• Carvalho, L. M. V., , C. Jones, , and B. Liebmann, 2004: The South Atlantic convergence zone: Intensity, form, persistence, relationships with intraseasonal to interannual activity and extreme rainfall. J. Climate, 17 , 88108.

    • Search Google Scholar
    • Export Citation
  • Cavalcanti, I. F. A., , and M. K. Kayano, 1999: High-frequency patterns of the atmospheric circulation over Southern Hemisphere and South America. Meteor. Atmos. Phys., 69 , 179193.

    • Search Google Scholar
    • Export Citation
  • Cavalcanti, I. F. A., , and V. E. Kousky, cited. 2001: Drought in Brazil during summer and fall 2001 and associated atmospheric circulation features. [Available online at http://www.cptec.inpe.br/products/climanalise/artigos/criseing.PDF.].

  • Cavalcanti, I. F. A., , and V. Kousky, 2003: Climatology of the South American cold fronts. Preprints, Seventh Int. Conf. on Southern Hemisphere Meteorology and Oceanography, Wellington, New Zealand, Amer. Meteor. Soc., 6.5.

  • Cavalcanti, I. F. A., , C. K. Folland, , and A. W. Colman, 2002: Note on “Predictability of Northeast Brazil rainfall and real-time forecast skill, 1987–98”. J. Climate, 15 , 19931996.

    • Search Google Scholar
    • Export Citation
  • Cunningham, C. A. C., , and I. F. A. Cavalcanti, 2006: Intraseasonal modes of variability affecting the South Atlantic Convergence Zone. Int. J. Climatol., 26 , 11651180.

    • Search Google Scholar
    • Export Citation
  • Doyle, M. E., , and V. R. Barros, 2002: Midsummer low-level circulation and precipitation in subtropical South America and related sea surface temperature anomalies in the South Atlantic. J. Climate, 15 , 33943410.

    • Search Google Scholar
    • Export Citation
  • Duchon, C. E., 1979: Lanczos filtering in one and two dimensions. J. Appl. Meteor., 18 , 10161022.

  • Fasullo, J., , and P. J. Webster, 2002: Hydrologic signatures relating the Asian summer monsoon and ENSO. J. Climate, 15 , 30823095.

  • Fasullo, J., , and P. J. Webster, 2003: A hydrological definition of Indian monsoon onset and withdrawal. J. Climate, 16 , 32003211.

  • Figueroa, S. N., , P. Satyamurty, , and P. L. S. Dias, 1995: Simulations of the summer circulation over the South American region with an Eta coordinate model. J. Atmos. Sci., 52 , 15731584.

    • Search Google Scholar
    • Export Citation
  • Fu, R., , B. Zhu, , and R. E. Dickinson, 1999: How do atmosphere and land surface influence seasonal changes of convection in the tropical Amazon. J. Climate, 12 , 13061321.

    • Search Google Scholar
    • Export Citation
  • Gan, M. A., , V. E. Kousky, , and C. F. Ropelewski, 2004: The South America monsoon circulation and its relationship to rainfall over west-central Brazil. J. Climate, 17 , 4766.

    • Search Google Scholar
    • Export Citation
  • Gan, M. A., , V. B. Rao, , and C. L. Moscati, 2006: South American monsoon indices. Atmos. Sci. Lett., 6 , 219223.

  • Gandu, A. W., , and J. E. Geisler, 1991: A primitive equations model study of the effect of topography on the summer circulation over tropical South America. J. Atmos. Sci., 48 , 18221836.

    • Search Google Scholar
    • Export Citation
  • Garreaud, R., 2000: Intraseasonal variability of moisture and rainfall over the South American altiplano. Mon. Wea. Rev., 128 , 33373346.

    • Search Google Scholar
    • Export Citation
  • Garreaud, R., , and J. M. Wallace, 1998: Summertime incursions of midlatitude air into subtropical and tropical South America. Mon. Wea. Rev., 126 , 27132733.

    • Search Google Scholar
    • Export Citation
  • Grimm, A. M., , C. Vera, , and R. Mechoso, 2004: The South American Monsoon System. The Global Monsoon System: Research and Forecast, WMO Tech. Doc. 1266, TMRP Rep. 70, 219–238.

  • Hastenrath, S., , and L. Heller, 1977: Dynamics of climatic hazards in Northeast Brazil. Quart. J. Roy. Meteor. Soc., 103 , 7792.

  • Holland, G., 1986: Interannual variability of the Australian monsoon at Darwin. Mon. Wea. Rev., 114 , 594604.

  • Jones, C., , and L. M. V. Carvalho, 2002: Active and break phases in the South American monsoon system. J. Climate, 15 , 905914.

  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Kistler, R., and Coauthors, 2001: The NCEP–NCAR 50-Year Reanalysis: Monthly means CD-ROM and documentation. Bull. Amer. Meteor. Soc., 82 , 247267.

    • Search Google Scholar
    • Export Citation
  • Kousky, V., 1988: Pentad outgoing longwave radiation climatology for the South American sector. Rev. Bras. Meteor., 3 , 217231.

  • Kousky, V., , and C. F. Ropelewski, 1997: The Tropospheric Seasonally Varying Mean Climate over the Western Hemisphere (1979–1995). Atlas No. 3, NCEP/Climate Prediction Center, 135 pp.

    • Search Google Scholar
    • Export Citation
  • Lenters, J., , and K. Cook, 1995: Simulation and diagnosis of the regional summertime precipitation climatology of South America. J. Climate, 8 , 29883005.

    • Search Google Scholar
    • Export Citation
  • Li, W., , and R. Fu, 2004: Transition of the large-scale atmospheric and land surface conditions from the dry to the wet season over Amazonia as diagnosed by the ECMWF reanalysis. J. Climate, 17 , 26372651.

    • Search Google Scholar
    • Export Citation
  • Li, W., , and R. Fu, 2006: Influence of cold air intrusions on the wet season onset over Amazonia. J. Climate, 19 , 257275.

  • Liebmann, B., , and J. Marengo, 2001: The seasonality and interannual variability of rainfall in the Brazilian Amazon basin. J. Climate, 14 , 43084318.

    • Search Google Scholar
    • Export Citation
  • Liebmann, B., , G. N. Kiladis, , J. A. Marengo, , T. Ambrizzi, , and J. D. Glick, 1999: Submonthly convective variability over South America and the South Atlantic convergence zone. J. Climate, 12 , 18771891.

    • Search Google Scholar
    • Export Citation
  • Marengo, J., , B. Liebmann, , V. Kousky, , N. Filizola, , and I. Wainer, 2001: On the onset and end of the rainy season in the Brazilian Amazon basin. J. Climate, 14 , 833852.

    • Search Google Scholar
    • Export Citation
  • Marengo, J., , W. R. Soares, , C. Saulo, , and M. Nicolini, 2004: Climatology of the low-level jet east of the Andes as derived from the NCEP–NCAR reanalyses: Characteristics and temporal variability. J. Climate, 17 , 22612280.

    • Search Google Scholar
    • Export Citation
  • Mo, K. C., , and J. N. Paegle, 2001: The Pacific–South American modes and their downstream effects. Int. J. Climatol., 21 , 12111229.

  • Nogués-Paegle, J., , and K. Mo, 1997: Alternating wet and dry conditions over South America during summer. Mon. Wea. Rev., 125 , 279291.

    • Search Google Scholar
    • Export Citation
  • Nogués-Paegle, J., , L. A. Byerle, , and K. C. Mo, 2000: Intraseasonal modulation of South American summer precipitation. Mon. Wea. Rev., 128 , 837850.

    • Search Google Scholar
    • Export Citation
  • Nogués-Paegle, J., and Coauthors, 2002: Progress in Pan American CLIVAR research: Understanding the South American monsoon. Meteorologica, 1–2 , 332.

    • Search Google Scholar
    • Export Citation
  • Pearce, R. P., , and U. C. Monhanty, 1984: Onsets of the Asia summer monsoon 1979–82. J. Atmos. Sci., 41 , 16201639.

  • Rao, V. B., , I. F. A. Cavalcanti, , and K. Hada, 1996: Annual variation of rainfall over Brazil and water vapor characteristics over South America. J. Geophys. Res., 101 , 2653926551.

    • Search Google Scholar
    • Export Citation
  • Silva, A. E., , and L. M. V. Carvalho, 2007: Large-scale index for South American Monsoon (LISAM). Atmos. Sci. Lett., 8 , 5157.

  • Siqueira, J. R., , and L. A. T. Machado, 2004: Influence of the frontal systems on the day-to-day convection variability over South America. J. Climate, 17 , 17541766.

    • Search Google Scholar
    • Export Citation
  • Soman, M. K., , and K. K. Kumar, 1993: Space–time evolution of meteorological features associated with the onset of Indian summer monsoon. Mon. Wea. Rev., 121 , 11771194.

    • Search Google Scholar
    • Export Citation
  • Vera, C., and Coauthors, 2006: Toward a unified view of the American monsoon systems. J. Climate, 19 , 49775000.

  • Zhou, J., , and K. M. Lau, 1998: Does a monsoon climate exist over South America? J. Climate, 11 , 10201040.

  • View in gallery

    Twelve areas for the mean annual cycle analysis. The shaded areas correspond to the 8 areas that have a monsoon regime.

  • View in gallery

    Mean annual cycle (1984–2004) of precipitation (mm day−1) (▪), specific humidity at 925 hPa (g kg−1) (○), and potential evapotranspiration rate (W m−2) (dot–dash line) in 12 selected areas of Fig. 1.

  • View in gallery

    Mean annual cycle (1984–2004) of zonal wind at 850 hPa (m s−1) (○) and surface temperature (°C) (▪) in 12 selected areas of Fig. 1.

  • View in gallery

    Monthly mean (1984–2004) integrated humidity transport from 1000 to 700 hPa (g g−1 m s−1). (a) June, (b) July, (c) August, (d) September, (e) October, (f) November, (g) December, (h) January, (i) February, (j) March, (k) April, (l) May. Shading represents flux magnitude.

  • View in gallery

    Normalized mean pentads, for the period of 1984–2004, of integrated zonal humidity transport from 1000 to 700 hPa, over A3 area and average precipitation in A3 (+), A4 (○), A12 (•) areas.

  • View in gallery

    Composites for normal onset (selected years from Table 4). The first line refers to 5-pentad average before onset, the second line shows the onset, and in the third line is the 5-pentad average after onset. The fourth line displays the difference between the onset and the previous 5-pentad average: (a) wind vector at 850 hPa (m s−1) and SLP (hPa; contour); (b) integrated humidity flux and magnitude (contour) from 1000 to 700 hPa (g g−1 m s−1). Shading indicates statistical significance at 95% confidence level.

  • View in gallery

    Same as Fig. 6 for (a) omega (Pa s−1) (dashed lines represent negative values); (b) streamlines at 300 hPa. Shading indicates statistical significance at 95% confidence level.

  • View in gallery

    Same as Fig. 6 but for the demise phase.

  • View in gallery

    Same as Fig. 7 but for the demise phase.

  • View in gallery

    Daily evolution of SLP (hPa), meridional wind (m s−1), temperature (°C), and specific humidity (g kg−1) at 925 hPa, latent and sensible heat flux (W m−2), soil moisture and potential evapotranspiration rate (W m−2), averaged in A12p (17.5°–22.5°S, 42.5°–47.5°W) area, from August to December: (a) 1998 (normal onset), (b) 1991 (late onset), (c) 2001 (early onset).

  • View in gallery

    Longitudinal cross section of daily integrated humidity transport from 1000 to 700 hPa (vector) and zonal humidity transport (shading) at 15°S (g g−1 m s−1) and daily precipitation time series (mm day−1) of A12p area (bar): (a) 1998 (normal onset), (b) 1991 (late onset), (c) 2001 (early onset).

  • View in gallery

    Precipitation over A12p area and geopotential anomaly at 200 hPa filtered in 30–90-day band at grid point (30°S, 50°W), during August to December: (a) 1998 (normal onset), (b) 1991 (late onset), and (c) 2001 (early onset).

  • View in gallery

    Geopotential anomaly at 200 hPa filtered in 30–90-day band (Lanczos filter; shading are negative and contours are positive values). (a) Pentad 61 (before onset of normal case), (b) pentad 62 (onset of normal case), (c) pentad 67 (before onset of late case), (d) pentad 68 (onset of late case), (e) pentad 51 (before onset of early case), (f) pentad 52 (onset of early case).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 289 289 39
PDF Downloads 255 255 33

The Life Cycle of the South American Monsoon System

View More View Less
  • 1 Pontifícia Universidade Católica de Minas Gerais, Contagem–Minas Gerais, and Centro de Previsão de Tempo e Estudos Climáticos/Instituto Nacional de Pesquisas Espaciais, Cachoeira Paulista, São Paulo, Brazil
  • | 2 Centro de Previsão de Tempo e Estudos Climáticos/Instituto Nacional de Pesquisas Espaciais, Cachoeira Paulista, São Paulo, Brazil
© Get Permissions
Full access

Abstract

The South American monsoon system (SAMS) life cycle plays an important role in the distribution and duration of the rainy season mainly over southwestern Amazonia, and the central west and southeast Brazil regions, affecting the economy through impacts on the agriculture and hydrology sectors. In this study a new criterion is applied to identify the monsoon onset and demise that was not used before in the SAMS region. This criterion is based on the atmospheric humidity flux over an area recognized as the monsoon core because of zonal wind reversal and changes in humidity from the transition seasons to summer and winter. Areas in Brazil that have a monsoon regime are identified, and several features associated with the life cycle are discussed. The climatological onset and demise are identified as the end of October and the end of March, respectively, and an interannual variability is found in the times of onset/demise. The main observed features in the two phases are discussed, such as the role of the South Atlantic subtropical high displacement, the northwestern moisture flux east of Andes and from the Atlantic Ocean, the zonal wind intensity and direction over central South America, the vertical motion over the continent, and the expansion/reduction of the Bolivian high circulation with associated high-level divergence. The frontal systems contribute to the pressure decrease, wind direction changes, and soil moisture increase previous to the onset. Low-frequency troughs with intraseasonal variability establish conditions favorable to the monsoon onset, and low-frequency ridges are related to late onset.

Corresponding author address: Adma Raia, Pontifícia Universidade Católica de Minas Gerais, MG Tempo, Rua Rio Comprido 4580 Contagem–MG, 32010-025 Brazil. Email: adma@cptec.inpe.br

Abstract

The South American monsoon system (SAMS) life cycle plays an important role in the distribution and duration of the rainy season mainly over southwestern Amazonia, and the central west and southeast Brazil regions, affecting the economy through impacts on the agriculture and hydrology sectors. In this study a new criterion is applied to identify the monsoon onset and demise that was not used before in the SAMS region. This criterion is based on the atmospheric humidity flux over an area recognized as the monsoon core because of zonal wind reversal and changes in humidity from the transition seasons to summer and winter. Areas in Brazil that have a monsoon regime are identified, and several features associated with the life cycle are discussed. The climatological onset and demise are identified as the end of October and the end of March, respectively, and an interannual variability is found in the times of onset/demise. The main observed features in the two phases are discussed, such as the role of the South Atlantic subtropical high displacement, the northwestern moisture flux east of Andes and from the Atlantic Ocean, the zonal wind intensity and direction over central South America, the vertical motion over the continent, and the expansion/reduction of the Bolivian high circulation with associated high-level divergence. The frontal systems contribute to the pressure decrease, wind direction changes, and soil moisture increase previous to the onset. Low-frequency troughs with intraseasonal variability establish conditions favorable to the monsoon onset, and low-frequency ridges are related to late onset.

Corresponding author address: Adma Raia, Pontifícia Universidade Católica de Minas Gerais, MG Tempo, Rua Rio Comprido 4580 Contagem–MG, 32010-025 Brazil. Email: adma@cptec.inpe.br

1. Introduction

Since the investigations of Zhou and Lau (1998) on the monsoon system in South America, several studies have addressed the subject, mainly related to precipitation and circulation regimes over the continent. A review of the fundamental features related to this system over South America was documented in Vera et al. (2006) and Nogués-Paegle et al. (2002). Although there is not an obvious reversal of winds over large areas of South America, as occurs in Asia and India, it is possible to identify the reversal when the annual wind average is removed, as shown by Zhou and Lau (1998). Another important monsoon characteristic is the precipitation behavior in several areas, which presents a well-defined annual cycle, with a wet phase during the summer and a dry phase during the winter (Rao et al. 1996; Gan et al. 2004). In these areas there is a dry period in winter [June–August (JJA)] followed by a fast increase in precipitation during spring [September–November (SON)], reaching a maximum during the summer season [December–February (DJF)] and decreasing during autumn [March–May (MAM); Kousky and Ropelewski 1997; Nogués-Paegle et al. (2002); Cavalcanti et al. 2002].

In general, the South American monsoon system (SAMS) onset occurs during Southern Hemisphere (SH) spring (SON), when the convection is displaced from extreme northwestern South America to central Amazonia and central Brazil (Nogués-Paegle et al. (2002)). During the SH summer (DJF) there is maximum precipitation over central Amazonia and southeastern Brazil, a change in the atmospheric circulation associated with the Bolivian high, an upper-level cyclonic vortex over Northeast Brazil and the Atlantic Ocean, and also the occurrence of the South Atlantic convergence zone (SACZ; Kousky and Ropelewski 1997; Grimm et al. 2004).

Several studies have been performed to identify the onset and demise phases of monsoon systems around the world (e.g., Holland 1986; Pearce and Monhanty 1984; Soman and Kumar 1993). A monsoon index based on the hydrological cycle was defined by Fasullo and Webster (2003) for the Indian monsoon. They used vertically integrated moisture transport to identify the onset and withdrawal of the Indian monsoon by linking this flux to the basic monsoon forces (Fasullo and Webster 2002). Particularly over South America, several criteria were applied, such as those based on wind and precipitation (e.g., Gan et al. 2004, 2006); on outgoing longwave radiation (OLR; Kousky 1988); only on precipitation (Marengo et al. 2001; Liebmann and Marengo 2001); and on combined EOFs using anomalies of precipitation, specific humidity, air temperature, and zonal and meridional winds at 850 hPa (Silva and Carvalho 2007).

The beginning and length of the rainy season as well as the precipitation distribution in the SAMS region are important issues to be understood. They affect the economy directly by their impact on agriculture and hydrological resources management, mainly over southeastern Brazil, which has the highest population concentration per square kilometer and highest gross domestic product (GDP) compared to other regions of the country. The objective of this study is to analyze the SAMS life cycle based on the Fasullo and Webster (2003) humidity transport criterion and to identify the atmospheric features associated with onset and demise. Some aspects of the monsoon life cycle, such as the vertical humidity flux analysis, the role of the Atlantic subtropical high, frontal systems, and intraseasonal variability, as well as the atmospheric characteristics at low and high levels, and other analyses that were not presented in previous studies in detail, are explored in this paper. Section 2 describes data and procedures to identify the life cycle phases. Analysis of the annual cycle in section 3 is performed to identify areas affected by the monsoon regime over South America. The SAMS life cycle, interannual variability and the atmospheric patterns associated with onset/demise are analyzed in section 4. In section 5, there is a discussion on the role of frontal systems and intraseasonal variability. Summary and conclusions are given in section 6.

2. Data and methods

Daily precipitation was obtained from a dataset interpolated to a 0.25° × 0.25° grid point field, provided by several sources, such as surface stations of the National Institute of Meteorology (INMET), the National Agency of Energy (ANEEL), and automatic stations of the Center for Weather Forecasts and Climate Studies/National Institute for Space Research (CPTEC/INPE). These data were used to analyze the mean annual cycle (1984–2004) in several areas of Brazil and also in the analysis of specific years. The onset and demise criteria calculations and atmospheric and surface field analysis were performed using National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis of daily data (Kalnay et al. 1996), except for soil moisture, which was obtained from the NCEP–Department of Energy (DOE) reanalysis 2 (Kistler et al. 2001). The assimilation of observed precipitation in this latter dataset resulted in a more realistic soil wetness.

The monsoon onset and end have been identified in previous studies using methods based on precipitation, OLR, or wind (e.g., Marengo et al. 2001; Liebmann and Marengo 2001; Kousky 1988; Gan et al. 2004). In the present study, the SAMS phases were identified using a criterion that takes into account the atmospheric circulation and also the available atmospheric humidity, represented by the atmospheric humidity transport. This criterion was based on Fasullo and Webster (2003). There are several reasons to apply this criterion, as discussed by Fasullo and Webster (2003), and they can be transferred to the SAMS region. The spatial gradient of water vapor, clouds, and precipitation contribute to the seasonal mean and interannual variation of radiation and latent heating, which drives the monsoon circulation. In addition, hydrological fields, such as moisture flux are directly associated with basic monsoon forcing (Fasullo and Webster 2003).

Another reason to apply the Fasullo and Webster criterion is that, by using wind and humidity field rather than precipitation, the monsoon onset can be monitored independent of rainfall onset. Results from numerical models could be used to forecast the monsoon onset and length, information that is useful in hydrological management. Reanalysis data and model outputs comprise variables of type A (like wind and temperature), which are influenced by data assimilation and considered the most reliable; type B (humidity), which is influenced by data assimilation and model; and type C (precipitation), which is purely derived from the model.

In the present analysis, the criterion was modified to consider only the zonal humidity flux inside the core of the South America monsoon region and the persistence of the flux. This decision was taken based on the annual cycle of the humidity transport analysis, which revealed stronger variations from the dry to the wet season in the zonal component of humidity transport than in the meridional component, in the monsoon core. In addition, the criterion to apply only the zonal component followed the analysis of both the meridional and zonal components, and each component separately. The use of zonal flux presented a better relation with precipitation than did the use of both components. A relation between precipitation and zonal wind was also shown in Gan et al. (2004), which considered the first occurrence of westerly winds and 4 mm day−1 for 75% of the subsequent 8 pentads to identify the monsoon onset in the central region of Brazil. Therefore, the criterion proposed in the present study consists of calculating the vertical integrated moisture transport by the zonal wind component (UVIMT) [Eq. (1)] for pentads of all years during the period of analysis (1984–2004) and then applying a normalization following Eq. (2) to get a result between −1 and +1. The minimum zonal humidity transport yields the value of −1, and the maximum zonal humidity transport is indicated by the value of +1. Therefore, the monsoon onset is considered when normalized UVIMT shows the first positive value followed by at least 3 positive pentads from the 4 subsequent pentads. The demise phase occurs when normalized UVIMT presents the first negative value, followed by 3 negative pentads in the 4 subsequent pentads:
i1520-0442-21-23-6227-e1
where q is the specific humidity and u is the zonal wind component.
i1520-0442-21-23-6227-e2
where X is the UVIMT annual mean and χ is the normalized time series.

In the next section, the mean annual cycle of several variables is analyzed to select areas that have a monsoon regime. Total precipitation during the monsoon life cycle and during DJF is also analyzed for each year. Additional analysis comprises discussions on the atmospheric features associated with normal and anomalous (later or earlier onset/demise) life cycle. The main atmospheric patterns associated with the onset and demise of the SAMS are discussed based on composite fields of normal and anomalous years. The synoptic systems’ role in the onset is investigated with daily analyses from August to December of 3 yr (normal, early, and late onset) identifying the atmospheric changes due to frontal systems passages over southeast Brazil. Intraseasonal variability influence is discussed based on the filtered 200-hPa geopotential in the 30–90-days band (Lanczos filter; Duchon 1979).

3. Mean annual cycle

Precipitation distribution over South America depends on geographical location and seasons. To analyze details of annual cycle in several South America regions, monthly means of observed precipitation, evapotranspiration, humidity at 925 hPa, zonal wind at 850 hPa, and surface temperature from NCEP–NCAR reanalysis were calculated in 12 areas distributed over Brazil (Fig. 1). The annual cycles of these variables are shown in Figs. 2 and 3. In most of analyzed areas (A1, A2, A3, A4, A5, A6, A10, and A12) there is a seasonal precipitation cycle very well defined, showing a maximum during DJF and a minimum in JJA, mainly in areas A3, A4, and A12, where around 50% of the annual accumulated precipitation occurs in DJF (Table 1). The Amazon region (A1, A2, and A9) presents the highest annual accumulated precipitation, and the lowest values are found over the southern northeast region (A6). Low values are also registered in the central northeast (A7) and northern southeast (A5) regions. Over the northern part of the northeast the highest values occur in MAM, the rainy season in that region, associated with the Intertropical convergence zone (ITCZ; Hastenrath and Heller 1977).

Area A11, which comprises large parts of southern Brazil and Uruguay, is the only area that does not show large precipitation variability through the year. A uniform distribution is seen through the whole year, from 22% of total precipitation in JJA to 27% in SON (Fig. 2 and Table 1). The small annual precipitation variability in this region can also be observed in Grimm et al. (2004). However, there is a slight evapotranspiration annual cycle, with maximum values from September to January and minimum values from April to July. The uniform precipitation over this area is related to the influence of transient systems, such as cold fronts, that affect this southern region during the whole year (Cavalcanti and Kousky 2003).

Considering the seasonal cycle of precipitation, the areas with a monsoon regime are A1, A2, A3, A4, A5, A6, A10, and A12. In these areas the annual cycle of evapotranspiration and specific humidity is well defined. The strongest seasonal cycle occurs in areas A3 and A4. During summer the mean precipitation rate reaches 10 mm day−1 in area A3 and the evapotranspiration reaches minimum values of 150 W m−2, while in the winter the precipitation rate decreases to 1 mm day−1 and evapotranspiration increases to 650 W m−2, characterizing a dry winter and a wet summer. In areas A3 and A4 the maximum temperatures are found in later winter/early spring, reaching values of 26°C and 27°C (Fig. 3). This maximum is associated with lack of cloudiness and rains during these seasons.

A reversal of zonal wind from October to November and from March to April occurs in Area A3, which is the only area with this behavior, with easterly winds in the winter and westerly winds in the summer. Before the wind reversal, the specific humidity has already increased in this area in the spring. When the easterlies return in April, there is a reduction of specific humidity. In this area there is also a surface heating in August and September that contributes to the atmosphere instability. The role of atmospheric instability, originating from surface heating associated with the rainy season in the Amazon region was discussed by Li and Fu (2004). The increase of humidity from September to November and the wind reversal in November are associated with the increase of precipitation in area A3 (Fig. 2). In the present study, this area is defined as the core of the South American monsoon where the UVIMT index was calculated to analyze its life cycle.

4. SAMS life cycle

a. Mean features

In this section, the transition phases of humidity transport over South America are discussed based on the vertical integrated humidity from 1000 to 700 hPa, averaged during the 1984 to 2004 period (Fig. 4). In June, July, and August, there is an intense southeasterly humidity transport over Northeast Brazil, associated with the South Atlantic subtropical high, which is close to the continent during this period, as described by Rao et al. (1996) and Doyle and Barros (2002). The intense westward humidity transport over Amazonia (without a southward component) reduces the humidity transport to extratropical latitudes of South America. From September to November, there is a gradual change in wind direction over tropical South America, which contributes to the humidity increase in the central and southeastern areas of Brazil. The maximum flux over the equatorial Atlantic in spring shifts to the north coast of South America in summer, changing direction over northern Amazonia. The humidity is thus transported from the ocean southwestward inside the continent and then, southeastward east of Andes, where the flux is intensified. The Andes cordillera affects the easterly flow as a barrier, forcing the winds to turn southeastward. The humidity flux associated with this turning reaches the SAMS core (area A3), where the wind reversal from winter to summer is seen. The maximum flux over Bolivia in the summer months can contribute to the high frequency of low-level jet (LLJ) occurrences during this season, as noticed in Marengo et al. (2004) and mentioned by Vera et al. (2006) and Grimm et al. (2004). The flux intensifies also over southeastern Brazil, and is associated with the flux intensification in the South Atlantic subtropical high. A great change is observed in the humidity flux following the subtropical high flow in the summer months. Thus, there is a large contribution of South Atlantic humidity to the rainy season of southeast Brazil, in addition to the humidity from northwestern Amazonia. From March to May (autumn), the humidity fluxes change intensity and direction again over Amazonia, east of the Andes, and in northeastern and southeastern Brazil, and the winds in the SAMS core change again to easterlies. These changes reduce the humidity transport to southeast and central Brazil, and the winter conditions begin in the following months.

Based on the summer wind reversal and the humidity flux intensification in area A3, considered the SAMS core, the monsoon phases were identified, following the modified index of Fasullo and Webster (2003) as mentioned in section 2. The annual cycle of the normalized zonal humidity transport index shows positive values, associated with a wet period, and negative values with a dry period (Fig. 5). Considering this index, the climatological onset occurs at the end of October (pentad 60: 23–27 October) and the demise at the end of March (pentad 18: 27–31 March). Using the present dataset and based on the same criterion of Gan et al. (2004), who applied an index related to zonal wind and precipitation to calculate South American monsoon phases, similar results were obtained (Table 2). Therefore, there is agreement between the two methods. The slight difference from the dates obtained in Gan et al. (2004) is probably related to the different analysis period and precipitation data.

b. Interannual variability

The modified Fasullo and Webster (2003) criterion was also applied to each year in the period from 1984 to 2004, and results show an interannual variability around the climatological value in both onset and demise phases. The onset and demise dates, the duration and accumulated precipitation during wet season and in DJF are presented in Table 3. Onset pentads could not be found in 1997/98 and 2000/01 because of high variability of the UVIMT index. During the summer of those years, low precipitation occurred mainly in area A12, likely related to El Niño conditions in 1997/98, and a strong intraseasonal oscillation in 2000/01 (Cavalcanti and Kousky 2001). Considering the other analyzed years, a standard deviation of 5 pentads around the climatological pentad occurs at the onset and demise dates, indicating a variation up to 25 days. During the analyzed period, 72% of the years presented onset around climatology plus or minus one standard deviation. In 17% of the years, onset happened later and in 11% it happened earlier. During demise, 56% of the years presented pentads around climatology plus or minus one standard deviation. Late demise occurred in 11% of total years and early demise, in 33%. This indicates that, although there is variability, the majority of years presented onset and demise in an interval that was within one standard deviation. There is a larger percentage of years with late onset (5 yr) than early onset (3 yr) and larger percentage of years with early demise (6 yr) than late demise (2 yr), but it was not possible to identify a relation between variability of the two phases, because there are years with late onset/demise, as 1986/87, and years with early onset/demise (1999/2000).

Values in Table 3 indicate that in 78% of the years the monsoon had a duration close to climatology (31 pentads) plus or minus one standard deviation. The other cases had longer (5%) or smaller (17%) duration than average. The extreme monsoon duration occurred in 2001/02 (40 pentads) and in 1991/92 (18 pentads). The accumulated precipitation between onset and demise and in the mature SAMS phase (DJF) for 2 monsoon areas, A3 and A12, is shown in Table 3. The accumulated precipitation in area A3 can be related to the monsoon duration, while this relation is not always followed in area A12, which represents southeast Brazil. Although there is not a systematic relation between accumulated precipitation and the duration of the rainy season over area A12, 60%–70% of the accumulated precipitation occurs during DJF, with maximum in December and/or January. The results also indicate that even with variability of onset/demise and length of life cycle, the accumulated precipitation has low interannual variability during the mature phase (DJF).

c. Atmospheric characteristics associated with the SAMS life cycle

The main characteristics of the SAMS life cycle were obtained through composite analyses for onset and demise of normal cases (climatological plus or minus one standard deviation), late cases (later than normal) and early cases (earlier than normal). Table 4 indicates years and pentads in each group. Atmospheric fields were analyzed according to onset/demise pentad and 5 pentads average before and after the onset/demise. A difference between two periods (onset/demise pentad − average of previous 5 pentads) was calculated to enhance the main features of monsoon transition periods.

1) Onset

The atmospheric characteristics during onset for years selected as normal (Table 4), are seen in Fig. 6. During the transition to onset, there is a reduction of sea level pressure (SLP) over the southeastern South Atlantic extending to the continent, mainly over southern Brazil, Uruguay, and Argentina, comprising the Chaco low region (Fig. 6a). During this period, the South Atlantic subtropical high is displaced eastward and shows a gradual weakening, contributing to the pressure decrease over southeastern and eastern South America. There is a noticeable change in the wind field over the continent from the previous period to the onset, mainly over extreme southwestern Amazonia, from northerlies to northwesterlies, and over eastern Brazil, where the winds turn from easterlies to northeasterlies because of the displacement eastward of the South Atlantic subtropical high. These changes in the low-level flow contribute to the increase of moisture flux from Amazonia and from South Atlantic Ocean to the monsoon region. These contributions were discussed by Lenters and Cook (1995), Rao et al. (1996), and Nogués-Paegle et al. (2002). Water vapor flux provided by the western flank of the Atlantic subtropical high is one of the mechanisms important for the generation of precipitation associated with the SACZ, a typical system of the South America summer season (Lenters and Cook 1995). Features of the circulation changes in several phases of SAMS were also discussed in Zhou and Lau (1998). The North Atlantic trade winds are intensified and there is also intensification of moisture flux over extreme northern South America, contributing to the southward flux. The cross-equatorial flow over northern South America related to the rainy season onset in Amazonia was discussed in Marengo et al. (2001), Fu et al. (1999), and Liebmann and Marengo (2001). After the onset, the westerly flow extends to central South America where there is a confluence with the flow from the subtropical high.

From the transition to the onset there is an intensification of the humidity transport east of Andes, which is the region of frequent LLJ occurrence (Marengo et al. 2004) and an increase in magnitude over the tropical Atlantic Ocean close to Northeast Brazil (Fig. 6b). The difference field enhances the westerly flux over central South America. This configuration is favorable to moisture increase over central and southeast Brazil. Although there is intensification of moisture flux east of Andes from the transition to the onset, the associated flow cannot be identified as the LLJ. Analysis of mean vertical structure of the wind magnitude and meridional component at 15° and 18°S, for the composite (not shown), indicated the occurrence of maximum values at around 850 hPa over Bolivia. However, the wind shear and intensity of this maximum required by the LLJ criterion (Marengo et al. 2004) were not observed. In addition, monthly mean analyses indicated that there were changes in this profile configuration from September/October to November, when the configuration was similar to the composite cases.

The onset is also characterized by an intensification of the ascending vertical motion over the continent with a northwest–southeast orientation extending from the extreme northwest South America to the Atlantic Ocean (Fig. 7a). This situation suggests favorable conditions for the coupling between tropical convection and frontal systems that reach the southeast region, often initiating the SACZ. Several studies have shown the influence of transient systems on the SACZ, such as Siqueira and Machado (2004) who used infrared satellite data to analyze the influence of frontal system and tropical convection on the SACZ occurrences. High-frequency wave trains crossing the extratropical South Pacific and bending over South America were related to wave pulses from frontal systems in Cavalcanti and Kayano (1999). They suggested that these systems could be responsible for maintaining the convective activity in the SACZ. The influence of transient systems and low-frequency variability on SACZ convection was also discussed in Cunningham and Cavalcanti (2006).

At high levels the anticyclonic circulation associated with the Bolivian high extends eastward and the upper-level trough over South Atlantic/Northeast Brazil is displaced eastward, from the transition to the onset (Fig. 7b). The expansion of the Bolivian high and associated divergence, as well as the amplification of the upper-level trough, represent the high-level features during the monsoon onset. The mechanisms for the Bolivian high development have been associated with diabatic heating and topography in Figueroa et al. (1995) and Gandu and Geisler (1991). In the mature phase the typical summer features become established, with the Bolivian high well organized and the upper-level trough amplified. The largest differences between the previous period and the onset are seen associated with the upper-level trough amplification over tropical Atlantic.

2) Demise

During the demise phase, there is a SLP increase over extratropical South America, as the South Atlantic subtropical high is displaced westward again (Fig. 8a). The flow east of the Andes changes from northwesterly to northerly again, and over Northeast Brazil changes from northeasterlies to easterlies (Fig. 8a). The difference field indicates the weakening of northwesterlies/westerlies over southeast Amazonia and central Brazil, and intensification of easterlies over eastern Brazil. The humidity transport behavior is opposite to that of the onset and the northwesterly flux east of the Andes changes to northerly flux, at the same time that easterly flux reaches the mountains (Fig. 8b). These features are enhanced in the difference field. During this phase, the vertical motion also has behavior opposite to that seen at the onset (Fig. 9a). There is a gradual reduction of ascending motion mainly over southeastern and eastern Brazil. Strong ascent continues over the extreme north of the northeast, consistent with the ITCZ influence during the autumn season. At upper levels, the typical summer conditions weaken; that is, there is a change in the Bolivian high and upper-level trough configuration (Fig. 9b).

The anomalous onset/demise groups showed characteristics similar to those of the normal group in several variables, which allows us to infer that the typical large-scale patterns are associated with the beginning and decay of SAMS phases whatever the period of occurrence. The positioning and intensity of the South Atlantic subtropical high and humidity flux from Amazonia, the presence of ascending (subsiding) vertical motion and the integrated humidity transport from northwest (northeast) of the Andes cordillera are associated with favorable conditions for the initiation (reduction) of convection over the monsoon region. In years with late demise, there is a delay of pressure reduction over the continent, while in early onset years the pressure is reduced earlier than the normal, allowing the occurrence of other features associated with the onset. The pressure reduction is related to the subtropical high’s eastward displacement, which occurs earlier than normal in the early monsoon, and later than normal in the late onset monsoon.

5. Influences of frontal systems and intraseasonal variability

a. Influence of frontal systems in three selected years

Frontal systems are the most common high-frequency system over South America, mainly over subtropical and middle latitudes. They affect Brazil during the whole year (Cavalcanti and Kousky 2003) and are the main contributors to the precipitation in those latitudes. Influences of frontal systems on precipitation or convection over South America were discussed in Garreaud and Wallace (1998), Siqueira and Machado (2004), and Cunningham and Cavalcanti (2006). The influence of cold air outbreaks on the southern Amazonia rainy season onset was discussed by Li and Fu (2006). Besides their contribution to rainfall, frontal systems play an important role in the heat, humidity, and momentum distribution.

The frontal systems’ role in monsoon onset was investigated in three selected years, 1998 (normal onset on pentad 62: 2–6 November); 1991 (late onset on pentad 68: 2–6 December), and 2001 (early onset on pentad 52: 13–17 September). The last two years had an extreme life cycle duration, as mentioned in section 4. Daily evolution of SLP, precipitation, meridional wind, temperature and specific humidity at 925 hPa, latent and sensible flux, soil moisture, and potential evapotranspiration, averaged in a small area inside area A12 (A12p: 17.5°–22.5°S, 42.5°–47.5°W), from August to December in these selected years, are seen in Fig. 10.

During the normal onset, several cold frontal passages are identified in the time series of Fig. 10a (SLP increase, change in meridional wind direction, temperature decrease), sometimes related to rainfall over the area, before the monsoon onset. During these occurrences, the atmosphere humidity, soil moisture, and latent heating flux increase gradually, and sensible heating flux as well as potential evapotranspiration decrease. Consistent with the discussion in section 3 of the annual behavior of the A3 area, the atmosphere humidity reaches values above 14 g kg−1 also in area A12p, remaining high after the monsoon onset. The associated surface processes in area A12p also reach maximum (soil moisture, latent heating) or minimum (sensible heating, potential evapotranspiration) values after the monsoon onset in area A3.

In the late onset case, the SLP average is higher over area A12p than the normal case, from August to middle September, and there were no precipitation episodes during this period, even with frontal system passages (Fig. 10b). When SLP decreases over the analyzed area, the first precipitation episodes occur associated with the presence of frontal systems. Two false onsets occur in October and November, when the persistence of precipitation for several days and the increase of humidity variables suggest the establishment of conditions for the beginning of the rainy season. However, after a few days there is a decrease in those variables and a reduction or lack of precipitation, characterizing the situation as a false rainy season onset. Afterward, after the passage of another frontal system, the atmospheric specific humidity and soil moisture increase again, and the area reaches necessary conditions for the beginning of the rainy season. Previous to the early onset, the SLP was also high during the first two weeks of August, but reduced sharply in the middle of September, when there was a strong frontal system passage, identified by changes in the meridional wind direction, reduction of temperature, and increase of SLP (Fig. 10c). This episode contributed to increase the atmospheric humidity, soil moisture, and latent heat flux.

A complementary analysis enhances differences in the atmospheric conditions related to the normal, late and early onset, displayed in Fig. 11, which shows a longitudinal cross section of daily integrated humidity flux from 1000 to 700 hPa and precipitation over area A12p at 15°S. In the transition period before the onset, when the passage of frontal systems is accompanied by the northwesterly humidity flux east of the Andes, there is precipitation over the area. In the normal case (Fig. 11a), there is an increase of northwesterly humidity flux and a reduction of easterly flux, associated with the eastward displacement of the South Atlantic subtropical high, reaching maximum/minimum intensity in the beginning of November (monsoon onset). Afterward, the northwesterly flux remains strong and northeasterly flux is observed on some days, indicating the contribution of the South Atlantic humidity to precipitation over southeastern Brazil. In the late onset case, there are several episodes of humidity flux increase/reduction associated with the northwesterly/easterly flux, and they reflect on the precipitation over the area (Fig. 11b). When the easterly flux is strong, extending westward, there is a reduction of humidity flux from northwest, feature that is related to the two false monsoon onsets. The interruption of the humidity flux contributed to the late establishment of the monsoon. The same mechanism occurs in the early onset case (Fig. 11c). The easterly flux was weaker than in the other two cases since August and the northwesterly flux was stronger, reaching eastern longitudes earlier than in the other two cases.

Therefore, a combination of conditions provided by frontal systems (which contribute to the atmospheric convective instability previous to the onset, as seen in Li and Fu (2006)) and by northwesterly moist flux increase in central/southeast Brazil (associated with the South Atlantic subtropical high eastward displacement) is essential to the establishment of SAMS onset.

b. Intraseasonal variability influence

Previous studies indicate that besides disturbances of high-frequency (frontal systems), fluctuations on intraseasonal time scales affect convection intensity and position over central South America (Garreaud 2000). The intraseasonal variability has influences on summer precipitation over southeast Brazil, mainly related to SACZ development. Depending on the phase of the intraseasonal variability, the SACZ is intensified or not (Nogués-Paegle and Mo 1997; Nogués-Paegle et al. 2000; Carvalho et al. 2004; Cunningham and Cavalcanti 2006). Break periods during the monsoon season have also been connected to intraseasonal variability (Jones and Carvalho 2002). Thus, in this section, the role of low-frequency (30–90 days) influence during the SAMS onset is discussed for the three years analyzed in the previous section. Time series of geopotential anomaly at 200 hPa, filtered for frequencies of 30–90 days at grid point 35°S, 55°W and precipitation pentads averaged on area A12p are analyzed in Fig. 12. This grid point was chosen to represent the position of troughs or ridges associated with vertical motion over the southeastern Brazil region.

In general, maximum precipitation was associated with low-frequency troughs over southern Brazil. Very low precipitation values occurred when low-frequency ridges were present. It is observed that in the late monsoon onset the low-frequency troughs and ridges were stronger than in the normal and early cases. Two periods of maximum precipitation occurred when persistent troughs developed, followed by persistent ridges and precipitation reduction, which were related to the two false onsets discussed in the previous section.

Low-frequency geopotential anomaly fields at 200 hPa during the onset and in the preceding pentad for 1998, 1991, and 2001 are shown in Fig. 13. In the normal case, a Pacific–South America (PSA; Mo and Paegle 2001) type wave train is identified previous to onset and a trough develops over southern Brazil. During the onset pentad, the trough intensified in the favorable position for convection over A12p area. In the late onset case, an anomalous ridge that was part of a zonal wave train over the South Pacific Ocean was located over southeastern South America in the previous pentad, a feature that is consistent with late onset. During the onset, a trough is situated over the ocean, close to southeastern Brazil, while a blocking configuration is observed over southwestern South Pacific. On the other hand, some features observed previous to the normal case occur in the early onset pentad, mainly close to South America, and a trough develops also over southern Brazil. In the three cases, the SAMS onset occurred when there was a trough over southeast South America (normal and early) or over the ocean close to southeast Brazil (late onset). A low-frequency trough over southern Brazil following the PSA pattern and its influence on SACZ convective activity (which is the main precipitation system of the South American monsoon), was also identified in studies of Cunningham and Cavalcanti (2006), Liebmann et al. (1999), and Nogués Paegle and Mo (1997).

6. Summary and conclusions

The main atmospheric large-scale features associated with the SAMS life cycle as well as the influence of intraseasonal variability and frontal systems on the onset phase have been discussed. Through the annual variability analysis, the regions affected by the monsoon regime were identified. These regions comprise Amazonia and central and southeastern Brazil. The onset and demise SAMS phases were identified based on vertical integrated humidity flux in the monsoon core area, which has a strong annual cycle of humidity variables and a wind reversal between summer and winter. The gradual monthly changes of vertical integrated humidity flux from spring to summer (onset) and from autumn to winter (demise) support also the use of this method. Interannual variability was observed in the two phases, but for the majority of years the onset and demise occurred within the normal period plus or minus one standard deviation. However, some years presented late or early onset/demise.

The atmospheric characteristics associated with the onset or demise are similar in the normal, early, and late cases, indicating the need for the establishment of a specific atmospheric pattern, such as the position and intensity of the South Atlantic subtropical high, the moisture flux intensity and direction and vertical motion over the continent. In the onset phase the South Atlantic subtropical high displaced eastward, and this feature was reflected in pressure reduction over the continent and on intensity and direction of the zonal flow over tropical and subtropical regions. The northerly moisture flux east of the Andes changed to northwesterly, bringing humidity to the central and southeast Brazil regions. At this time the Bolivian high became established and strong ascent extended northwest–southeast over the continent. During demise, there were strong sea level pressure increases over the continent and strong easterly moisture flux directed toward the Amazon region. There was a reduction of the northerly/northwesterly flow over the extreme northern South America/east of Andes, and ascending motion gradually diminished over the central and southeast regions.

Daily analysis of atmosphere and surface characteristics over southeast Brazil in three selected years indicated that previous to the monsoon onset there was a passage of several frontal systems over the region, sometimes accompanied by episodes of rainfall. They increased the soil moisture and latent heat flux, as well as contributing to the northwesterly flow increase east of the Andes, as the pressure decreases with their approach over southeast Brazil. These features and the convective instability generated by frontal systems provide favorable conditions for the beginning of the rainy season over the region. In these specific cases, the main differences between normal onset and late or early cases were very clear and related to the increase of humidity flux from the west and the reduction of the easterly component over central South America. In the late onset case, the easterly component was stronger than in the other two cases, inhibiting the westerly component and the arrival of humidity in central South America. A weaker easterly component occurred in the early onset, in August, allowing the westerly component and the humidity to reach eastern longitudes earlier than normal. The monsoon and rainy season onset were also affected by intraseasonal variability, through low-frequency trough or ridge occurrence over southern Brazil. The three selected years’ analysis indicated the establishment of a low-frequency trough over southern Brazil for the normal and early onsets. The high intensity and persistence of low-frequency troughs or ridges previous to the late onset case were associated with false rainy season onset and monsoon delay.

The SAMS life cycle is influenced by several features, such as position and intensity of the South Atlantic subtropical high, the intensity and direction of trade winds over northern South America, northwestern moisture flux east of Andes (sometimes associated with the LLJ), reduction of SLP over the continent leading to Chaco low intensification, and the Bolivian high/upper-level trough behavior. In addition to these influences, the intensity and persistence of low-frequency troughs or ridges of a PSA pattern wave train can affect the SAMS onset and the rainfall variability over southeastern Brazil. Frontal system passages over southeastern Brazil also contribute to the monsoon onset, through the reduction of pressure, change of wind direction and increase of convective instability, soil moisture and humidity. Since SAMS is a complex system, its characteristics depend on a combination of these features to establish favorable conditions for monsoon development. Additional investigations on the role of teleconnections and their influence also on the demise phase, as well as the role of frontal systems during this phase, need to be explored in future studies.

Acknowledgments

Thanks are due to CNPq for research support (306319/2004-7) and to FAPESP Serra do Mar project (04/09469-0). Thanks also to Dr. Ruibran dos Reis and to the anonymous reviewers for their comments and suggestions.

REFERENCES

  • Carvalho, L. M. V., , C. Jones, , and B. Liebmann, 2004: The South Atlantic convergence zone: Intensity, form, persistence, relationships with intraseasonal to interannual activity and extreme rainfall. J. Climate, 17 , 88108.

    • Search Google Scholar
    • Export Citation
  • Cavalcanti, I. F. A., , and M. K. Kayano, 1999: High-frequency patterns of the atmospheric circulation over Southern Hemisphere and South America. Meteor. Atmos. Phys., 69 , 179193.

    • Search Google Scholar
    • Export Citation
  • Cavalcanti, I. F. A., , and V. E. Kousky, cited. 2001: Drought in Brazil during summer and fall 2001 and associated atmospheric circulation features. [Available online at http://www.cptec.inpe.br/products/climanalise/artigos/criseing.PDF.].

  • Cavalcanti, I. F. A., , and V. Kousky, 2003: Climatology of the South American cold fronts. Preprints, Seventh Int. Conf. on Southern Hemisphere Meteorology and Oceanography, Wellington, New Zealand, Amer. Meteor. Soc., 6.5.

  • Cavalcanti, I. F. A., , C. K. Folland, , and A. W. Colman, 2002: Note on “Predictability of Northeast Brazil rainfall and real-time forecast skill, 1987–98”. J. Climate, 15 , 19931996.

    • Search Google Scholar
    • Export Citation
  • Cunningham, C. A. C., , and I. F. A. Cavalcanti, 2006: Intraseasonal modes of variability affecting the South Atlantic Convergence Zone. Int. J. Climatol., 26 , 11651180.

    • Search Google Scholar
    • Export Citation
  • Doyle, M. E., , and V. R. Barros, 2002: Midsummer low-level circulation and precipitation in subtropical South America and related sea surface temperature anomalies in the South Atlantic. J. Climate, 15 , 33943410.

    • Search Google Scholar
    • Export Citation
  • Duchon, C. E., 1979: Lanczos filtering in one and two dimensions. J. Appl. Meteor., 18 , 10161022.

  • Fasullo, J., , and P. J. Webster, 2002: Hydrologic signatures relating the Asian summer monsoon and ENSO. J. Climate, 15 , 30823095.

  • Fasullo, J., , and P. J. Webster, 2003: A hydrological definition of Indian monsoon onset and withdrawal. J. Climate, 16 , 32003211.

  • Figueroa, S. N., , P. Satyamurty, , and P. L. S. Dias, 1995: Simulations of the summer circulation over the South American region with an Eta coordinate model. J. Atmos. Sci., 52 , 15731584.

    • Search Google Scholar
    • Export Citation
  • Fu, R., , B. Zhu, , and R. E. Dickinson, 1999: How do atmosphere and land surface influence seasonal changes of convection in the tropical Amazon. J. Climate, 12 , 13061321.

    • Search Google Scholar
    • Export Citation
  • Gan, M. A., , V. E. Kousky, , and C. F. Ropelewski, 2004: The South America monsoon circulation and its relationship to rainfall over west-central Brazil. J. Climate, 17 , 4766.

    • Search Google Scholar
    • Export Citation
  • Gan, M. A., , V. B. Rao, , and C. L. Moscati, 2006: South American monsoon indices. Atmos. Sci. Lett., 6 , 219223.

  • Gandu, A. W., , and J. E. Geisler, 1991: A primitive equations model study of the effect of topography on the summer circulation over tropical South America. J. Atmos. Sci., 48 , 18221836.

    • Search Google Scholar
    • Export Citation
  • Garreaud, R., 2000: Intraseasonal variability of moisture and rainfall over the South American altiplano. Mon. Wea. Rev., 128 , 33373346.

    • Search Google Scholar
    • Export Citation
  • Garreaud, R., , and J. M. Wallace, 1998: Summertime incursions of midlatitude air into subtropical and tropical South America. Mon. Wea. Rev., 126 , 27132733.

    • Search Google Scholar
    • Export Citation
  • Grimm, A. M., , C. Vera, , and R. Mechoso, 2004: The South American Monsoon System. The Global Monsoon System: Research and Forecast, WMO Tech. Doc. 1266, TMRP Rep. 70, 219–238.

  • Hastenrath, S., , and L. Heller, 1977: Dynamics of climatic hazards in Northeast Brazil. Quart. J. Roy. Meteor. Soc., 103 , 7792.

  • Holland, G., 1986: Interannual variability of the Australian monsoon at Darwin. Mon. Wea. Rev., 114 , 594604.

  • Jones, C., , and L. M. V. Carvalho, 2002: Active and break phases in the South American monsoon system. J. Climate, 15 , 905914.

  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Kistler, R., and Coauthors, 2001: The NCEP–NCAR 50-Year Reanalysis: Monthly means CD-ROM and documentation. Bull. Amer. Meteor. Soc., 82 , 247267.

    • Search Google Scholar
    • Export Citation
  • Kousky, V., 1988: Pentad outgoing longwave radiation climatology for the South American sector. Rev. Bras. Meteor., 3 , 217231.

  • Kousky, V., , and C. F. Ropelewski, 1997: The Tropospheric Seasonally Varying Mean Climate over the Western Hemisphere (1979–1995). Atlas No. 3, NCEP/Climate Prediction Center, 135 pp.

    • Search Google Scholar
    • Export Citation
  • Lenters, J., , and K. Cook, 1995: Simulation and diagnosis of the regional summertime precipitation climatology of South America. J. Climate, 8 , 29883005.

    • Search Google Scholar
    • Export Citation
  • Li, W., , and R. Fu, 2004: Transition of the large-scale atmospheric and land surface conditions from the dry to the wet season over Amazonia as diagnosed by the ECMWF reanalysis. J. Climate, 17 , 26372651.

    • Search Google Scholar
    • Export Citation
  • Li, W., , and R. Fu, 2006: Influence of cold air intrusions on the wet season onset over Amazonia. J. Climate, 19 , 257275.

  • Liebmann, B., , and J. Marengo, 2001: The seasonality and interannual variability of rainfall in the Brazilian Amazon basin. J. Climate, 14 , 43084318.

    • Search Google Scholar
    • Export Citation
  • Liebmann, B., , G. N. Kiladis, , J. A. Marengo, , T. Ambrizzi, , and J. D. Glick, 1999: Submonthly convective variability over South America and the South Atlantic convergence zone. J. Climate, 12 , 18771891.

    • Search Google Scholar
    • Export Citation
  • Marengo, J., , B. Liebmann, , V. Kousky, , N. Filizola, , and I. Wainer, 2001: On the onset and end of the rainy season in the Brazilian Amazon basin. J. Climate, 14 , 833852.

    • Search Google Scholar
    • Export Citation
  • Marengo, J., , W. R. Soares, , C. Saulo, , and M. Nicolini, 2004: Climatology of the low-level jet east of the Andes as derived from the NCEP–NCAR reanalyses: Characteristics and temporal variability. J. Climate, 17 , 22612280.

    • Search Google Scholar
    • Export Citation
  • Mo, K. C., , and J. N. Paegle, 2001: The Pacific–South American modes and their downstream effects. Int. J. Climatol., 21 , 12111229.

  • Nogués-Paegle, J., , and K. Mo, 1997: Alternating wet and dry conditions over South America during summer. Mon. Wea. Rev., 125 , 279291.

    • Search Google Scholar
    • Export Citation
  • Nogués-Paegle, J., , L. A. Byerle, , and K. C. Mo, 2000: Intraseasonal modulation of South American summer precipitation. Mon. Wea. Rev., 128 , 837850.

    • Search Google Scholar
    • Export Citation
  • Nogués-Paegle, J., and Coauthors, 2002: Progress in Pan American CLIVAR research: Understanding the South American monsoon. Meteorologica, 1–2 , 332.

    • Search Google Scholar
    • Export Citation
  • Pearce, R. P., , and U. C. Monhanty, 1984: Onsets of the Asia summer monsoon 1979–82. J. Atmos. Sci., 41 , 16201639.

  • Rao, V. B., , I. F. A. Cavalcanti, , and K. Hada, 1996: Annual variation of rainfall over Brazil and water vapor characteristics over South America. J. Geophys. Res., 101 , 2653926551.

    • Search Google Scholar
    • Export Citation
  • Silva, A. E., , and L. M. V. Carvalho, 2007: Large-scale index for South American Monsoon (LISAM). Atmos. Sci. Lett., 8 , 5157.

  • Siqueira, J. R., , and L. A. T. Machado, 2004: Influence of the frontal systems on the day-to-day convection variability over South America. J. Climate, 17 , 17541766.

    • Search Google Scholar
    • Export Citation
  • Soman, M. K., , and K. K. Kumar, 1993: Space–time evolution of meteorological features associated with the onset of Indian summer monsoon. Mon. Wea. Rev., 121 , 11771194.

    • Search Google Scholar
    • Export Citation
  • Vera, C., and Coauthors, 2006: Toward a unified view of the American monsoon systems. J. Climate, 19 , 49775000.

  • Zhou, J., , and K. M. Lau, 1998: Does a monsoon climate exist over South America? J. Climate, 11 , 10201040.

Fig. 1.
Fig. 1.

Twelve areas for the mean annual cycle analysis. The shaded areas correspond to the 8 areas that have a monsoon regime.

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 2.
Fig. 2.

Mean annual cycle (1984–2004) of precipitation (mm day−1) (▪), specific humidity at 925 hPa (g kg−1) (○), and potential evapotranspiration rate (W m−2) (dot–dash line) in 12 selected areas of Fig. 1.

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 3.
Fig. 3.

Mean annual cycle (1984–2004) of zonal wind at 850 hPa (m s−1) (○) and surface temperature (°C) (▪) in 12 selected areas of Fig. 1.

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 4.
Fig. 4.

Monthly mean (1984–2004) integrated humidity transport from 1000 to 700 hPa (g g−1 m s−1). (a) June, (b) July, (c) August, (d) September, (e) October, (f) November, (g) December, (h) January, (i) February, (j) March, (k) April, (l) May. Shading represents flux magnitude.

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 5.
Fig. 5.

Normalized mean pentads, for the period of 1984–2004, of integrated zonal humidity transport from 1000 to 700 hPa, over A3 area and average precipitation in A3 (+), A4 (○), A12 (•) areas.

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 6.
Fig. 6.

Composites for normal onset (selected years from Table 4). The first line refers to 5-pentad average before onset, the second line shows the onset, and in the third line is the 5-pentad average after onset. The fourth line displays the difference between the onset and the previous 5-pentad average: (a) wind vector at 850 hPa (m s−1) and SLP (hPa; contour); (b) integrated humidity flux and magnitude (contour) from 1000 to 700 hPa (g g−1 m s−1). Shading indicates statistical significance at 95% confidence level.

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 7.
Fig. 7.

Same as Fig. 6 for (a) omega (Pa s−1) (dashed lines represent negative values); (b) streamlines at 300 hPa. Shading indicates statistical significance at 95% confidence level.

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 8.
Fig. 8.

Same as Fig. 6 but for the demise phase.

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 9.
Fig. 9.

Same as Fig. 7 but for the demise phase.

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 10.
Fig. 10.

Daily evolution of SLP (hPa), meridional wind (m s−1), temperature (°C), and specific humidity (g kg−1) at 925 hPa, latent and sensible heat flux (W m−2), soil moisture and potential evapotranspiration rate (W m−2), averaged in A12p (17.5°–22.5°S, 42.5°–47.5°W) area, from August to December: (a) 1998 (normal onset), (b) 1991 (late onset), (c) 2001 (early onset).

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 11.
Fig. 11.

Longitudinal cross section of daily integrated humidity transport from 1000 to 700 hPa (vector) and zonal humidity transport (shading) at 15°S (g g−1 m s−1) and daily precipitation time series (mm day−1) of A12p area (bar): (a) 1998 (normal onset), (b) 1991 (late onset), (c) 2001 (early onset).

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 12.
Fig. 12.

Precipitation over A12p area and geopotential anomaly at 200 hPa filtered in 30–90-day band at grid point (30°S, 50°W), during August to December: (a) 1998 (normal onset), (b) 1991 (late onset), and (c) 2001 (early onset).

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Fig. 13.
Fig. 13.

Geopotential anomaly at 200 hPa filtered in 30–90-day band (Lanczos filter; shading are negative and contours are positive values). (a) Pentad 61 (before onset of normal case), (b) pentad 62 (onset of normal case), (c) pentad 67 (before onset of late case), (d) pentad 68 (onset of late case), (e) pentad 51 (before onset of early case), (f) pentad 52 (onset of early case).

Citation: Journal of Climate 21, 23; 10.1175/2008JCLI2249.1

Table 1.

Annual climatological precipitation and percentages for each season over the 12 areas of Fig. 1.

Table 1.
Table 2.

Climatological onset and demise pentads.

Table 2.
Table 3.

Onset and demise pentads, accumulated precipitation during the monsoon life cycle and during DJF in areas A3 and A12, and monsoon duration. (Onset could not be determined by the method in 1997/98 and 2000/01)

Table 3.
Table 4.

Set of years with normal, late, and early onset/demise.

Table 4.
Save