• Barros, V., , M. Doyle, , M. González, , I. Camilloni, , R. Bejaran, , and R. M. Caffera, 2002: Climate variability over subtropical South America and the South American monsoon: A review. Meteorologica, 27 , 3357.

    • Search Google Scholar
    • Export Citation
  • Beljaars, A. C. M., , P. Viterbo, , M. J. Miller, , and A. K. Betts, 1996: The anomalous rainfall over the United States during July 1993: Sensitivity to land surface parameterization and soil moisture anomalies. Mon. Wea. Rev., 124 , 362383.

    • Search Google Scholar
    • Export Citation
  • Berbery, E. H., , and E. A. Collini, 2000: Springtime precipitation and water vapor flux over southeastern South America. Mon. Wea. Rev., 128 , 13281346.

    • Search Google Scholar
    • Export Citation
  • Berbery, E. H., , and V. R. Barros, 2002: The hydrologic cycle of the La Plata basin in South America. J. Hydrometeor., 3 , 630645.

  • Berbery, E. H., , E. A. Collini, , V. Silva, , M. Peña, , D. L. Lettenmaier, , V. R. Barros, , and M. Pyle, 2004: The water cycle of La Plata basin at regional scales. GEWEX News, Vol. 14, No. 1, International GEWEX Project Office, Silver Spring, MD, 4–5.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , and M. J. Miller, 1986: A new convective adjustment scheme. Part II: Single column tests using GATE wave, BOMEX, ATEX and Arctic air-mass data sets. Quart. J. Roy. Meteor. Soc., 112 , 693710.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , and C. Jakob, 2002: Evaluation of the diurnal cycle of precipitation, surface thermodynamics, and surface fluxes in the ECMWF model using LBA data. J. Geophys. Res., 107 .8045, doi:10.1029/2001JD000427.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , and P. Viterbo, 2005: Land-surface, boundary layer, and cloud-field coupling over the southwestern Amazon in ERA-40. J. Geophys. Res., 110 .D14108, doi:10.1029/2004JD005702.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , J. H. Ball, , A. C. M. Beljaars, , M. J. Miller, , and P. Viterbo, 1996: The land surface–atmosphere interaction: A review based on observational and global modeling perspectives. J. Geophys. Res., 101 , D3. 72097226.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , J. D. Fuentes, , M. Garstang, , and J. H. Ball, 2002: Surface diurnal cycle and boundary layer structure over Rondônia during the rainy season. J. Geophys. Res., 107 .8065, doi:10.1029/2001JD000356.

    • Search Google Scholar
    • Export Citation
  • Black, T. L., 1994: The new NMC mesoscale Eta Model: Description and forecast examples. Wea. Forecasting, 9 , 265278.

  • Bosilovich, M. G., , and W-Y. Sun, 1999: Numerical simulation of the 1993 of the midwestern flood: Land–atmosphere interactions. J. Climate, 12 , 14901505.

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

    • Search Google Scholar
    • Export Citation
  • Chou, S. C., , C. A. S. Tanajura, , Y. Xue, , and C. A. Nobre, 2002: Validation of the coupled Eta/SSiB model over South America. J. Geophys. Res., 107 .8088, doi:10.1029/2000JD000270.

    • Search Google Scholar
    • Export Citation
  • de Goncalves, L. G. G., , W. J. Shuttleworth, , S. C. Chou, , Y. Xue, , P. R. Houser, , D. L. Toll, , J. Marengo, , and M. Rodell, 2006: Impact of different initial soil moisture fields on Eta model weather forecasts for South America. J. Geophys. Res., 111 .D17102, doi:10.1029/2005JD006309.

    • Search Google Scholar
    • Export Citation
  • Delworth, T., , and S. Manabe, 1988: The influence of potential evaporation on the variabilities of simulated soil wetness and climate. J. Climate, 1 , 523547.

    • Search Google Scholar
    • Export Citation
  • Delworth, T., , and S. Manabe, 1989: The influence of soil wetness on near-surface atmospheric variability. J. Climate, 2 , 14471462.

  • Diaz, A., , and P. Aceituno, 2003: Atmospheric circulation anomalies during episodes of enhanced and reduced convective cloudiness over Uruguay. J. Climate, 16 , 31713185.

    • 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
  • Ek, M. B., , K. E. Mitchell, , Y. Lin, , P. Grunmann, , E. Rogers, , G. Gayno, , V. Koren, , and J. D. Tarpley, 2003: Implementation of Noah land surface model advances in the National Centers for Environmental Prediction operational mesoscale Eta Model. J. Geophys. Res., 108 .8851, doi:10.1029/2002JD003296.

    • Search Google Scholar
    • Export Citation
  • Eltahir, E. A. B., 1998: A soil moisture–rainfall feedback mechanism. 1. Theory and observations. Water Resour. Res., 34 , 765776.

  • Fan, Y., , and H. Van den Dool, 2004: Climate Prediction Center global monthly soil moisture data set at 0.5° resolution for 1948 to present. J. Geophys. Res., 109 .D10102, doi:10.1029/2003JD004345.

    • 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
  • Garratt, J. R., 1993: Sensitivity of climate simulations to land-surface and atmospheric boundary-layer treatments—A review. J. Climate, 6 , 419448.

    • Search Google Scholar
    • Export Citation
  • GLACE Team, 2004: Regions of strong coupling between soil moisture and precipitation. Science, 305 , 11381140.

  • Hahmann, A. N., , and R. E. Dickinson, 1997: RCCM2-BATS model over tropical South America: Applications to tropical deforestation. J. Climate, 10 , 19441964.

    • Search Google Scholar
    • Export Citation
  • Herdies, D. L., , A. da Silva, , M. A. F. Silva Dias, , and R. Nieto Ferreira, 2002: Moisture budget of the bimodal pattern of the summer circulation over South America. J. Geophys. Res., 107 .8075, doi:10.1029/2001JD000997.

    • Search Google Scholar
    • Export Citation
  • Horel, J. D., , A. N. Hahmann, , and J. Geisler, 1989: An investigation of the annual cycle of convective activity over the tropical Americas. J. Climate, 2 , 13881403.

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

  • Kodama, Y. M., 1992: Large-scale common features of the subtropical precipitation zones (the Baiu frontal zone, the SPCZ, and the SACZ). Part I: Characteristics of subtropical frontal zones. J. Meteor. Soc. Japan, 70 , 813836.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., , M. J. Suarez, , and M. Heiser, 2000: Variance and predictability of precipitation at seasonal-to-interannual timescales. J. Hydrometeor., 1 , 2646.

    • Search Google Scholar
    • Export Citation
  • Kumar, A., , A. G. Barnston, , and M. P. Hoerling, 2001: Seasonal predictions, probabilistic verifications, and ensemble size. J. Climate, 14 , 16711676.

    • Search Google Scholar
    • Export Citation
  • Janjić, Z. I., 1990: The step-mountain coordinate: Physical package. Mon. Wea. Rev., 118 , 14291443.

  • Janjić, Z. I., 1994: The step-mountain eta coordinate model: Further developments of the convection, viscous sublayer, and turbulence closure schemes. Mon. Wea. Rev., 122 , 927945.

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

  • Labraga, J. C., , O. Frumento, , and M. López, 2000: The atmospheric water vapor cycle in South America and the tropospheric circulation. J. Climate, 13 , 18991915.

    • Search Google Scholar
    • Export Citation
  • Lenters, J. D., , and K. H. Cook, 1997: On the origin of the Bolivian high and related circulation features of the South American climate. J. Atmos. Sci., 54 , 656678.

    • 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 Re-Analysis. J. Climate, 17 , 26372651.

    • 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
  • Liebmann, B., , G. N. Kiladis, , C. S. Vera, , A. C. Saulo, , and L. M. V. Carvalho, 2004: Subseasonal variations of rainfall in South America in the vicinity of the low-level jet east of the Andes and comparison to those in the South Atlantic convergence zone. J. Climate, 17 , 38293842.

    • Search Google Scholar
    • Export Citation
  • Marengo, J. A., , 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
  • McCorcle, M. D., 1988: Simulation of surface-moisture effects on the Great Plains low-level jet. Mon. Wea. Rev., 116 , 17051720.

  • Mesinger, F., 1997: Dynamics of limited-area models: Formulation and numerical methods. Meteor. Atmos. Phys., 63 , 314.

  • Nicolini, M., , and A. C. Saulo, 2006: Modeled Chaco low-level jets and related precipitation patterns during the 1997–1998 warm season. Meteor. Atmos. Phys., 94 , 129143.

    • Search Google Scholar
    • Export Citation
  • Nogués-Paegle, J., , and K. C. 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., and Coauthors, 2002: Progress in Pan American CLIVAR research: Understanding the South American Monsoon. Meteorologica, 27 , 332.

    • Search Google Scholar
    • Export Citation
  • Nogués-Paegle, J., , E. H. Berbery, , E. A. Collini, , and C. Saulo, 2006: An evaluation of the Eta model during SALLJEX. Proc. Eighth Int. Conf. on Southern Hemisphere Meteorology and Oceanography, Foz do Iguaçu, Brazil, INPE, 51–56.

  • Paegle, J., , K. C. Mo, , and J. Nogués-Paegle, 1996: Dependence of simulated precipitation on surface evaporation during the 1993 United States summer floods. Mon. Wea. Rev., 124 , 345361.

    • Search Google Scholar
    • Export Citation
  • Rodwell, M. J., , and B. J. Hoskins, 2001: Subtropical anticyclones and summer monsoons. J. Climate, 14 , 31923211.

  • Satyamurty, P., , C. C. Ferreira, , and M. A. Gan, 1990: Cyclonic vortices over South America. Tellus, 42A , 194201.

  • Saulo, A. C., , M. Nicolini, , and S. C. Chou, 2000: Model characterization of the South American low-level flow during the 1997–1998 spring-summer season. Climate Dyn., 16 , 867881.

    • Search Google Scholar
    • Export Citation
  • Schaake, J. C., and Coauthors, 2004: An intercomparison of soil moisture fields in the North American Land Data Assimilation System (NLDAS). J. Geophys. Res., 109 .D01S90, doi:10.1029/2002JD003309.

    • Search Google Scholar
    • Export Citation
  • Seth, A., , and M. Rojas, 2003: Simulation and sensitivity in a nested modeling system for South America. Part I: Reanalyses boundary forcing. J. Climate, 16 , 24372453.

    • Search Google Scholar
    • Export Citation
  • Silva, V. B. S., , and E. H. Berbery, 2006: Intense rainfall events affecting the La Plata basin. J. Hydrometeor., 7 , 769787.

  • Vera, C., and Coauthors, 2006: The South American Low-Level Jet Experiment. Bull. Amer. Meteor. Soc., 87 , 6377.

  • Warner, T. T., , R. A. Peterson, , and R. E. Treadon, 1997: A tutorial on lateral boundary conditions as a basic and potentially serious limitation to regional numerical weather prediction. Bull. Amer. Meteor. Soc., 78 , 25992617.

    • Search Google Scholar
    • Export Citation
  • Xie, P., , and P. A. Arkin, 1997: Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor. Soc., 78 , 25392558.

    • Search Google Scholar
    • Export Citation
  • Xue, Y., , F. de Sales, , W-P. Li, , C. R. Mechoso, , C. A. Nobre, , and H-M. Juang, 2006: Role of land surface processes in South American monsoon development. J. Climate, 19 , 741762.

    • Search Google Scholar
    • Export Citation
  • Zhong, S., , J. D. Fast, , and X. Bian, 1996: A case study of the Great Plains low-level jet using wind profiler network data and a high-resolution mesoscale model. Mon. Wea. Rev., 124 , 785806.

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

  • View in gallery

    The regional Eta model domain (shaded area) and the analysis domain (rectangle).

  • View in gallery

    (a) Ratio (dimensionless) between potential evaporation and precipitation during SON. The shaded region represents intermediate values of magnitude 1–4. (b) Surface runoff (mm day−1), also for SON. All terms are 24-yr averages (1980–2003) computed from the NCEP–NCAR Reanalysis Project (NNRP) analyses. The square box in the panels represents the monsoon region.

  • View in gallery

    (a) 1980–2003 mean OND precipitation (mm day−1) estimated from CMAP; (b) CPC 1980–2003 mean OND soil moisture (mm); (c) CMAP precipitation averaged for October 1981, 1982, 1983, and 1989 for comparison to the model simulations. The square box in the panels represents the monsoon region. See text for further explanation and references.

  • View in gallery

    (a) October precipitation and (b) top-layer soil moisture computed from the control ensemble of the Eta model simulations. Units of precipitation are mm day−1 with values larger than 5 mm day−1 shaded. Soil moisture units are m3 m−3 with values larger than 0.25 shaded. The square box in the panels represents the monsoon region.

  • View in gallery

    Land-only precipitation anomalies for (a) the ensemble of reduced initial soil moisture simulations and (b) the ensemble of increased initial soil moisture simulations. Units are mm day−1 and values larger than +/− 0.2 mm day−1 are shaded dark/light. The square box in the panels represents the monsoon region.

  • View in gallery

    Area average (20°–10°S, 55°–45°W) of the monthly precipitation changes (%) as a function of the initial soil moisture changes (%) for the individual simulations that represent the ensembles in this study.

  • View in gallery

    Monsoon area-averaged time series and mean diurnal cycle of (a), (b) sensible heat flux, (c), (d) latent heat flux, and (e), (f) Bowen ratio. Units of the sensible and latent heat fluxes are W m−2. The Bowen ratio is dimensionless. The solid line is the control ensemble, and the dashed line is the reduced soil moisture ensemble.

  • View in gallery

    As in Fig. 7 but for (a), (b) the outgoing longwave radiation, (c), (d) incoming shortwave radiation, and (e), (f) total radiation. The dotted line in (f) is the net radiation difference between the control and reduced soil moisture ensembles, with the scale on the outer left side. Units are W m−2.

  • View in gallery

    (a) Temperature and dewpoint distribution in the boundary layer for the control (solid/dashed lines), and reduced (open circles) and increased (closed circles) soil moisture ensembles. The inset represents thermodynamic structure of the control ensemble in the troposphere up to 100 hPa. (b) Amplitude of the diurnal cycle of temperature for the control (solid line) and reduced soil moisture (dashed line) ensembles. (c) As in (b) but for the dewpoint. All units are °C.

  • View in gallery

    As in Fig. 7 but for (a), (b) CAPE; (c), (d) CIN; and (e), (f) precipitation. Units of CAPE and CIN are W m−2. Units of precipitation are mm day−1 for (e) and mm (3 h)−1 for (f).

  • View in gallery

    (a) Vertically integrated moisture flux and (b) its convergence for the control ensemble; (c), (d) the corresponding anomalies for the reduced soil moisture ensemble. The vertically integrated moisture flux has units of kg m−1 s−1 and its convergence has units of mm day−1. Moisture flux convergence is defined as positive and the divergence as negative. In (b), positive values (convergence) are shaded. In (d), negative values (divergence anomalies, or reduction of convergence) are shaded. The square box in the panels represents the monsoon region.

  • View in gallery

    Cross sections along line AB in Fig. 11 of (a) winds of the control ensemble (m s−1); (b) wind anomalies (w.r.t. the control ensemble) of the reduced soil moisture case (m s−1); (c) moisture flux of the control ensemble (g kg−1 m s−1); and (d) anomalies of moisture flux for the reduced soil moisture ensemble (g kg−1 m s−1).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 47 47 9
PDF Downloads 43 43 6

How Does Soil Moisture Influence the Early Stages of the South American Monsoon?

View More View Less
  • 1 Argentine Navy Weather Service/Argentine Hydrographic Service, Buenos Aires, Argentina
  • | 2 Department of Atmospheric and Oceanic Science, and ESSIC, University of Maryland, College Park, College Park, Maryland
  • | 3 Department of Atmospheric and Oceanic Sciences, University of Buenos Aires, Buenos Aires, Argentina
  • | 4 National Centers for Environmental Prediction, Camp Springs, Maryland
© Get Permissions
Full access

Abstract

This article discusses the feedbacks between soil moisture and precipitation during the early stages of the South American monsoon. The system achieves maximum precipitation over the southern Amazon basin and the Brazilian highlands during the austral summer. Monsoon changes are associated with the large-scale dynamics, but during its early stages, when the surface is not sufficiently wet, soil moisture anomalies may also modulate the development of precipitation. To investigate this, sensitivity experiments to initial soil moisture conditions were performed using month-long simulations with the regional mesoscale Eta model. Examination of the control simulations shows that they reproduce all major features and magnitudes of the South American circulation and precipitation patterns, particularly those of the monsoon. The surface sensible and latent heat fluxes, as well as precipitation, have a diurnal cycle whose phase is consistent with previous observational studies. The convective inhibition is smallest at the time of the precipitation maximum, but the convective available potential energy exhibits an unrealistic morning maximum that may result from an early boundary layer mixing.

The sensitivity experiments show that precipitation is more responsive to reductions of soil moisture than to increases, suggesting that although the soil is not too wet, it is sufficiently humid to easily reach levels where soil moisture anomalies stop being effective in altering the evapotranspiration and other surface and boundary layer variables. Two mechanisms by which soil moisture has a positive feedback with precipitation are discussed. First, the reduction of initial soil moisture leads to a smaller latent heat flux and a larger sensible heat flux, and both contribute to a larger Bowen ratio. The smaller evapotranspiration and increased sensible heat flux lead to a drier and warmer boundary layer, which in turn reduces the atmospheric instability. Second, the deeper (and drier) boundary layer is related to a stronger and higher South American low-level jet (SALLJ). However, because of the lesser moisture content, the SALLJ carries less moisture to the monsoon region, as evidenced by the reduced moisture fluxes and their convergence. The two mechanisms—reduced convective instability and reduced moisture flux convergence—act concurrently to diminish the core monsoon precipitation.

Corresponding author address: Ernesto Hugo Berbery, Department of Atmospheric and Oceanic Science, and ESSIC, 3427 Computer and Space Sciences Building, University of Maryland, College Park, College Park, MD 20742-2425. Email: berbery@atmos.umd.edu

Abstract

This article discusses the feedbacks between soil moisture and precipitation during the early stages of the South American monsoon. The system achieves maximum precipitation over the southern Amazon basin and the Brazilian highlands during the austral summer. Monsoon changes are associated with the large-scale dynamics, but during its early stages, when the surface is not sufficiently wet, soil moisture anomalies may also modulate the development of precipitation. To investigate this, sensitivity experiments to initial soil moisture conditions were performed using month-long simulations with the regional mesoscale Eta model. Examination of the control simulations shows that they reproduce all major features and magnitudes of the South American circulation and precipitation patterns, particularly those of the monsoon. The surface sensible and latent heat fluxes, as well as precipitation, have a diurnal cycle whose phase is consistent with previous observational studies. The convective inhibition is smallest at the time of the precipitation maximum, but the convective available potential energy exhibits an unrealistic morning maximum that may result from an early boundary layer mixing.

The sensitivity experiments show that precipitation is more responsive to reductions of soil moisture than to increases, suggesting that although the soil is not too wet, it is sufficiently humid to easily reach levels where soil moisture anomalies stop being effective in altering the evapotranspiration and other surface and boundary layer variables. Two mechanisms by which soil moisture has a positive feedback with precipitation are discussed. First, the reduction of initial soil moisture leads to a smaller latent heat flux and a larger sensible heat flux, and both contribute to a larger Bowen ratio. The smaller evapotranspiration and increased sensible heat flux lead to a drier and warmer boundary layer, which in turn reduces the atmospheric instability. Second, the deeper (and drier) boundary layer is related to a stronger and higher South American low-level jet (SALLJ). However, because of the lesser moisture content, the SALLJ carries less moisture to the monsoon region, as evidenced by the reduced moisture fluxes and their convergence. The two mechanisms—reduced convective instability and reduced moisture flux convergence—act concurrently to diminish the core monsoon precipitation.

Corresponding author address: Ernesto Hugo Berbery, Department of Atmospheric and Oceanic Science, and ESSIC, 3427 Computer and Space Sciences Building, University of Maryland, College Park, College Park, MD 20742-2425. Email: berbery@atmos.umd.edu

1. Introduction

The warm-season precipitation in subtropical South America exhibits the typical features of a monsoon system, with strong precipitation developing during the austral spring over the highlands of central and southeastern Brazil (Horel et al. 1989; Zhou and Lau 1998; Nogués-Paegle et al. 2002). There, the monsoon precipitation accounts for about 90% of the annual total (Gan et al. 2004). The intense precipitation extends over the Atlantic Ocean in a band known as the South Atlantic convergence zone (SACZ; Kodama 1992). The SACZ is not a uniform band, with differences between the continental and oceanic portions (Carvalho et al. 2004). The monsoon system has active and break phases lasting a few days (Jones and Carvalho 2002) and is also modulated at subseasonal time scales by the passage of large-scale Rossby waves (Liebmann et al. 1999, 2004). Associated with the large precipitation, the circulation is defined by an upper-air high pressure system called the Bolivian high (e.g., Lenters and Cook 1997), the thermal low called Chaco low (e.g., Satyamurty et al. 1990; Gan et al. 2004), and an important inflow of moisture supplied by the low-level jet (LLJ) east of the Andes or South American LLJ (SALLJ) (see, e.g., Barros et al. 2002; Marengo et al. 2004; Silva and Berbery 2006). Although the LLJ is present throughout the year (Berbery and Barros 2002), during the warm season it helps maintain the dynamic balance of the monsoon system (Rodwell and Hoskins 2001).

The monsoon precipitation and a precipitation center to the south form a dipole or seesaw pattern, so that when precipitation increases in one center the other shows reduced precipitation (see, e.g., Nogués-Paegle and Mo 1997; Herdies et al. 2002; Doyle and Barros 2002; Diaz and Aceituno 2003). The dipole structure manifests in many other fields including the convergence of moisture flux. The corresponding changes in the circulation are also noted in the SALLJ, which has lateral shifts toward one or the other dipole centers. The characteristics of the LLJ when supplying moisture to the southern center have been discussed in many articles (e.g., Berbery and Collini 2000; Saulo et al. 2000; Nicolini and Saulo 2006). In addition, the SALLJ has been the subject of a field campaign to further study its relation to precipitation, mostly due to mesoscale convective systems (Vera et al. 2006).

The large-scale dynamics plays a major role in the monsoon system, but the effect of surface conditions is also recognized. Betts and Viterbo (2005), employing the 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40), found that the total cloud and precipitation fields in the southwestern Amazon seem closely linked to the vertical velocity, implying that dynamic forcings are more relevant than the surface effects. Nevertheless, these authors also found a strong coupling between the top-layer soil moisture and relative humidity, in turn affecting the lifting condensation level (LCL), the cloud base, the depth of the boundary layer, and the net outgoing longwave radiation. Other studies have shown that the timing of the onset of the wet season in the Amazon also seems to respond, at least in part, to the surface conditions (Fu et al. 1999). Li and Fu (2004) showed that increases in the Bowen ratio during the dry season that precedes the rainy season delay the transitions in large-scale circulation that are typical of the onset stages. At later stages of the seasonal cycle the same effect becomes secondary.

Numerical experiments have shown that a better representation of the surface conditions can lead to improved forecasts and simulations (e.g., Betts et al. 1996; Beljaars et al. 1996). Xue et al. (2006) achieved a more realistic evolution of the South American monsoon through modifications to the Bowen ratio due to the explicit representation of the vegetation processes. Likewise, the experiments of de Goncalves et al. (2006) indicate that a realistic depiction of the initial soil moisture conditions reduces the root-mean-square error and increases the equitable threat score of 7-day forecasts of South American weather. The sensitivity of simulated precipitation to soil moisture anomalies may be dependent on the regional model’s domain size (Seth and Rojas 2003), but this should not be a problem in larger domains that allow interactions with the larger scales.

It is known that the effect of soil moisture on the surface fluxes and consequently on the Bowen ratio can lead to changes in precipitation (e.g., Eltahir 1998; Betts and Viterbo 2005). However, other questions can be raised in the case of South America: (i) Is the monsoon region a place where land–atmosphere feedbacks can be expected? (ii) How does the atmosphere respond to changes of soil moisture of opposite sign? (iii) Are there processes, apart of the Bowen ratio effect, that link soil moisture changes with precipitation? This study will address these questions by examining the land surface–atmosphere feedbacks that may affect the initial stages of the monsoon precipitation. To this end, numerical simulations with the regional Eta model are performed to examine the sensitivity of precipitation to soil moisture changes and the mechanisms involved.

The structure of the article is as follows: Section 2 presents the model and describes the experimental design and the methodology followed to perform the simulations. Section 3 discusses the background state and the control runs that are used as reference. Section 4 examines the overall effects of soil moisture changes on land–atmosphere interactions at the local scales. The impact of these changes in the low-level jet east of the Andes and the corresponding moisture flux convergence are addressed next in section 5. Finally, a summary and the main conclusions are presented in section 6.

2. Experimental design

The experiments performed for this research were done with the National Centers for Environmental Prediction (NCEP) workstation version of the Eta model, which is the same as the mesoscale regional model used in 2003 for NCEP’s operational short-term forecasts over the United States. The model configuration, described in Silva and Berbery (2006), includes a horizontal grid spacing of about 80 km with a domain that covers all South America and portions of neighboring oceans (see Fig. 1). Although 80 km is in the upper limit of the mesoscale processes, the aspects discussed in this article should not be affected by changes of resolution, as they are widespread over the monsoon region. As with any model with grid spacing of tens of kilometers or more, mesoscale convective systems’ precipitation is produced through parameterizations and not by their explicit representation. The vertical structure is resolved with 38 unevenly distributed levels, of which about half are below 700 hPa. Treatment of the lateral boundary conditions is discussed in Black (1994) and Mesinger (1997), and summarized in Berbery and Collini (2000).

The model physics includes the boundary layer formulation presented in Janjić (1990, 1994) and the convective scheme discussed in Betts and Miller (1986) with modifications by Janjić (1994). The land surface physics for surface energy/water fluxes and temperature/soil moisture are resolved by a land surface model (LSM) known as Noah. The Noah LSM is a model of intermediate complexity for use in operational weather and seasonal prediction models, and is documented in Ek et al. (2003). Further description of the convection scheme, surface processes, and radiation scheme can be found directly at NCEP’s Environmental Modeling Center Web site (http://www.emc.ncep.noaa.gov/mmb/mesoscale.html).

Versions of the Eta model are being employed with success at several national weather services and educational institutions in South America (e.g., Chou et al. 2002). At the University of Maryland, it is routinely used for research purposes (http://www.atmos.umd.edu/~berbery/etasam), and its performance is discussed in Berbery et al. (2004) and Silva and Berbery (2006).

The Eta Model experiments in this study are one-month simulations forced by the NCEP–National Center for Atmospheric Research (NCAR) global reanalysis (Kalnay et al. 1996) as initial and lateral boundary conditions. Although incompatibilities between the Eta model physics and the boundary conditions provided by the reanalyses could be expected, they have not substantially affected our integrations because of the large domain used in our experiments [see Warner et al. (1997) for a discussion on domain sizes using the Eta model]. The simulations employ prescribed sea surface temperature (SST) conditions and treat soil moisture anomalies as an initial value problem.

The simulations are performed for October months that represent the early stages of development of the monsoon. To cover different possible evolutions, the experiments are performed for (a) El Niño (1982); (b) La Niña (1983); and (c) neutral years (1981, 1999). In addition to the control experiments, sensitivity runs were performed to assess the soil moisture effect by superimposing spatially uniform soil moisture anomalies of different magnitude (±15%, ±30%, and ±45%). The results are presented in the form of ensembles, so that the reduced soil moisture ensemble and the increased soil moisture ensemble have 12 members each (3 soil moisture anomaly values and 4 October months). According to Kumar et al. (2001), ensembles of 10–20 members are sufficient to ensure average skill close to what is expected based on infinite ensemble size. When anomalies are shown, they are computed as the difference between the ensemble of reduced soil moisture anomaly simulations and the ensemble control run. Modifications to other parameters, like albedo, are not included although they can be affected by soil moisture changes. [See, e.g., Garratt (1993) and Eltahir (1998) for discussions on soil moisture and albedo relations.]

The choice of uniform anomalies is to simplify the analysis of the resulting effects, but the assumption is not expected to influence the results: as will be discussed in section 3a, soil moisture only interacts with the atmosphere in preferred regions (Koster et al. 2000); thus the presence of an anomaly over a noninteracting region will not have an impact on the overlying boundary layer. In other words, even when the soil moisture anomaly is uniformly distributed its impact will only be sensed in preferred regions like the monsoon region.

An EOF analysis of summer precipitation over South America (see Silva and Berbery 2006) reveals that the mode representing the monsoon [second mode, the first one being related to the intertropical convergence zone (ITCZ)] is centered approximately at 13°–17°S, 46°–52°W. In this article we define the monsoon region as the box bounded by 10°–20°S and 45°–55°W.

3. Time mean states

a. Background states

The stronger interactions between land surface conditions and the atmosphere occur in regions that are not too dry and not too wet (Koster et al. 2000) because moisture anomalies in very dry soils are rapidly evaporated before having enough time to act on the atmosphere, while with wet soils the evapotranspiration may be close to its potential value and thus changes in soil moisture will not necessarily imply changes in evapotranspiration. Delworth and Manabe (1988, 1989 proposed that the ratio between potential evaporation (PET) and precipitation (P) can be a helpful diagnostic to determine the regions where stronger land–atmosphere interactions are expected. Very large values of the ratio (P ≪ PET) reflect very dry regions, while small values (P ≫ PET) represent wet regions with large precipitation that saturates the soil. Values close to one and lower single digits are considered to represent the intermediate regions that are more favorable for land–atmosphere interactions.

Figure 2a shows a 24-yr climatology (1980–2003) of the austral springtime [September–November (SON)] ratio between potential evaporation and precipitation for South America as estimated from the NCEP–NCAR global reanalysis. The ratio PET/P is small over the Amazon region (due to abundant precipitation in comparison to potential evaporation), while large values of the ratio are found to the south (toward Patagonia) due to the small precipitation. Over and near the Andes Mountains the ratio exhibits large gradients that are possibly due to the poor representation of the fields by the coarse-grid global reanalysis. The region approximately between 10°–30°S and 60°–45°W has intermediate values where feedback effects can be expected, and if so, they may contribute to improve the predictive skill of the precipitation (as in The GLACE Team 2004). The intermediate region encompasses the monsoon, thus supporting the hypothesis that soil moisture and the early monsoon precipitation may have positive feedbacks that will contribute to the precipitation processes. Notice that very wet regions, like the Amazon and southern Chile have large runoff (Fig. 2b), because as discussed by Delworth and Manabe (1988), this is the mechanism by which the system removes water when evapotranspiration is not large enough (evapotranspiration has reached the potential value) to avoid the unphysical case when an anomaly could grow indefinitely.

Figure 2 presents SON averages to reflect the conditions during the early spring season. Figure 3a depicts the late springtime [October–November–December (OND)] 1980–2003 mean satellite-estimated precipitation field once the monsoon is established [employing the Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP); Xie and Arkin 1997]. The largest precipitation is noticed over the Atlantic portion of the ITCZ. Over the continent, a band of precipitation extends from the western Amazon toward southeastern Brazil and over the Atlantic Ocean. This band represents the monsoon precipitation and the SACZ.

Soil moisture is not measured except at local scales, but spatial estimates can be derived from observations and land surface models. CPC developed a soil moisture dataset (Fan and Van den Dool 2004) using a “leaky” one-layer bucket model and with a 0.5° × 0.5° resolution. (Notice that the magnitude of soil moisture is model dependent; see, e.g., Schaake et al. 2004.) Figure 3b reveals that during springtime the estimated soil moisture is largest over the northwestern portion of the South American continent, followed by a secondary maximum over southern Brazil. Notice also the large gradients near the central Andes Mountains and Patagonia. Large gradients have been found in precipitation analysis with station data and supported by large vegetation gradients in satellite imagery (J. Nogués-Paegle 2007, personal communication). At this point it cannot be stated, beyond speculation, whether these features are realistic or fictitious, as is the case of the NCEP–NCAR global reanalysis.

b. Control runs

Following the experimental design presented in section 2, month-long simulations were performed for four Octobers to represent the basic states responding to interannual variability. The ensemble precipitation of the four control simulations (Fig. 4a) depicts all the features of the observed CMAP precipitation for the same months (Fig. 3c), including the ITCZ, the monsoon precipitation/SACZ, and the maximum over southern Chile. The simulations tend to produce an onshore maximum of precipitation along the monsoon–SACZ band that is not evident in CMAP. Notice also the additional maximum toward the east of Paraguay in the observations (Fig. 3c) and model simulations (Fig. 4a). This precipitation center constitutes, together with the monsoon precipitation, the dipole described in section 1 and discussed in earlier literature (e.g., Nogués-Paegle and Mo 1997; Doyle and Barros 2002; Diaz and Aceituno 2003; Silva and Berbery 2006).

The model top soil moisture layer (0–10 cm) in Fig. 4b also reveals a pattern whose spatial structure resembles the soil moisture estimated from observations. Largest values are found again over northwestern South America (Colombia) and southern Brazil, while the monsoon region (defined here as the box bounded by 10°–20°S and 45°–55°W) exhibits intermediate values. The Eta model–simulated precipitation does not have the large values and gradients over the mountains noticed in Fig. 2b, suggesting that the model may be able to treat better the processes near steep mountains.

4. Soil moisture reduction experiments

a. Effects at the surface

The sensitivity of the monsoon precipitation to changes in initial soil moisture conditions was explored using ensembles of month-long simulations with the regional mesoscale Eta model. As stated earlier, the individual simulations correspond to the four Octobers with ±15%, ±30%, and ±45% soil moisture anomalies. Thus, the ensembles for either reduced or increased initial soil moisture have 12 members each.

The reduced soil moisture ensemble shows an overall reduction of precipitation (Fig. 5a), reflecting a positive feedback between the two variables, particularly over the band that represents the monsoon. To simplify the analysis, land-only precipitation is shown. Only small areas depict slightly positive values. In contrast, Fig. 5b shows that increasing the initial soil moisture results in an undetermined pattern over the monsoon region with patches of increased and decreased precipitation. It is only away from this region, like northwestern South America and northeastern Argentina, where a clearer pattern of increased precipitation (positive feedback) is found. The dark gray areas in Fig. 5b show increased precipitation with increased soil moisture, indicating a positive feedback. The light shades indicate that there are areas with decreased precipitation resulting from increased soil moisture. The reason is likely in dynamic forcings that are acting together with the soil moisture change. For example, the positive feedback over Colombia implies stronger ascending motions that will need to be compensated for elsewhere. If the compensating motions occur to the south, then the dynamic effect may produce descending (or reduced ascending motions) that offset the local positive effect of the soil moisture, and a negative anomaly in precipitation results. The results of Fig. 5 suggest that the soil in the monsoon region, although not too wet, may rapidly reach saturation. Therefore, adding water to the soil does not greatly affect the atmosphere and precipitation processes, unlike the opposite case, which indicates that when reducing the soil moisture there is an effective change in the soil conditions that is conducive to less precipitation.

To test how systematic the results are, Fig. 6 presents the area average precipitation over the monsoon region (box 45°–55°W, 10°–20°S) for the individual experiments. The reduction of soil moisture produces an almost linear reduction of precipitation, while the increases of soil moisture have a negligible effect that does not depend on the magnitude of the soil moisture anomaly. These results are mostly independent of the year considered regardless if it is an El Niño, La Niña, or neutral year.

Following the results associated with Figs. 5 and 6, most of the subsequent discussion will focus on the reduced soil moisture experiments. Although the anomalies of soil moisture are introduced in the initial conditions, their effects last the full integration. This can be seen in Fig. 7, where the monsoon area-averaged time series of the ensemble latent and sensible heat fluxes are displayed. The reduction of the initial soil moisture results in a larger sensible heat flux during the whole month (Fig. 7a). In general there is an increase of the sensible heat flux of about 10–20 W m−2, and there also seems to be a slight increase with time in the difference between the control and reduced soil moisture experiments. The latent heat flux of the reduced soil moisture ensemble (Fig. 7c) is smaller than the control ensemble at all times of the integration (implying less evapotranspiration). Their difference, which is about 10–30 W m−2, also tends to increase with time. The changes in the two heat fluxes have associated changes in the Bowen ratio (Fig. 7e), which increases by about one unit with respect to the control ensemble.

Although the changes are systematic throughout the time integration, most of the effects take place during daytime when the heat fluxes achieve a maximum (Figs. 7b,d). The timing of the maximum (1800 UTC or 1500 LST) is consistent with other studies (Hahmann and Dickinson 1997; Betts et al. 2002); however, the magnitudes differ largely even between observations, as they are highly dependent of the type of surface conditions (e.g., vegetation types) dominant at the measuring sites.

According to Fig. 7b, the sensible heat flux of the reduced soil moisture ensemble is about 50 W m−2 larger than that of the control ensemble during the middle of the day. At nighttime, both control and reduced soil moisture ensembles have a reversal of the sensible heat flux sign, with the reduced soil moisture experiment having slightly larger negative values. The latent heat flux difference between the control and reduced soil moisture ensemble is also largest during daytime (Fig. 7d) and of about the same magnitude as that of the sensible heat flux (∼50 W m−2). Notice that in this case the values are always smaller than the control run, even during nighttime. In the case of the Bowen ratio (Fig. 7f) both daytime latent and sensible heat fluxes lead to a larger ratio, but during nighttime the sign reverses, with more negative values than the control ensemble. This latter effect is likely the result of the very small changes in the sensible heat flux during nighttime.

The changes in soil moisture also affect the surface radiation budget (Fig. 8). The net incoming shortwave radiation is defined as SWnet = SW↓ − SW↑ where the arrows represent the direction of the flux, and positive values indicate a net gain of shortwave radiation. The net outgoing longwave radiation is defined as LWnet = LW↓ − LW↑, and in this case, negative values reflect the loss of energy. The total radiation is then defined as NR = SWnet + LWnet. Figure 8a shows that the reduction of soil moisture in the initial conditions results in a systematic increase of the net outgoing longwave radiation (negative values), probably due to the reduction of low cloud coverage (Betts and Viterbo 2005) and warmer surface. Although the increase of outgoing longwave radiation flux is seen throughout the day, the largest differences between the control and reduced soil moisture ensembles are found in the midafternoon (Fig. 8b).

The reduced soil moisture effect on the incoming shortwave radiation flux is less evident, but according to Fig. 8c, there is a slight increase (the lower cloud coverage would decrease the overall albedo of the region). In this case, the nighttime and morning hours show no difference (Fig. 8d), while the larger difference takes place during the afternoon and evening, between 1500 and 2100 UTC, or approximately 1200 and 1800 LST. The net shortwave and longwave radiation flux terms tend to cancel each other in the total radiation budget (Fig. 8e), with a net reduction of the order of about −10 W m−2. The reduction of total radiation is distributed throughout the day (Fig. 8f), except for a slight increase occurring around 1800 UTC (1500 LST). This is better seen in the second curve in the figure that represents the difference between control and soil moisture experiments.

b. Boundary layer effects

The changes in the surface energy and radiation fluxes are expected to alter the structure of the boundary layer (BL). The lower evapotranspiration (latent heat flux) in the reduced soil moisture experiments implies less water into the BL, particularly below the 700-hPa level, as seen in Fig. 9a. On the other hand, the increased sensible heat flux and mixing of the lower levels leads to higher temperatures with respect to the control ensemble for the layer from the surface up to 700–650 hPa. The increased soil moisture ensemble, while having the opposite behavior, has deviations of a much smaller magnitude and on a shallower layer.

It is also of interest to examine the amplitude of the diurnal cycle in the BL. The amplitude is defined here as the difference between the time of maximum temperature (1800 UTC or ∼1500 LST) and the time of minimum temperature (0900 or ∼0600 LST). According to Fig. 9b, the control ensemble has relatively small amplitude in the temperature’s diurnal cycle near the surface (about 5°C), probably reflecting the tropical/subtropical nature of the region. The amplitude diminishes at higher levels and remains at about 0.5° above 700 hPa. The drier atmosphere of the reduced soil moisture ensemble increases the amplitude of the temperature’s diurnal cycle (Fig. 9b), particularly in the first 100 hPa, with hints of smaller amplitude above 700 hPa. The largest amplitude of the diurnal cycle of the dewpoint temperature is not found at the lower levels, but around 800 hPa (Fig. 9c), likely the result of vertical mixing. Notice that the negative values in the lower layers indicate a nighttime accumulation of moisture probably because of the lack of vertical mixing, while the positive values at the upper levels reflect the increased mixing during daytime. The diurnal amplitude of the atmospheric moisture content is somewhat reduced while the maximum is at a slightly higher elevation (Fig. 9c), probably because the increase of sensible heat flux leads to increased vertical mixing in the deeper BL.

The changes in surface conditions and BL structure are expected to affect the instability of the atmosphere. The development of convective clouds and precipitation is related to the convective available potential energy (CAPE), which is the amount of energy available to an air parcel to ascend once it has reached the level of free convection. When CAPE is large, the atmosphere will be more unstable. However, convection depends also on the convective inhibition (CIN), that is, the amount of energy needed to supply an air parcel to elevate it up to the level of free convection. Large values of CIN imply large resistance to convective development. Therefore, the more favorable conditions for convection and precipitation are identified by large values of CAPE and small values of CIN. Figure 10a shows that the reduced soil moisture ensemble, with its drier and warmer BL, reduces the CAPE with a systematic decrease of about 30% throughout the month. [Silva and Berbery (2006) found little relation between the monsoon precipitation and CAPE, but their composites cover the austral summer months when the monsoon is already established and the soil is too wet for surface effects to influence the BL structure.] In the same way as for the sensible and latent heat flux, a trend toward larger differences in time is noticed. The overall CAPE values are relatively small due to removal of the diurnal variations in the time series and the area averaging over a large region. The diurnal cycle has larger amplitude (Fig. 10b), but still, these curves are not categorized in rain/no-rain days, which probably would better reflect realistic values of CAPE. The timing of the diurnal cycle, with a peak at 1200 UTC (about 0900 LST), seems premature and may reflect an early mixing of the BL in the Eta model. Difficulties in representing the diurnal cycle have been found in other models: for example, Betts and Jakob (2002) found that the ECMWF forecast model over Rondônia, Brazil, produces precipitation too early in the day, which was attributed to an improper representation of the development of the shallow cumulus layer. Betts and Jakob (2002) also notice that despite the failure to reproduce the diurnal cycle of precipitation, the model mean daily precipitation compares well with observations. This has been observed in other models as well. Although CAPE differences between control and reduced soil moisture ensembles are present at all hours, the largest differences occur at daytime.

The time series of area-averaged CIN in the control ensemble (Fig. 10c) stays in the range 25–50 W m−2. CIN has a well-defined diurnal cycle (Fig. 10d) with negative values as large as 65 W m−2 of magnitude near dawn, and as small as 10 W m−2 in magnitude in the afternoon (∼1800 LST). Unlike CAPE, this diurnal cycle is in agreement with the evolution of the surface variables and the BL evolution. The reduced soil moisture experiments lessen the amplitude of CIN’s diurnal cycle, with somewhat larger negative values during the afternoon, implying a slight increase in the resistance to convection development.

The last two panels of Fig. 10 present the evolution of the precipitation in the control and reduced soil moisture ensembles. The control ensemble has the largest precipitation at 2100 UTC (∼1800 LST), and a nighttime to early morning minimum. The precipitation maximum occurs about 3 h after the maximum in surface fluxes and is in phase with the diurnal cycle of CIN. Given the known difference in CAPE’s diurnal cycle, it can be speculated then that the reduction in CIN may be more relevant than the increase in CAPE to determine the diurnal cycle of precipitation. The day-to-day variability in the reduced soil moisture ensemble is similar to that of the control ensemble, albeit with a general reduction in the magnitude. The precipitation diurnal cycle in the reduced soil moisture ensembles has the same phase as the control ensemble except that the reduction in precipitation, although distributed throughout the day, has the largest values at 2100 UTC (∼1800 LST).

5. The low-level jet and the moisture flux

The reduction in precipitation is not necessarily the only consequence of changes in the atmospheric instability discussed in the previous section. In fact, the changes in the BL also alter the moisture transport into the monsoon region. This section will inspect the moisture flux characteristics and the changes that result from modifying the initial conditions of soil moisture. The vertically integrated moisture flux of the control ensemble (Fig. 11a) shows similar features as previous climatologies based on global reanalyses (Labraga et al. 2000; Berbery and Barros 2002; Doyle and Barros 2002; Marengo et al. 2004) and short-term regional forecasts (Silva and Berbery 2006). The largest westward moisture transport is found over tropical regions supplying moisture from the tropical Atlantic Ocean to the Amazon basin. The moisture flux exhibits a counterclockwise rotation turning south near the Andes Mountains and then east into the monsoon region. The second maximum in moisture flux east of the Andes reflects the SALLJ that supplies moisture to the monsoon system. Figure 11a also shows that the SALLJ is not the only source of moisture for the monsoon as some contributions come directly from the north, consistent with the results of Barros et al. (2002).

For convenience, the vertically integrated moisture flux convergence is defined as positive, and the divergence as negative. This convergence, together with the precipitation and the evapotranspiration, constitutes one of the three most relevant terms of the atmospheric water cycle in relation to the monsoon. Figure 11b shows that the vertically integrated moisture flux convergence has a maximum downstream of the SALLJ over the central and southern parts of Brazil, where the maximum precipitation was found (Fig. 4a). Other convergence maxima, like the one corresponding to the ITCZ, are also noticeable. The northeastern part of Brazil, which does not have precipitation, has consistently a large divergence of moisture flux (negative values). The moisture flux convergence field near mountains exhibits a distorted pattern typical of all models and their difficulties to resolve the circulation near the steep Andes Mountains (e.g., Nogués-Paegle et al. 2006).

The reduction of soil moisture in the initial conditions leads to changes of the moisture flux pattern that are better represented in the anomalies depicted in Fig. 11c. The westward moisture flux anomalies indicate a reduction of the total moisture flux into the monsoon region. Consequently, this region has negative anomalies of moisture flux convergence (Fig. 11d), resulting in a reduction of the total moisture flux convergence in the region. On the other hand, nearby areas to the north and south have positive anomalies. Notice in particular that the reduction of moisture flux convergence over the monsoon and the increase to the south are consistent with the dipolar mode discussed in section 1. The patterns in Figs. 11c,d have a resemblance to one phase of the dipole pattern discussed in studies like Nogués-Paegle and Mo (1997) and Herdies et al. (2002) and presented in the introduction. The reason is that our experiments with reduced soil moisture anomalies in fact can result in excitation of the dipolar mode.

The changes in the moisture flux and its convergence over the monsoon can be explained with the analysis of the SALLJ structure. Although the SALLJ is not the only source of moisture (direct flux from the north is also noticed), it best exemplifies a second effect of soil moisture through changes in the vertical structure of the wind and atmospheric moisture content. Figure 12a presents the control ensemble wind cross section along the line AB in Fig. 11a. The SALLJ is found between 900 and 600 hPa, with the maximum at about 750 hPa (the negative values indicate a southward component). The specific humidity exhibits a strong vertical stratification with largest values below 700 hPa (not shown) so that the maximum moisture flux occurs at about 800 hPa (Fig. 12c). Figure 12a also shows a second wind maximum at about 300 hPa (direction not shown, but it represents the upper-level westerlies), which is not relevant for the moisture transports because of the negligible specific humidity at those levels.

When the initial soil moisture is reduced in the simulations, changes in the SALLJ structure are noticed as well: according to Fig. 12b, the wind anomalies show an increased southeastward component at higher levels (negative anomalies at 600–500 hPa), while a hint of decreased wind intensity is noticed near the surface. This pattern suggests a higher and stronger SALLJ, which is consistent with the drier BL and a more elevated BL top. However, as seen in Fig. 12d, because of the stratification of the specific humidity and the reduction in atmospheric moisture content, the southeastward moisture flux is noticeably reduced (positive anomalies) below 750 hPa and particularly between 800 and 900 hPa. At the same time the upper levels have a slight increase of moisture flux as indicated by the negative anomalies centered at 650 hPa, which would not be enough to compensate the lower levels’ decrease in moisture transport.

Earlier studies—many addressing the U.S. Great Plains floods of 1993—have discussed possible roles of soil moisture in the intensity of an LLJ and the resulting effect on downstream precipitation. In some cases, these studies have offered a divergent view of the likely soil moisture effects (see Paegle et al. 1996 and a summary in Bosilovich and Sun 1999). The differences are attributed to dissimilar interpretations of the dominant processes that are affected by the surface conditions, but they are also probably due to different model performance, different model configuration including domain size, initial soil moisture conditions, and even the position of the soil moisture anomaly with respect to the LLJ location. Paegle et al. (1996) concluded that surface evapotranspiration may be more favorable to enhance the buoyancy rather than in providing additional moisture to the already abundant supply of the LLJ. While finding a similar result with respect to the LLJ strength and amplitude of the diurnal cycle, Zhong et al. (1996) also found stronger ascending motions downstream of the LLJ, thus favoring more precipitation with wet soils in contrast to opposite results by McCorcle (1988). In this case, Zhong et al. attributed the differences to the lack of inclusion of relevant processes in McCorcle’s model.

The results reported in our research indicate that when soil moisture is reduced the SALLJ is stronger in terms of wind speed, but the more elevated LLJ in combination with a reduction of the BL moisture content results in a reduction of the advected moisture into the monsoon region. Thus, the changes in the LLJ and the reduction of moisture transports into the monsoon region constitute a second mechanism by which precipitation is reduced when the initial soil moisture is reduced.

6. Summary and conclusions

This study has examined the feedbacks between soil moisture and precipitation during the early stages of the South American monsoon. With this purpose, a sensitivity analysis to soil moisture initial conditions was performed with the regional Eta model. The month-long simulations were carried out for different October months in order to ensure that the results are not dependent on a particular year. The choice of October corresponds with the early austral spring when surface conditions over the monsoon region are shown to favor the land–atmosphere interactions. Analysis of the individual runs indicates that the surface and dynamical forcings act in the monsoon region independently of the large-scale conditions (El Niño, La Niña, and neutral years), thus suggesting that the results are robust. This is in agreement with Fu et al. (1999) who found that El Niño or non–El Niño events do not change the processes that control the onset of the wet season over the Amazon.

The control simulations bear a consistent resemblance with satellite-based observations and global reanalysis-based climatologies. The ensemble precipitation field exhibits all the major features of the observations, both over ocean and land, and particularly the monsoon precipitation. Likewise, the simulations reproduce the main features of the moisture flux fields as represented by global reanalysis, including the inflow from the Atlantic Ocean and the low-level jet east of the Andes Mountains. However, the higher resolution of the model simulations represents the narrow SALLJ better than the global reanalysis.

The sensitivity experiments to soil moisture changes reveal that while a reduction of soil moisture produces an almost linear reduction in precipitation over the monsoon region, an equivalent soil moisture increase has only a negligible effect, suggesting that the surface rapidly reaches large values of soil moisture where evapotranspiration is not significantly modified. The imposed soil moisture anomalies are introduced as initial conditions, but their effects persist for the entire length of the simulation, as noticed in all the analyzed variables. The reduction of soil moisture is accompanied by an increase of the sensible heat flux as well as a reduction of the latent heat flux (less evapotranspiration) and consequently an increase of the Bowen ratio. The surface effects also modify the radiation budget probably through cloud cover changes, so that reduced soil moisture implies a larger loss of longwave radiation and a minor gain in shortwave radiation. While the longwave and shortwave radiation terms tend to cancel each other, the total radiation flux still depicts a reduction in magnitude.

Reduction of the soil moisture gives rise to changes in the boundary layer structure and thermodynamic stability: the increased sensible heat flux favors mixing and a deeper BL, while the reduction of latent heat flux (evapotranspiration) results in a drier BL. The changes in all variables are most apparent during daytime. The sensible and latent heat fluxes have differences as large as ∼50 W m−2 about 1500 LST, but decay to nearly zero during nighttime (and the fluxes themselves have near-zero values). These changes affect the convective instability (CAPE and CIN): CAPE is reduced by about 30%, but its diurnal cycle, with a maximum in the early morning, disagrees with either the expected evolution or the model precipitation. CIN, on the other hand, has a small increase (resistance to convection) that is more evident at 1800 LST, which is the time when precipitation shows the largest reduction with respect to the control ensemble.

The previous analysis discusses the local effects that soil moisture anomalies have on the overlying atmosphere. However, the changes in the boundary layer also result in other regional changes through the modification of the moisture transports. The reduced soil moisture experiments produce a deeper boundary layer, which in turn results in a stronger but more elevated LLJ. Given the lesser moisture content at higher altitude and the overall reduction of moisture in the BL due to the smaller evapotranspiration, there is a consequent reduction in the LLJ moisture transports and in the moisture flux convergence and, hence, in the core monsoon precipitation.

In summary, two mechanisms are acting as positive feedbacks when the initial soil moisture is reduced, and both concurrently favor the reduction of precipitation. On the one hand the direct effect of locally reduced evapotranspiration, and reduced instability, and on the other hand the regional reduction in the moisture flux convergence due to a higher LLJ and less moisture content in the lower layers.

Acknowledgments

The comments of Professors Julia Nogués-Paegle and Jan Paegle are much appreciated. Comments by the anonymous reviewers helped clarify several aspects of the manuscript. This research was supported by NOAA Grant NA76GP0479 (PACS) and NSF Grant EAR0450089.

REFERENCES

  • Barros, V., , M. Doyle, , M. González, , I. Camilloni, , R. Bejaran, , and R. M. Caffera, 2002: Climate variability over subtropical South America and the South American monsoon: A review. Meteorologica, 27 , 3357.

    • Search Google Scholar
    • Export Citation
  • Beljaars, A. C. M., , P. Viterbo, , M. J. Miller, , and A. K. Betts, 1996: The anomalous rainfall over the United States during July 1993: Sensitivity to land surface parameterization and soil moisture anomalies. Mon. Wea. Rev., 124 , 362383.

    • Search Google Scholar
    • Export Citation
  • Berbery, E. H., , and E. A. Collini, 2000: Springtime precipitation and water vapor flux over southeastern South America. Mon. Wea. Rev., 128 , 13281346.

    • Search Google Scholar
    • Export Citation
  • Berbery, E. H., , and V. R. Barros, 2002: The hydrologic cycle of the La Plata basin in South America. J. Hydrometeor., 3 , 630645.

  • Berbery, E. H., , E. A. Collini, , V. Silva, , M. Peña, , D. L. Lettenmaier, , V. R. Barros, , and M. Pyle, 2004: The water cycle of La Plata basin at regional scales. GEWEX News, Vol. 14, No. 1, International GEWEX Project Office, Silver Spring, MD, 4–5.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , and M. J. Miller, 1986: A new convective adjustment scheme. Part II: Single column tests using GATE wave, BOMEX, ATEX and Arctic air-mass data sets. Quart. J. Roy. Meteor. Soc., 112 , 693710.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , and C. Jakob, 2002: Evaluation of the diurnal cycle of precipitation, surface thermodynamics, and surface fluxes in the ECMWF model using LBA data. J. Geophys. Res., 107 .8045, doi:10.1029/2001JD000427.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , and P. Viterbo, 2005: Land-surface, boundary layer, and cloud-field coupling over the southwestern Amazon in ERA-40. J. Geophys. Res., 110 .D14108, doi:10.1029/2004JD005702.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , J. H. Ball, , A. C. M. Beljaars, , M. J. Miller, , and P. Viterbo, 1996: The land surface–atmosphere interaction: A review based on observational and global modeling perspectives. J. Geophys. Res., 101 , D3. 72097226.

    • Search Google Scholar
    • Export Citation
  • Betts, A. K., , J. D. Fuentes, , M. Garstang, , and J. H. Ball, 2002: Surface diurnal cycle and boundary layer structure over Rondônia during the rainy season. J. Geophys. Res., 107 .8065, doi:10.1029/2001JD000356.

    • Search Google Scholar
    • Export Citation
  • Black, T. L., 1994: The new NMC mesoscale Eta Model: Description and forecast examples. Wea. Forecasting, 9 , 265278.

  • Bosilovich, M. G., , and W-Y. Sun, 1999: Numerical simulation of the 1993 of the midwestern flood: Land–atmosphere interactions. J. Climate, 12 , 14901505.

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

    • Search Google Scholar
    • Export Citation
  • Chou, S. C., , C. A. S. Tanajura, , Y. Xue, , and C. A. Nobre, 2002: Validation of the coupled Eta/SSiB model over South America. J. Geophys. Res., 107 .8088, doi:10.1029/2000JD000270.

    • Search Google Scholar
    • Export Citation
  • de Goncalves, L. G. G., , W. J. Shuttleworth, , S. C. Chou, , Y. Xue, , P. R. Houser, , D. L. Toll, , J. Marengo, , and M. Rodell, 2006: Impact of different initial soil moisture fields on Eta model weather forecasts for South America. J. Geophys. Res., 111 .D17102, doi:10.1029/2005JD006309.

    • Search Google Scholar
    • Export Citation
  • Delworth, T., , and S. Manabe, 1988: The influence of potential evaporation on the variabilities of simulated soil wetness and climate. J. Climate, 1 , 523547.

    • Search Google Scholar
    • Export Citation
  • Delworth, T., , and S. Manabe, 1989: The influence of soil wetness on near-surface atmospheric variability. J. Climate, 2 , 14471462.

  • Diaz, A., , and P. Aceituno, 2003: Atmospheric circulation anomalies during episodes of enhanced and reduced convective cloudiness over Uruguay. J. Climate, 16 , 31713185.

    • 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
  • Ek, M. B., , K. E. Mitchell, , Y. Lin, , P. Grunmann, , E. Rogers, , G. Gayno, , V. Koren, , and J. D. Tarpley, 2003: Implementation of Noah land surface model advances in the National Centers for Environmental Prediction operational mesoscale Eta Model. J. Geophys. Res., 108 .8851, doi:10.1029/2002JD003296.

    • Search Google Scholar
    • Export Citation
  • Eltahir, E. A. B., 1998: A soil moisture–rainfall feedback mechanism. 1. Theory and observations. Water Resour. Res., 34 , 765776.

  • Fan, Y., , and H. Van den Dool, 2004: Climate Prediction Center global monthly soil moisture data set at 0.5° resolution for 1948 to present. J. Geophys. Res., 109 .D10102, doi:10.1029/2003JD004345.

    • 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
  • Garratt, J. R., 1993: Sensitivity of climate simulations to land-surface and atmospheric boundary-layer treatments—A review. J. Climate, 6 , 419448.

    • Search Google Scholar
    • Export Citation
  • GLACE Team, 2004: Regions of strong coupling between soil moisture and precipitation. Science, 305 , 11381140.

  • Hahmann, A. N., , and R. E. Dickinson, 1997: RCCM2-BATS model over tropical South America: Applications to tropical deforestation. J. Climate, 10 , 19441964.

    • Search Google Scholar
    • Export Citation
  • Herdies, D. L., , A. da Silva, , M. A. F. Silva Dias, , and R. Nieto Ferreira, 2002: Moisture budget of the bimodal pattern of the summer circulation over South America. J. Geophys. Res., 107 .8075, doi:10.1029/2001JD000997.

    • Search Google Scholar
    • Export Citation
  • Horel, J. D., , A. N. Hahmann, , and J. Geisler, 1989: An investigation of the annual cycle of convective activity over the tropical Americas. J. Climate, 2 , 13881403.

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

  • Kodama, Y. M., 1992: Large-scale common features of the subtropical precipitation zones (the Baiu frontal zone, the SPCZ, and the SACZ). Part I: Characteristics of subtropical frontal zones. J. Meteor. Soc. Japan, 70 , 813836.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., , M. J. Suarez, , and M. Heiser, 2000: Variance and predictability of precipitation at seasonal-to-interannual timescales. J. Hydrometeor., 1 , 2646.

    • Search Google Scholar
    • Export Citation
  • Kumar, A., , A. G. Barnston, , and M. P. Hoerling, 2001: Seasonal predictions, probabilistic verifications, and ensemble size. J. Climate, 14 , 16711676.

    • Search Google Scholar
    • Export Citation
  • Janjić, Z. I., 1990: The step-mountain coordinate: Physical package. Mon. Wea. Rev., 118 , 14291443.

  • Janjić, Z. I., 1994: The step-mountain eta coordinate model: Further developments of the convection, viscous sublayer, and turbulence closure schemes. Mon. Wea. Rev., 122 , 927945.

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

  • Labraga, J. C., , O. Frumento, , and M. López, 2000: The atmospheric water vapor cycle in South America and the tropospheric circulation. J. Climate, 13 , 18991915.

    • Search Google Scholar
    • Export Citation
  • Lenters, J. D., , and K. H. Cook, 1997: On the origin of the Bolivian high and related circulation features of the South American climate. J. Atmos. Sci., 54 , 656678.

    • 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 Re-Analysis. J. Climate, 17 , 26372651.

    • 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
  • Liebmann, B., , G. N. Kiladis, , C. S. Vera, , A. C. Saulo, , and L. M. V. Carvalho, 2004: Subseasonal variations of rainfall in South America in the vicinity of the low-level jet east of the Andes and comparison to those in the South Atlantic convergence zone. J. Climate, 17 , 38293842.

    • Search Google Scholar
    • Export Citation
  • Marengo, J. A., , 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
  • McCorcle, M. D., 1988: Simulation of surface-moisture effects on the Great Plains low-level jet. Mon. Wea. Rev., 116 , 17051720.

  • Mesinger, F., 1997: Dynamics of limited-area models: Formulation and numerical methods. Meteor. Atmos. Phys., 63 , 314.

  • Nicolini, M., , and A. C. Saulo, 2006: Modeled Chaco low-level jets and related precipitation patterns during the 1997–1998 warm season. Meteor. Atmos. Phys., 94 , 129143.

    • Search Google Scholar
    • Export Citation
  • Nogués-Paegle, J., , and K. C. 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., and Coauthors, 2002: Progress in Pan American CLIVAR research: Understanding the South American Monsoon. Meteorologica, 27 , 332.

    • Search Google Scholar
    • Export Citation
  • Nogués-Paegle, J., , E. H. Berbery, , E. A. Collini, , and C. Saulo, 2006: An evaluation of the Eta model during SALLJEX. Proc. Eighth Int. Conf. on Southern Hemisphere Meteorology and Oceanography, Foz do Iguaçu, Brazil, INPE, 51–56.

  • Paegle, J., , K. C. Mo, , and J. Nogués-Paegle, 1996: Dependence of simulated precipitation on surface evaporation during the 1993 United States summer floods. Mon. Wea. Rev., 124 , 345361.

    • Search Google Scholar
    • Export Citation
  • Rodwell, M. J., , and B. J. Hoskins, 2001: Subtropical anticyclones and summer monsoons. J. Climate, 14 , 31923211.

  • Satyamurty, P., , C. C. Ferreira, , and M. A. Gan, 1990: Cyclonic vortices over South America. Tellus, 42A , 194201.

  • Saulo, A. C., , M. Nicolini, , and S. C. Chou, 2000: Model characterization of the South American low-level flow during the 1997–1998 spring-summer season. Climate Dyn., 16 , 867881.

    • Search Google Scholar
    • Export Citation
  • Schaake, J. C., and Coauthors, 2004: An intercomparison of soil moisture fields in the North American Land Data Assimilation System (NLDAS). J. Geophys. Res., 109 .D01S90, doi:10.1029/2002JD003309.

    • Search Google Scholar
    • Export Citation
  • Seth, A., , and M. Rojas, 2003: Simulation and sensitivity in a nested modeling system for South America. Part I: Reanalyses boundary forcing. J. Climate, 16 , 24372453.

    • Search Google Scholar
    • Export Citation
  • Silva, V. B. S., , and E. H. Berbery, 2006: Intense rainfall events affecting the La Plata basin. J. Hydrometeor., 7 , 769787.

  • Vera, C., and Coauthors, 2006: The South American Low-Level Jet Experiment. Bull. Amer. Meteor. Soc., 87 , 6377.

  • Warner, T. T., , R. A. Peterson, , and R. E. Treadon, 1997: A tutorial on lateral boundary conditions as a basic and potentially serious limitation to regional numerical weather prediction. Bull. Amer. Meteor. Soc., 78 , 25992617.

    • Search Google Scholar
    • Export Citation
  • Xie, P., , and P. A. Arkin, 1997: Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor. Soc., 78 , 25392558.

    • Search Google Scholar
    • Export Citation
  • Xue, Y., , F. de Sales, , W-P. Li, , C. R. Mechoso, , C. A. Nobre, , and H-M. Juang, 2006: Role of land surface processes in South American monsoon development. J. Climate, 19 , 741762.

    • Search Google Scholar
    • Export Citation
  • Zhong, S., , J. D. Fast, , and X. Bian, 1996: A case study of the Great Plains low-level jet using wind profiler network data and a high-resolution mesoscale model. Mon. Wea. Rev., 124 , 785806.

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

Fig. 1.
Fig. 1.

The regional Eta model domain (shaded area) and the analysis domain (rectangle).

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Fig. 2.
Fig. 2.

(a) Ratio (dimensionless) between potential evaporation and precipitation during SON. The shaded region represents intermediate values of magnitude 1–4. (b) Surface runoff (mm day−1), also for SON. All terms are 24-yr averages (1980–2003) computed from the NCEP–NCAR Reanalysis Project (NNRP) analyses. The square box in the panels represents the monsoon region.

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Fig. 3.
Fig. 3.

(a) 1980–2003 mean OND precipitation (mm day−1) estimated from CMAP; (b) CPC 1980–2003 mean OND soil moisture (mm); (c) CMAP precipitation averaged for October 1981, 1982, 1983, and 1989 for comparison to the model simulations. The square box in the panels represents the monsoon region. See text for further explanation and references.

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Fig. 4.
Fig. 4.

(a) October precipitation and (b) top-layer soil moisture computed from the control ensemble of the Eta model simulations. Units of precipitation are mm day−1 with values larger than 5 mm day−1 shaded. Soil moisture units are m3 m−3 with values larger than 0.25 shaded. The square box in the panels represents the monsoon region.

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Fig. 5.
Fig. 5.

Land-only precipitation anomalies for (a) the ensemble of reduced initial soil moisture simulations and (b) the ensemble of increased initial soil moisture simulations. Units are mm day−1 and values larger than +/− 0.2 mm day−1 are shaded dark/light. The square box in the panels represents the monsoon region.

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Fig. 6.
Fig. 6.

Area average (20°–10°S, 55°–45°W) of the monthly precipitation changes (%) as a function of the initial soil moisture changes (%) for the individual simulations that represent the ensembles in this study.

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Fig. 7.
Fig. 7.

Monsoon area-averaged time series and mean diurnal cycle of (a), (b) sensible heat flux, (c), (d) latent heat flux, and (e), (f) Bowen ratio. Units of the sensible and latent heat fluxes are W m−2. The Bowen ratio is dimensionless. The solid line is the control ensemble, and the dashed line is the reduced soil moisture ensemble.

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Fig. 8.
Fig. 8.

As in Fig. 7 but for (a), (b) the outgoing longwave radiation, (c), (d) incoming shortwave radiation, and (e), (f) total radiation. The dotted line in (f) is the net radiation difference between the control and reduced soil moisture ensembles, with the scale on the outer left side. Units are W m−2.

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Fig. 9.
Fig. 9.

(a) Temperature and dewpoint distribution in the boundary layer for the control (solid/dashed lines), and reduced (open circles) and increased (closed circles) soil moisture ensembles. The inset represents thermodynamic structure of the control ensemble in the troposphere up to 100 hPa. (b) Amplitude of the diurnal cycle of temperature for the control (solid line) and reduced soil moisture (dashed line) ensembles. (c) As in (b) but for the dewpoint. All units are °C.

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Fig. 10.
Fig. 10.

As in Fig. 7 but for (a), (b) CAPE; (c), (d) CIN; and (e), (f) precipitation. Units of CAPE and CIN are W m−2. Units of precipitation are mm day−1 for (e) and mm (3 h)−1 for (f).

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Fig. 11.
Fig. 11.

(a) Vertically integrated moisture flux and (b) its convergence for the control ensemble; (c), (d) the corresponding anomalies for the reduced soil moisture ensemble. The vertically integrated moisture flux has units of kg m−1 s−1 and its convergence has units of mm day−1. Moisture flux convergence is defined as positive and the divergence as negative. In (b), positive values (convergence) are shaded. In (d), negative values (divergence anomalies, or reduction of convergence) are shaded. The square box in the panels represents the monsoon region.

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Fig. 12.
Fig. 12.

Cross sections along line AB in Fig. 11 of (a) winds of the control ensemble (m s−1); (b) wind anomalies (w.r.t. the control ensemble) of the reduced soil moisture case (m s−1); (c) moisture flux of the control ensemble (g kg−1 m s−1); and (d) anomalies of moisture flux for the reduced soil moisture ensemble (g kg−1 m s−1).

Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1846.1

Save