.,120, 1747–1763.

  • Ramage, C. S., 1971: Monsoon Meteorology. Academic Press, 391 pp.

  • Ropelewski, C. F., and M. S. Halpert, 1986: North American precipitation and temperature patterns associated with El Niño–Southern Oscillation (ENSO). Mon. Wea. Rev.,114, 2352–2362.

  • ——, and ——, 1996: Quantifying southern oscillation–precipitation relationships. J. Climate,9, 1043–1059.

  • Smith, T. M., R. W. Reynolds, R. E. Livezey, and D. C. Stokes, 1996:Reconstruction of historical sea surface temperatures using empirical orthogonal functions. J. Climate,9, 1403–1420.

  • Stensrud, D. J., R. L. Gall, S. L. Mullen, and K. W. Howard, 1995:Model climatology of the Mexican monsoon. J. Climate,8, 1775–1794.

  • Tang, M., and E. R. Reiter, 1984: Plateau monsoons of the Northern Hemisphere: A comparison between North America and Tibet. Mon. Wea. Rev.,112, 617–637.

  • Wallace, J. M., 1975: Diurnal variations in precipitation and thunderstorm frequency over the conterminous United States. Mon. Wea. Rev.,103, 406–419.

  • Webster, P. J., 1987: The elementary monsoon. Monsoons, J. S. Fein and P. L. Stephens, Eds., John Wiley and Sons, 3–32.

  • ——, and S. Yang, 1992: Monsoon and ENSO: Selectively interactive systems. Quart. J. Roy. Meteor. Soc.,118, 877–926.

Fig. 1.
Fig. 1.

Typical station distribution for hourly reporting stations in the United States and daily reporting stations in Mexico used in the US–MEXICO merged precipitation dataset. Gridlines represent the (33 × 26) grid to which the station data have been analyzed. The topography of the region is also included, courtesy of the United States National Geophysical Data Center. The resolution of the data is 5 min (0.083°). The topography data are available online at http://www:ngdc.noaa.gov/mgg/mggd.html. Shading intervals are at 500, 1000, 2000, and 3000 m.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 2.
Fig. 2.

Histograms of the mean (1963–88) daily and 5-day running mean precipitation (units: mm day−1) during May–August at grid points in the southwestern United States and Mexico from the merged US–MEXICO precipitation database. Grid points used for the AZNM, NWMEX, and SWMEX precipitation indices (defined in section 2b) are indicated by the “#,” “*,” and “+” symbols, respectively. The grid points at which the data are valid are located at the center of each box. The topography of the region is also included, as described in Fig. 1.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 3.
Fig. 3.

Mean (1963–88) daily and 5-day running mean area averaged precipitation (units: mm day−1) for the (a) AZNM, (b) NWMEX, and (c) SWMEX regions.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 4.
Fig. 4.

Evolution of the composite mean (1963–88) daily precipitation (mm day−1) for the (a) AZNM, (b) NWMEX, and (c) SWMEX regions relative to monsoon onset.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 5.
Fig. 5.

(a) Mean (1963–88) seasonal precipitation (units: mm) for July–September from the US–MEXICO merged analysis. The contours are 50, 100, 200, 400, 600, 800, and 1000 mm and values greater than 100 mm are shaded. (b) Contribution of the precipitation during July–September to the annual total, expressed in percent, from the US–MEXICO merged precipitation analysis. The contour interval is 5% and values greater than 40% are shaded.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 6.
Fig. 6.

Analysis of the contribution of the mean (1963–88) monthly precipitation to the annual mean (units: percent) over the conterminous United States and Mexico for each month from the US–MEXICO merged analysis. The contour interval is 5% and areas with values exceeding 15% are shaded.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 7.
Fig. 7.

Mean (1968–88) monthly 925-hPa vector wind (m s−1), 200-hPa streamlines, and US–MEXICO precipitation (shading) for (a) May, (b) June, (c) July, and (d) August. Circulation data are from the NCEP–NCAR Reanalysis. A topography mask has been applied to the 925-hPa winds. Precipitation amounts are in mm day−1 and values greater than 1 mm day−1 are shaded. The characteristic vector length is 10 m s−1.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 8.
Fig. 8.

Mean (1968–88) monthly 925-hPa vector wind (m s−1), 200-hPa streamlines, and US–MEXICO precipitation (shading) represented as a difference between consecutive months for (a) May–April, (b) June–May, (c) July–June, and (d) August–July. A topography mask has been applied to the 925-hPa winds. The characteristic vector length is 2 m s−1 and precipitation differences >0.5 mm day−1 (<−0.5 mm day−1) are shaded dark (light).

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 9.
Fig. 9.

Histograms of mean (1963–94) monthly precipitation (mm day −1) from selected grid points over the conterminous United States and Mexico: (a) (34°N, 110°W) in Arizona, (b) (30°N, 97.5°W) in Texas, (c) (36°N, 95°W) in Oklahoma, (d) (46°N, 110°W) in Montana, (e) (28°N, 110°W) in Sonora, (f) (20°N, 105°W) in Jalisco, and (g) (16°N, 97.5°W) in Oaxaca.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 10.
Fig. 10.

Standard deviation of (a) the seasonal (June–September) mean precipitation (mm) and (b) the seasonal (June–September) percent departure from normal (units: percent) based on 26 yr (1963–88) of data. In (a) the contours are 50, 100, and 200 mm and values greater than 50 mm are shaded. In (b) the contour interval is 10% and values greater than 20% are shaded. In each case, areas where the seasonal precipitation is less than 30 mm are masked.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 11.
Fig. 11.

Seasonal (June–September) percent departure from normal precipitation (%) for (a) AZNM, (b) NWMEX, and (c) SWMEX based on 26 yr (1963–88) of data. Anomalies are computed with respect to the apropriate area mean for the 26-yr period. On panel (c) the years with mature cold or warm episode conditions are indicated by a C or a W, respectively.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 12.
Fig. 12.

Mean (1963–88) calendar date of onset for the summer monsoon at each grid point used in the AZNM, NWMEX, and SWMEX precipitation indices.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 13.
Fig. 13.

Time series of the date of onset for the summer monsoon in (a) AZNM, (b) NWMEX, and (c) SWMEX for the period 1963–88. The mean onset date for each region is indicated by a solid line and one standard deviation is indicated by dashed lines. All dates are Julian dates in the calendar year (e.g., day 182 = 1 July, etc.).

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 14.
Fig. 14.

Composite evolution of monthly precipitation (mm day−1) over (a) AZNM, (b) NWMEX, and (c) SWMEX for early (dashed line), late (dash-dot line), and all (1963–88) monsoons (thick solid line).

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 15.
Fig. 15.

Time series of accumulated precipitation anomalies [departures from mean (1963–88) daily values] in SWMEX (dash-dot line), NWMEX (dashed line), and AZNM (solid line) for the 60-day period after monsoon onset in (a) SWMEX, (b) NWMEX, and (c) AZNM.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 16.
Fig. 16.

Correlation between rainfall for the 60-day period after monsoon onset in (a) SWMEX, (b) NWMEX, and (c) AZNM and rainfall over Mexico and the United States. The contour interval is 10%, the zero contour is omitted for clarity. Regions where the correlation is locally significant at the 95% confidence level are shaded.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 17.
Fig. 17.

Composite evolution of monthly precipitation (mm day−1) over (a) AZNM, (b) NWMEX, and (c) SWMEX for wet (dashed line), dry (dash-dot line), and all (1963–88) monsoons (thick solid line). Composite evolution of monthly precipitation (mm day−1) over (d) AZNM, (e) NWMEX, and (f) SWMEX for La Niña (dashed line) and El Niño (dash-dot) years, and all (1963–88) years (thick solid line).

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 18.
Fig. 18.

Maps of the composite seasonal (JJAS) precipitation anomalies (mm day−1) for (a) wet monsoons in AZNM, (b) dry monsoons in AZNM, (c) wet monsoons in NWMEX, (d) dry monsoons in NWMEX, (e) wet monsoons in SWMEX, and (f) dry monsoons in SWMEX. In each case the anomalies are departures from the mean (1963–88) seasonal values. The contour interval is 0.25 mm day−1. Dark (light) shading indicates areas where seasonal precipitation anomalies are greater than 10% of the mean and positive (negative). Areas where the seasonal precipitation is less than 30 mm are masked.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 19.
Fig. 19.

Same as Fig. 18, except for (a) El Niño and (b) La Niña.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 20.
Fig. 20.

Composite seasonal (JJAS) 200-hPa wind anomalies, 200-hPa streamfunction anomalies, and US–MEXICO precipitation anomalies for (a) wet monsoons in AZNM, (b) dry monsoons in AZNM, (c) wet monsoons in NWMEX, (d) dry monsoons in NWMEX, (e) wet monsoons in SWMEX, and (f) dry monsoons in SWMEX. All fields are represented as departures from mean (1968–88) seasonal values. The contour interval is 1×106m2 s−1 and the standard vetor length is 10 m s−1. Precipitation anomalies >0.25 mm day−1 (<−0.25 mm day−1) are shaded dark (light).

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 21.
Fig. 21.

Same as Fig. 20, except for (a) El Niño and (b) La Niña.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Fig. 22.
Fig. 22.

Composite seasonal (JJAS) sea surface temperature anomalies [departures from mean (1963–88) seasonal values] represented as the difference between wet and dry monsoons in (a) AZNM, (b) NWMEX, and (c) SWMEX, and (d) as the difference between El Niño and (b) La Niña. In each panel the dark (light) shading indicates positive (negative) anomalies significant at the 90% confidence level. The contour interval is 0.2°C.

Citation: Journal of Climate 12, 3; 10.1175/1520-0442(1999)012<0653:IVOTNA>2.0.CO;2

Table 1.

Seasonal (June–September) percent departure from normal seasonal precipitation (units: percent) in AZNM, NWMEX, and SWMEX for mature-warm and -cold episode conditions. Warm and cold episode conditions were identified using the 5-month running mean Southern Oscillation index (SOI).

Table 1.
Table 2.

Statistics associated with the onset of the summer monsoon in Arizona–New Mexico, northwest Mexico, and southwest Mexico for 1963–88. Both Julian dates in the calendar year and calendar dates are shown.

Table 2.
Table 3.

Seasonal (JJAS) and monthly percent departure from normal precipitation (%) for wet and dry monsoon years in AZNM, NWMEX, and SWMEX. In each case the anomalies are departures from the appropriate mean (1963–88) seasonal or mean monthly values.

Table 3.
Save