• Benton, G. S., and M. A. Estoque, 1954: Water-vapor transfer over the North American continent. J. Meteor.,11, 462–477.

  • Blackadar, A. K., 1957: Boundary-layer wind maxima and their significance for the growth of nocturnal inversions. Bull. Amer. Meteor. Soc.,38, 283–290.

  • Blackmon, M. L., V.-H. Lee, and J. M. Wallace, 1984: Horizontal structure of 500 mb height fluctuations with long, intermediate and short time scales. J. Atmos. Sci.,41, 961–979.

  • Bloom, S. C., L. L. Takacs, A. M. da Silva, and D. Ledvina, 1996: Data assimilation using incremental analysis updates. Mon. Wea. Rev.,124, 1256–1271.

  • Bonner, W. D., 1966: Case study of thunderstorm activity relation to the low-level jet. Mon. Wea. Rev.,94, 167–178.

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

  • ——, and J. Paegle, 1970: Diurnal variations in the boundary-layer winds over the south central United States in summer. Mon. Wea. Rev.,98, 735–744.

  • Bowen, B. M., 1996: Rainfall and climate variation over a sloping New Mexico plateau during the North American monsoon. J. Climate,9, 3432–3442.

  • Browning, K. A., and C. W. Pardoe, 1973: Structure of low-level jet streams ahead of midlatitude cold fronts. Quart. J. Roy. Meteor. Soc.,99, 619–638.

  • Carlson, T. N., 1991: Mid-Latitude Weather Systems. Harper Collins Academic, 507 pp.

  • Chang, F.-C., and J. M. Wallace, 1987: Meteorological conditions during heat waves and droughts in the United States Great Plains. Mon. Wea. Rev.,115, 1253–1269.

  • Chen, T.-C., and J. A. Kpaeyeh, 1993: The synoptic-scale environment associated with the low-level jet of the Great Plains. Mon. Wea. Rev.,121, 416–420.

  • ——, M.-C. Yen, and S. Schubert, 1996: Hydrological processes associated with cyclonic systems over the United States. Bull. Amer. Meteor. Soc.,77, 1557–1567.

  • Harrold, T. W., 1973: Mechanisms influencing the distribution of precipitation within baroclinic disturbances. Quart. J. Roy. Meteor. Soc.,99, 232–251.

  • Harshvardhan, R. Davies, D. A. Randall, and T. G. Corsetti, 1987: A fast radiation parameterization for atmospheric circulation models. J. Geophys. Res.,92, 1009–1016.

  • Helfand, H. M., and J. C. Labraga, 1988: Design of a non-singular level 2.5 second-order closure model for the prediction of atmospheric turbulence. J. Atmos. Sci.,45, 113–132.

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

  • Higgins, R. W., J. E. Janowiak, and Y. Yao, 1996a: A gridded hourly precipitation data base for the United States (1963–93). Atlas No. 1., NCEP/Climate Prediction Center, 47 pp.

  • ——, K. C. Mo, and S. D. Schubert, 1996b: The moisture budget of the central United States in spring as evaluated in the NCEP/NCAR and the NASA/DAO reanalyses. Mon. Wea. Rev.,124, 939–963.

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

  • Hoecker, W. J., 1963: Three southerly low-level jet systems delineated by the Weather Bureau special pibal network of 1961. Mon. Wea. Rev.,91, 573–582.

  • Hovanec, R. D., and L. H. Horn, 1975: Static stability and the 300 mb isotach field in the Colorado cyclonetic area. Mon. Wea. Rev.,103, 628–638.

  • Lindzen, R. S., 1967: Thermally driven diurnal tide in the atmosphere. Quart. J. Roy. Meteor. Soc.,93, 18–42.

  • Lyon, B., and R. Dole, 1995: A diagnostic comparison of the 1980 and 1988 U.S. summer heat wave–droughts. J. Climate,8, 1658–1675.

  • Maddox, R. A., 1983: Large-scale meteorological conditions associated with midlatitude, mesoscale convective complexes. Mon. Wea. Rev.,111, 1475–1493.

  • Meyers, S. D., B. G. Kelly, and J. J. O’Brien, 1993: An introduction to wavelet analysis in oceanography and meteorology: With application to the dispersion of Yanai waves. Mon. Wea. Rev.,121, 2858–2878.

  • Min, W., and S. Schubert, 1997: The climate signal in regional moisture fluxes: A comparison of three global data assimilation products. J. Climate,10, 2623–2642.

  • Molod, A., H. M. Helfand, and L. L. Takacs, 1996: The climatology of parameterized physical processes in the GEOS-1 GCM and their impact on the GEOS-1 Data Assimilation System. J. Climate,9, 764–785.

  • Moorthi, S., and M. J. Suarez, 1992: Relaxed Arakawa–Schubert: A parameterization of moist convection for general circulation models. Mon. Wea. Rev.,120, 978–1002.

  • Namias, J., 1955: Some meteorological aspects of drought with special reference to the summers of 1952–1954 over the United States. Mon. Wea. Rev.,83, 199–205.

  • ——, 1982: Anatomy of Great Plains protracted heat waves (especially the 1980 U.S. summer drought). Mon. Wea. Rev.,110, 824–838.

  • Newton, C. W., 1967: Severe convective storms. Advances in Geophysics, Vol. 12, Academic Press, 257–308.

  • Oglesby, R. J., 1991: Springtime soil moisture, natural climatic variability, and North American drought as simulated by the NCAR Community Climate Model 1. J. Climate,4, 890–897.

  • ——, K. A. Maasch, and B. Saltzman, 1989: Glacial meltwater cooling of the Gulf of Mexico: GCM implications for Holocene and present-day climate. Climate Dyn.,3, 115–133.

  • Paegle, J., 1984: Topographically bound low-level circulations. Riv. Meteor. Aeronaut.,44, 113–125.

  • Pfaendtner, J., S. Bloom, D. Lamich, M. Seablom, M. Sienkiewicz, J. Stobie, and A. da Silva, 1995: Documentation of the Goddard Earth Observing System (GEOS) Data Assimilation System-Version 1. NASA Tech. Memo. 104606, Vol. 4, 44 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • Pitchford, K. L., and J. London, 1962: The low-level jet as related to nocturnal thunderstorms over midwest United States. J. Appl. Meteor.,1, 43–47.

  • Rasmusson, E. M., 1967: Atmospheric water vapor transport and the water balance of North America: Part I. Characteristics of the water vapor flux field. Mon. Wea. Rev.,95, 403–426.

  • Reiter, E. R., 1969: Tropopause circulation and jet streams. World Survey of Climatology, Climate of the Free Atmosphere, D. F. Rex, Ed., Vol. 4, Elsevier, 85–193.

  • Roads, J. O., S.-C. Chen, A. K. Guetter, and K.-P. Georgakakos, 1994:Large-scale aspects of the United States hydrological cycle. Bull. Amer. Meteor. Soc.,75, 1589–1610.

  • Roebber, P. J., 1984: Statistical analysis and updated climatology of explosive cyclones. Mon. Wea. Rev.,112, 1577–1589.

  • Schemm, J.-K., S. Schubert, J. Terry, S. Bloom, and Y. Sud, 1992: Estimates of monthly mean soil moisture for 1979–89. NASA Tech. Memo. 104571, 252 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • Schubert, S., and Y. Chang, 1996: An objective method for inferring sources of model error. Mon. Wea. Rev.,124, 325–340.

  • ——, J. Pfaendtner, and R. Rood, 1993: An assimilated dataset for earth science applications. Bull. Amer. Meteor. Soc.,74, 2331–2342.

  • ——, and Coauthors, 1995: A multiyear assimilation with the GEOS-1 system: Overview and results. NASA Tech. Memo. 104606, Vol. 6, 183 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • Stensrud, D. J., 1996: Importance of low-level jets to climate: A review. J. Climate,9, 1698–1711.

  • Suarez, M. J., and L. L. Takacs, 1995: Documentation of the Aries-GEOS dynamical core: Version 2. NASA Tech. Memo. 104606, Vol. 5, 45 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • Sud, Y., and A. Molod, 1988: The roles of dry convection, cloud-radiation feedback processes and the influence of recent improvements in the parameterization of convection in the GLA GCM. Mon. Wea. Rev.,116, 2366–2387.

  • Takacs, L. L., and M. J. Suarez, 1996: Dynamical aspects of climate simulations using the GEOS General Circulation Model. NASA Tech. Memo. 104606, Vol. 10, 56 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • ——, A. Molod, and T. Wang, 1994: Goddard Earth Observing System (GEOS) General Circulation Model (GCM) Version 1. NASA Tech. Memo. 104606, Vol. 1, 100 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • Uccellini, L. W., and D. R. Johnson, 1979: The coupling of upper and lower tropospheric jet streams and implications for the development of severe convective storms. Mon. Wea. Rev.,107, 682–703.

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

  • ——, and F. R. Hartranft, 1969: Diurnal wind variations, surface to 30 kilometers. Mon. Wea. Rev.,97, 446–455.

  • Weng, H., and K.-M. Lau, 1994: Wavelets, period doubling, and time-frequency localization with application to organization of convection over the tropical western Pacific. J. Atmos. Sci.,51, 2523–2541.

  • Winkler, J. A., B. R. Skeeter, and P. D. Yamamoto, 1988: Seasonal variations in the diurnal characteristics of heavy hourly precipitation across the United States. Mon. Wea. Rev.,116, 1641–1658.

  • View in gallery

    The time-mean (May–June 1985–93) moisture transport in (a) the lower troposphere (surface to σ = 0.84) and (b) the middle and upper troposphere σ < 0.84. Units are m s−1 g kg−1.

  • View in gallery

    Time series of northward moisture transport (υq) profile at 32°N, 97.5°W. Vertical scale is approximate pressure level assuming 1000-mb surface pressure. Units are m s−1 g kg−1.

  • View in gallery

    (a) Power spectrum of υ (m s−1)2, (b) power spectrum of q (g kg−1)2, (c) power spectrum of υq (m s−1 g kg−1)2, and (d) q2 times the power spectrum of υ (m s−1 g kg−1)2 at 32°N, 97.5°W, σ = 0.97 for May–June 1985–93.

  • View in gallery

    (a) Profile of the standard deviation of υq (solid line) and approximate form q2 Var(υ) (dashed line) at 32°N, 97.5°W. Units are m s−1 g kg−1. (b) The vertical correlation with σ = 0.97 as a reference level at 32°N, 97.5°W.

  • View in gallery

    The standard deviation at σ = 0.97 of (a) υq and (b) uq for May–June 1985–93. Units are m s−1 g kg−1. The two components of the correlation (displayed as a vector) of the moisture flux (uq, υq) at σ = 0.97 with (c) υq at a base point at 32°N, 97.5°W and with (d) uq at base point at 38°N, 77.5°W for May/June 1985–93.

  • View in gallery

    Wavelet analysis of υq at 32°N, 97.5°W shown for the period May–August of 1993. The top panel shows the time series of υq. The bottom panel shows the real part of the wavelet transform for each frequency. Units are (m s−1 g kg−1)2. The wavelet analysis was performed on the entire 9-yr record.

  • View in gallery

    Longitude–pressure cross sections at 34°N of the composite (time lag = 0, see text) mean horizontal wind anomalies for (a) diurnal, (b) subsynoptic (2 < τ < 4 day), (c) synoptic (4 < τ < 8 days), and (d) supersynoptic (8 < τ < 16 day) timescales. Vectors pointing down denote northerly wind anomalies. Shading shows the full composite northward wind component for each frequency band. Units: m s−1.

  • View in gallery

    Same as Fig. 7 except for latitude–pressure cross sections at 97.5°W. Vectors pointing down denote westerly wind anomalies.

  • View in gallery

    Same as Fig. 7 except for profiles at 32°N, 97.5°W at different time lags. Vectors pointing down (to the right) are northerly (westerly) wind anomalies. The absisca in panel (a) is given in hours, and in days for the other panels.

  • View in gallery

    Composite mean diurnal surface pressure anomalies (contours, mb), moisture flux at σ = 0.97 (m s−1 g kg−1), and observed precipitation anomalies (shading, mm day−1) for lags (a) −12 h, (b) −6 h, (c) 0 h, and (d) + 6 h. Moisture flux vectors with magnitudes less than 20 m s−1 g kg−1 are omitted.

  • View in gallery

    Composite mean diurnal wind anomalies at σ = 0.97 (dark vectors) and σ = 0.5 (light vectors) for lags (a) −12 h, (b) −6 h, c) 0 h, and d) + 6 h. Units are m s−1. Wind vectors at σ = 0.97 with magnitudes less than 1 m s−1 are omitted.

  • View in gallery

    Composite mean subsynoptic surface pressure anomalies (contours, mb), moisture flux at σ = 0.97 (m s−1 g kg−1), and GEOS-1 precipitation anomalies (shading, mm day−1) for lags (a) −1 days, (b) −0.5 days, (c) 0 days, and (d) + 0.5 days. Moisture flux vectors with magnitudes less than 20 m s−1 g kg−1 are omitted.

  • View in gallery

    Composite mean subsynoptic wind anomalies at σ = 0.97 (dark vectors), σ = 0.5 (light vectors), and observed precipitation anomalies (shading, mm day−1) for lags (a) −1 day, (b) −0.5 day, (c) 0 day, and (d) + 0.5 day. Units are m s−1. In (b) the light contours indicate the speed of the wind anomalies at σ = 0.5. Wind vectors at σ = 0.97 with magnitudes less than 1 m s−1 are omitted.

  • View in gallery

    Composite mean synoptic surface pressure anomalies (contours, mb), moisture flux at σ = 0.97 (m s−1 g kg−1), and GEOS-1 precipitation anomalies (shading, mm day−1) for lags (a) −2 day, (b) −1 day, (c) 0 day, and (d) + 1 day. Moisture flux vectors with magnitudes less than 20 m s−1 g kg−1 are omitted.

  • View in gallery

    Composite mean synoptic wind anomalies at σ = 0.97 (dark vectors), σ = 0.5 (light vectors), and observed precipitation anomalies (shading, mm day−1) for lags (a) −2 day, (b) −1 day, (c) 0 day, and (d) +1 day. Units are m s−1. Wind vectors at σ = 0.97 with magnitudes less than 1 m s−1 are omitted.

  • View in gallery

    Composite mean supersynoptic surface pressure anomalies (contours, mb), moisture flux at σ = 0.97 (m s−1 g kg−1), and GEOS-1 precipitation anomalies (shading, mm day−1) for lags (a) −3 day, (b) −1 day, (c) 0 day, and (d) + 2 day. Moisture flux vectors with magnitudes less than 20 m s−1 g kg−1 are omitted.

  • View in gallery

    Composite mean supersynoptic wind anomalies at σ = 0.97 (dark vectors) and σ = 0.5 (light vectors), and observed precipitation anomalies (shading, mm day−1) for lags (a) −3 day, (b) −1 day, (c) 0 day, and (d) + 2 day. Units are m s−1. Wind vectors at σ = 0.97 with magnitudes less than 1 m s−1 are omitted.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 3 3 3
PDF Downloads 4 4 4

Subseasonal Variations in Warm-Season Moisture Transport and Precipitation over the Central and Eastern United States

View More View Less
  • 1 Data Assimilation Office, Laboratory for Atmospheres, NASA/Goddard Space Flight Center, Greenbelt, Maryland
Full access

Abstract

Subseasonal variations in warm-season (May–August) precipitation over the central and eastern United States are shown to be strongly linked to variations in the moisture entering the continent from the Gulf of Mexico within a longitudinally confined “channel” (referred to here as the Texas corridor or TC). These variations reflect the development of low-level southerly wind maxima (or jets) on a number of different timescales in association with distinct subcontinental and larger-scale phenomena. On the diurnal timescale, the TC moisture flux variations are tied to the development of the Great Plains low-level jet. The composite nighttime anomalies are characterized by a strong southerly moisture flux covering northeast Mexico and the southern Great Plains, and enhanced boundary layer convergence and precipitation over much of the upper Great Plains. The strongest jets tend to be associated with an anomalous surface low over the Great Plains, reflecting a predilection for periods when midlatitude weather systems are positioned to produce enhanced southerly flow over this region. On subsynoptic (2–4 days) timescales the TC moisture flux variations are associated with the development and evolution of a warm-season lee cyclone. These systems, which are most prevalent during the early part of the warm season (May and June), form over the central Great Plains in association with an upper-level shortwave and enhanced upper-tropospheric cross-mountain westerly flow. A low-level southerly wind maximum or jet develops underneath and perpendicular to the advancing edge of enhanced midtropospheric westerlies. The clash of anomalous southerly moisture flux and a deep intrusion of anomalous northerly low-level winds results in enhanced precipitation eventually stretching from Texas to the Great Lakes. On synoptic (4–8 days) timescales the TC moisture flux variations are associated with the propagation and intensification of a warm-season midlatitude cyclone. This system, which also occurs preferentially during May and June, develops offshore and intensifies as it crosses the Rocky Mountains and taps moisture from the Gulf of Mexico. Low-level southerly wind anomalies develop parallel to the mid- and upper-level winds on the leading edge of the trough. Widespread precipitation anomalies move with the propagating system with reduced rainfall occurring over the anomalous surface high, and enhanced rainfall occurring over the anomalous surface low. On still longer timescales (8–16 days) the variations in the TC moisture transport are tied to slow eastward-moving systems. The evolution and structure of the mid- and low-level winds are similar to those of the synoptic-scale system with, however, a somewhat larger zonal scale and spatially more diffuse southerly moisture flux and precipitation anomalies.

* Additional affiliation: General Sciences Corporation, a subsidiary of Science Applications International Corporation, Laurel, Maryland.

Corresponding author address: Dr. Siegfried D. Schubert, NASA/Goddard Space Flight Center, Mail Code 910.3, Greenbelt, MD 20771.

Abstract

Subseasonal variations in warm-season (May–August) precipitation over the central and eastern United States are shown to be strongly linked to variations in the moisture entering the continent from the Gulf of Mexico within a longitudinally confined “channel” (referred to here as the Texas corridor or TC). These variations reflect the development of low-level southerly wind maxima (or jets) on a number of different timescales in association with distinct subcontinental and larger-scale phenomena. On the diurnal timescale, the TC moisture flux variations are tied to the development of the Great Plains low-level jet. The composite nighttime anomalies are characterized by a strong southerly moisture flux covering northeast Mexico and the southern Great Plains, and enhanced boundary layer convergence and precipitation over much of the upper Great Plains. The strongest jets tend to be associated with an anomalous surface low over the Great Plains, reflecting a predilection for periods when midlatitude weather systems are positioned to produce enhanced southerly flow over this region. On subsynoptic (2–4 days) timescales the TC moisture flux variations are associated with the development and evolution of a warm-season lee cyclone. These systems, which are most prevalent during the early part of the warm season (May and June), form over the central Great Plains in association with an upper-level shortwave and enhanced upper-tropospheric cross-mountain westerly flow. A low-level southerly wind maximum or jet develops underneath and perpendicular to the advancing edge of enhanced midtropospheric westerlies. The clash of anomalous southerly moisture flux and a deep intrusion of anomalous northerly low-level winds results in enhanced precipitation eventually stretching from Texas to the Great Lakes. On synoptic (4–8 days) timescales the TC moisture flux variations are associated with the propagation and intensification of a warm-season midlatitude cyclone. This system, which also occurs preferentially during May and June, develops offshore and intensifies as it crosses the Rocky Mountains and taps moisture from the Gulf of Mexico. Low-level southerly wind anomalies develop parallel to the mid- and upper-level winds on the leading edge of the trough. Widespread precipitation anomalies move with the propagating system with reduced rainfall occurring over the anomalous surface high, and enhanced rainfall occurring over the anomalous surface low. On still longer timescales (8–16 days) the variations in the TC moisture transport are tied to slow eastward-moving systems. The evolution and structure of the mid- and low-level winds are similar to those of the synoptic-scale system with, however, a somewhat larger zonal scale and spatially more diffuse southerly moisture flux and precipitation anomalies.

* Additional affiliation: General Sciences Corporation, a subsidiary of Science Applications International Corporation, Laurel, Maryland.

Corresponding author address: Dr. Siegfried D. Schubert, NASA/Goddard Space Flight Center, Mail Code 910.3, Greenbelt, MD 20771.

1. Introduction

The amount of moisture entering the United States along the Gulf Coast is one of the key factors determining the climate of the central and eastern parts of the continent.1 The concentrated and intense nature of the southerly transport, and its confinement to the very lowest layers of the atmosphere during the summer (in contrast to winter when the flow is vertically more extensive) was documented more than 40 years ago by Benton and Estoque (1954) employing one year (1949) of geostrophic winds derived from constant-pressure charts. Rasmusson (1967) expanded upon the work of Benton and Estoque employing a more comprehensive station dataset of winds and humidity for the time period 1 May 1961–30 April 1963. In particular, that study found a pronounced diurnal cycle in the July transport with a substantial increase in transport between 0000 and 1200 UTC. It was also noted that while the January inflow is primarily due to eddy transport, the July inflow is primarily due to transport by the mean wind associated with the western edge of the subtropical high. More recently Roads et al. (1994) reexamined the moisture budget over the United States using precipitation observations and employing National Meteorological Center [now known as the National Centers for Environmental Prediction (NCEP)] analyses for the wind and moisture information for the period 1984–90. The strong Gulf Coast moisture flux was clearly evident in the summer climatology based on the global analyses. They found that, on short timescales, precipitation fluctuations are strongly coupled with fluctuations in moisture flux convergence, whereas in the long-term mean precipitation is balanced by evaporation.

Helfand and Schubert (1995) focused on the role of the Great Plains low-level jet (GPLLJ; e.g., Bonner 1968; Paegle 1984; Stensrud 1996) in the moisture budget of the United States. They examined a simulation of the GPLLJ with the Goddard Earth Observing System (GEOS-1) GCM during spring and found that as much as one-third of the moisture entering the United States is associated with transport by the GPLLJ, most of which occurs during the nighttime. They also found that, while the transport is primarily due to the mean flow, the mean flow itself is determined largely by the strong nighttime acceleration of the jet. Recently Higgins et al. (1997) examined the influence of the GPLLJ on the summertime precipitation over the United States employing the NCEP–National Center for Atmospheric Research (NCAR) and National Aeronautics and Space Administration/Data Assimilation Office (NASA/DAO) reanalysis products. They found that the dominant diurnal signal in the Great Plains precipitation during spring and summer is associated with jet events.

The relationship between variations of moisture transport from the Gulf of Mexico and the weather and climate of the continental United States has been studied for various timescales. The role of the low-level jet (LLJ) on subsynoptic and synoptic timescales has received considerable attention. Pitchford and London (1962) demonstrated a strong relationship between nocturnal thunderstorms and the occurrence of convergence associated with an LLJ over the Midwest (centered on Omaha, Nebraska) during nonfrontal summer days. Bonner (1966) found, in a case study of thunderstorm activity, that the large-scale vertical velocity field associated with an LLJ was responsible for the persistence and dissipation of thunderstorm activity that had formed previously from frontal activity or convection over the mountains. Newton (1967) reviewed synoptic conditions favorable for convective storms. The conditions include strong LLJs bringing in warm, moist air, and the winds veering with height so that the upper-level and LLJs tend to be orthogonal. The combination of low-level heat and moisture influx (from the Gulf) and upper-level divergence produces a tendency for severe convection to occur at the intersection of the two jets. Maddox (1983) highlighted the role of the LLJ stream as a recurrent feature of mesoscale convective complex’s precursor environment. Harrold (1973) defined the concept of a “conveyor belt” for baroclinic disturbances, which is a flow (a few hundred kilometers wide and a few kilometers deep) parallel to and immediately ahead of the surface cold front thought to be responsible for much of the initial precipitation reaching the ground. Browning and Pardoe (1973) employed a case study approach to show the typical structure of LLJs associated with midlatitude cold fronts over the British Isles. The jets were found to occur just ahead of the surface cold front in a region of reversed horizontal temperature gradient within the convective boundary layer leading to jet maxima of 25–30 m s−1 between 900 and 850 mb. Diurnal effects were found to be unimportant and the winds within the jet were essentially in geostrophic balance.

On longer timescales, planetary circulation changes play an important role in modulating the moisture entering the United States from the Gulf of Mexico. For example, Namias (1955) showed that an anomalous anticyclonic cell dominated the south-central United States and adjacent oceans during the summer drought of 1952–54, acting to steer storm systems away from the continent and inhibit the flow of moist air from the Gulf of Mexico. Namias argued that the explanation for such droughts lies in the persistent recurrence of particular flow patterns in the general circulation and, in particular, the development and coupling of an anomalous North American summer anticyclone with circulation anomalies over the Pacific and Atlantic Oceans. More recent studies with more extensive datasets have confirmed the tendency for anomalous anticyclones to be present over the United States during heat waves and/or droughts (e.g., Namias 1982; Chang and Wallace 1987). Oglesby (1991) further examined the importance of moisture transport from the Gulf of Mexico and its role in drought over the United States. Lyon and Dole (1995) found reductions in Gulf Coast moisture transport during the recent 1988 drought over the midwest. Oglesby et al. (1989) showed that reductions in Gulf of Mexico SST reduced the moisture transport from the Gulf of Mexico in a GCM simulation, and suggested this as a mechanism for explaining climatic fluctuations associated with meltwater from the last glaciation.

It is clear from the above studies that LLJs are rather ubiquitous features of the circulation at a number of timescales. The terminology LLJ is, however, somewhat vague and has been used to designate somewhat different phenomena by different authors. For example, Reiter (1969) distinguishes between an LLJ and an (boundary layer) inversion wind maxima. The latter, while it may have vertical shear as strong as upper-tropospheric jets, has generally much weaker horizontal shear. Uccellini and Johnson (1979) point out that LLJs associated with leeside cyclogenesis/troughing have minimal diurnal cycles and extend above the boundary layer to 850 mb. Lacking in previous studies is a systematic characterization of the basic phenomenology of the various low-level wind maxima and their role in the weather and climate of the United States.

The current study is an attempt to characterize the variability of the moisture flux entering the United States from the Gulf of Mexico on subseasonal timescales. In particular, we examine how the low-level wind maxima differ as a function of timescale in terms of their contribution to the moisture transport and the embedding larger-scale circulation. A compositing approach is used to determine the characteristic large-scale circulation patterns and precipitation anomalies associated with enhanced northward moisture flux as a function of timescale. The study employs a 9-yr reanalysis of atmospheric observations carried out with the GEOS-1 data assimilation system. Section 2 provides an overview of the moisture flux and its variability. The results of the compositing are presented in section 3. The discussion and conclusions are given in section 4.

2. Variations in moisture transport

This study employs moisture and wind fields from 9 yr (1985–93) of a reanalysis performed with the GEOS-1 Data Assimilation System (DAS; Schubert et al. 1993). The GEOS-1 DAS consists of the GEOS-1 GCM (Takacs et al. 1994), and an optimal interpolation scheme described in Pfaendtner et al. (1995). The GEOS-1 system is summarized in the appendix. Early results from the assimilation (Schubert et al. 1995; Higgins et al. 1996b; Higgins et al. 1997) show that there is considerable climate bias in the estimated moisture budget over the United States. The twice-daily radiosonde wind observations are insufficient to adequately resolve the diurnal cycle, and in particular, they fail to capture the peak phase of the Great Plains LLJ, which falls between the 0000 and 1200 UTC observing times. This, together with the poor vertical resolution of the observations, makes the quality of the low-level moisture flux dependent on the model’s ability to capture the key wind fluctuations in the planetary boundary layer. Recent results (Min and Schubert 1997) show that, while there are substantial differences between the NASA/DAO and NCEP–NCAR reanalyses in the estimated moisture divergence, the monthly anomalies in the moisture fluxes are quite reasonable. Also, offline comparisons employing significant-level observations (not shown) verify that the assimilated data employed here are providing quite reasonable estimates of the diurnal variations in the Great Plains low-level moisture flux.

All results shown here are based on output saved at the model sigma levels [see Fig. 3 in Schubert et al. (1993)] to avoid errors resulting from vertical interpolation to pressure. The sigma coordinate is a terrain-following coordinate system with the top of the model at sigma equal to zero and the surface at sigma equal to one. For example, sigma equal to 0.97 is approximately 30 mb (300 m) above the surface. Figures showing profiles are labeled with approximate pressure levels assuming a surface pressure of 1000 mb. Figure 1 shows the mean transport of moisture for May–June (1985–93) for the lower troposphere (pressures greater than approximately 850 mb) and the middle and upper troposphere (pressures less than 850 mb).2 The low-level transport (Fig. 1a) is dominated by an influx into the United States from the Gulf Coast between about 92.5° and 102.5°W. Very little moisture enters from the West Coast (the transport is primarily northerly and parallel to the coast) presumably due to the blocking effect of the Rocky Mountains. The main outflow occurs along the East Coast north of about 35°N. The moisture flux in the middle and upper troposphere (Fig. 1b) is quite different from that at low levels, showing a generally zonal pattern. The inflow from the west coast is larger than at low levels, but still small compared with the strong outflow along the East Coast at these levels. The mean flux of water over the United States can be summarized as consisting primarily of a rather confined region of low-level southerly/southeasterly inflow along the Gulf Coast (referred to here as the Texas corridor or TC), turning westerly and upward over the central United States, and eventually exiting along a broad vertically extensive region of the East Coast. Qualitatively similar results have been obtained for July and August (not shown).

Clearly, the mean TC moisture flux is very important, and any significant variations in this very narrow moisture conduit would have a substantial impact on the weather and/or climate of the United States. Figure 2 shows a time series for May and June 1993 of the vertical profile of the northward moisture flux at a grid point near Ft. Worth (32°N, 97.5°W) within the TC. The time series is dominated by pronounced and regular diurnal variations with the largest northward flux occurring during the night and at very low levels, consistent with the occurrence of the GPLLJ (e.g., Hoecker 1963;Bonner 1968). There are also substantial variations on longer timescales. For example, there is a synoptic timescale “envelope” of intensified and vertically more extensive flux during the second week of May. A spectral analysis of the low-level northward flux (Fig. 3c) clearly reveals the strong diurnal component as well as substantial variability in the synoptic (4–8 day) timescales. The individual spectra of q and υ both show the strong diurnal signal, though the moisture spectrum tends to show more low-frequency variations. The flux spectrum is, in fact, almost entirely determined by the variations in the wind; that is,
i1520-0442-11-10-2530-e1
The wind spectrum weighted by the square of the mean moisture (Fig. 3d) is almost indistinguishable from the flux spectrum (Fig. 3c). The vertical profile of the northward flux variability (Fig. 4a) near Ft. Worth shows a similar result, with the strongest variability in the boundary layer, which is dominated by the wind variations. The vertical correlations with the flux at low levels (σ = 0.97, Fig. 4b) show that the variations tend to be coherent in the first 2 km of the atmosphere (correlations greater than 0.6). The correlations decrease approximately linearly with pressure to a value of about 0.4 just below the midtroposphere, beyond which they decay at a slower rate.

The region of high variability in the low-level northward flux is not limited to the immediate vicinity of Ft. Worth but exists throughout the Great Plains region (Fig. 5a), with the largest variability occurring in a rather narrow region extending from southern Texas to Iowa. This variability is large with respect to the mean northward flux at this level (not shown), which has a maximum of just over 70 g kg−1 m s−1 over Texas. The analogous map for the eastward flux (Fig. 5b) has the largest variance extending eastward across the upper midwest consistent with the mean eastward flux shown in Fig. 1. The region of variability in the flux over the Great Plains is not only extensive but also spatially coherent. Figure 5c shows, for example, the correlation of the northward flux at Ft. Worth with both components of the flux everywhere else. The largest correlations (displayed as vectors) extend from the Gulf Coast to the upper Midwest, indicating coherent northward (southward) and eastward (westward) fluctuations in transport. Substantial correlations also occur over the western States showing that northward flux at Ft. Worth is associated with northwesterly flux over this region. Correlations over the oceans further suggest that the fluctuations at Ft. Worth are part of spatially coherent (continental-scale) modes of variability. Similarly, Fig. 5d suggests that westerly fluctuations in the outflow region over the East Coast are associated with large-scale circulation features: in this case a cyclonic anomaly centered over the northeast and a weakly correlated anticyclonic anomaly to the south.

The strong spatial coherence shown above suggests that the flux at a single point within the TC can serve as a useful indicator of the variations in the flux over a much broader region of the United States. In the following analysis, the moisture flux near Ft. Worth (i.e., at the grid point at 32°N, 97.5°W) is employed to index the moisture entering the United States from the Gulf of Mexico. The time series for low-level (sigma = 0.97) northward moisture flux at Ft. Worth is spectrally decomposed locally (in time) by using the wavelet analysis technique (e.g., Meyers et al. 1993; Weng and Lau 1994). The algorithm employed here is that described in Weng and Lau based upon the Morlet wavelet:
i1520-0442-11-10-2530-e2
where hΨ = 5.4. The Morlet transform isolates waves that are locally sinusoidal at time = b.

Figure 6 shows shows the wavelet transform of the low-level moisture transport at Ft. Worth for May–August 1993. The upper panel shows the original (untransformed) time series of the transport. The evolution is characterized by a number of transitions and asymmetries. The diurnal signal is particularly strong and regular during July and August. The diurnal signal is weaker and more intermittent in May, when it occurs primarily during the positive peaks of longer-term (weekly) fluctuations. As we shall see, this reflects the tendency for the Great Plains LLJ to develop most strongly in the presence of enhanced southerlies associated with weather systems. Beginning in late June there is a general shift to positive (northward) flux corresponding to the time when the diurnal signal becomes quite regular. The wavelet analysis provides a local (in time) spectrum of these variations. The lower panel of Fig. 6 shows the real part of the Morlet transform. Note that the transform has two dimensions: period and time or translation. The real part of the transform essentially reconstructs that portion of the time series that oscillates locally at time b. The results clearly show the strong diurnal signal associated with the GPLLJ (cf. Fig. 3) and the transition in the strength of the diurnal signal from spring to summer. Subsynoptic (2–4 day), synoptic (4–8 day), and supersynoptic (8–16 day) events are strongest during May and June, whereas a clear low-frequency (16–32 day) fluctuation occurred during June of that year. There is a strong seasonal component characterized by the enhanced northward transport during most of July; this is apparently associated with the excessive precipitation and flooding that occurred in the midwest at this time. Other years show qualitatively similar variability, though the timing of the transitions of the dominant timescales differs considerably from year to year.

The above results suggest that the variability in the low-level moisture flux at Ft. Worth occurs on a wide range of distinct timescales, which can be separated into the following frequency bands: 1) diurnal, 2) subsynoptic (2 < τ < 4 day), 3) synoptic (4 < τ < 8 day), and 4) supersynoptic (8 < τ < 16 day). There is also evidence of a low-frequency signal (16 < τ < 32 day) during this year; however, the 9 yr of assimilation analyzed here proved to be insufficient to produce stable results for the compositing approach outlined below. While the separation of the diurnal signal is readily justifiable by the strong spectral peak, the choice of the other three frequency bands is somewhat arbitrary. We note, however, that these frequency bands are not inconsistent with the spectra shown in Fig. 3 and, with the exception of the 2–4-day band, they are generally consistent with those of many previous studies of midlatitude intraseasonal variability though much of that work has focused on the winter season (e.g., Blackmon et al. 1984).

The compositing approach employs all 9 yr (May–August 1985–93) of the wavelet time series at the grid point 32°N, 97.5°W as indices. The indices are defined as the band-averaged amplitude of the northward moisture flux from the wavelet transform (e.g., those shown in the lower panel of Fig. 6) for each of the four frequency ranges listed above. Composites of various fields are produced for each frequency band by averaging these fields over periods of enhanced northward (positive) moisture transport near Ft. Worth with a single standard deviation in each index as a cutoff. In addition, composites are produced at different time lags with respect to the time of maximum northward transport at Ft. Worth to help show the evolution of the composites. Table 1 shows the number of events counted for each frequency band over the nine warm seasons. The distribution of the diurnal events is rather uniform with some tendency for more events to occur during the latter two months of the warm season (27% vs 23%). This indicates that 1993 was a rather unusual year in producing such a distinct difference in the occurrence of the diurnal signal between May–June and July–August. The longer timescales clearly have a predilection for spring and early summer. For example, the subsynoptic events during May occur about twice as often as during July or August. When grouped according to time of day (Table 1) it is clear that the diurnal events are primarily a nighttime phenonenon, with approximately three-fourths of the events occurring at local midnight and the other one-fourth occurring at 0600 LT. The events for the other frequency bands are distributed more evenly over the day.

3. Composite fields

This section presents the composite fields based on the moisture flux indices developed for Ft. Worth, Texas, and described in the previous section. We present both full composite and anomaly fields. The former are the averages over the appropriate composite members, while the latter are the composite averages with the long-term (1985–93) May through August mean removed. The section is organized into subsections dealing with each of the frequency bands discussed above. Precipitation anomalies are shown from both the GEOS-1 assimilation and from a gridded hourly U.S. station dataset developed by NCEP (Higgins et al. 1996a). We begin by presenting several composite cross sections to help establish the basic space–time structures associated with the moisture variability entering through the TC.

Figure 7 shows longitude–pressure cross sections of the composite horizontal wind field at 34°N for each frequency band. This depicts the average structure during those times when the northward moisture flux at Ft. Worth is at a maximum. On the diurnal timescale (Fig. 7a) strong low-level southerly wind anomalies occur between about 90° and 102.5°W. The largest anomalies in this region occur in the lowest 100 mb. Westerly low-level wind anomalies occur over the mountainous region of the western United States. Wind anomalies are small (less than 5 m s−1) in the middle troposphere, whereas somewhat larger southwesterly anomalies occur above 300 mb west of about 97.5°W, and northeasterly anomalies occur to the east. The largest composite southerly winds occur near 100°W in the lowest 50 mb, with maximum winds greater than 12 m s−1. On subsynoptic timescales (Fig. 7b) the wind anomalies are vertically extensive and primarily westerly above 800 mb. There is a suggestion of a wave structure in the middle and upper troposphere with a tendency for northerly winds west of 100°W and southerly winds to the east. At low levels the wind anomalies are northerly just west of 100°W and southerly to the east, with peak southerly winds of 4–8 m s−1 centered at 97.5°W. On synoptic timescales (Fig. 7c) the wind anomalies show a well-defined and vertically extensive trough extending over much of the western and central United States at this latitude. The vertically extensive wind anomalies between about 92.5° and 102.5°W occur as part of the leading (east) edge of the anomalous trough: the full composite low-level southerly winds (anomaly plus mean) in that region are greater than 8 m s−1. On longer timescales (Fig. 7d) a relatively strong anomalous trough occurs above 800 mb centered near 112.5°W (about 10° farther west than the synoptic composite). Southwesterly anomalies occur throughout the middle and lower troposphere between about 92.5° and 102.5°W.

Figure 8 shows latitude–pressure cross sections of the composite horizontal wind field for each frequency band at the approximate center (97.5°W) of the strongest low-level southerly winds shown in the Fig. 7. On the diurnal timescale (Fig. 8a), strong low-level southerly wind anomalies (greater than 5 m s−1) extend from 22° to 38°N: these anomalies are southeasterly south of 32°N and become southwesterly to the north. This latitude band shows the strongest low-level southerly winds (greater than 8 m s−1 in the lowest 100 mb), with maximum southerly winds greater than 12 m s−1 occurring at 32°N at about 300 m above the surface. Wind anomalies are relatively weak in the middle troposphere, becoming stronger above 300 mb where they are characterized by northeasterlies south of 30°N and southwesterlies to the north of 36°N. On subsynoptic timescales (Fig. 8b) the wind anomalies are westerly over much of the upper and middle troposphere south of 42°N. At low levels the wind anomalies are southerly to the south of 36°N and northerly to the north of that latitude, suggesting an anomalous convergence of warm southerly air and cold northerly air at that latitude. The remaining frequency bands (Figs. 7c,d) show a similar north–south wind profile, with generally southwesterly wind anomalies extending from low levels and low latitudes to high latitudes and the upper troposphere. Maximum low-level southerly winds (greater than 8 m s−1) are confined to the lowest 100 mb and occur in a broad latitude band extending from about 24° to about 38°N. Relatively strong southerly winds also occur above 300 mb north of about 36°N.

Figure 9 summarizes the time evolution of the composite Ft. Worth horizontal wind profile for each frequency band. The time lags are with respect to the time of maximum (t = 0) northward moisture transport. The strongest diurnal wind variation (Fig. 9a) occurs in the lowest 50 mb. The wind anomalies in the lowest 100 mb show a marked turning from southeasterly (t = −6 h) to southerly (at t = 0, the time of maximum southerly flux) to southwesterly (t = +6 h). There is a generally enhanced southerly flow throughout the lower troposphere, and a weak diurnal signal extending well into the upper troposphere. As mentioned in the previous section, approximately three-fourths of the composite members occur at local midnight and the other one-fourth occur at 0600 LT, indicating the evolution is tied to the life cycle of the GPLLJ (Helfand and Schubert 1995). The subsynoptic composite (Fig. 9b) shows weaker wind anomalies than the diurnal composite but the local maximum in the southerlies, while still confined to the lower troposphere, is less concentrated in the boundary layer. The low-level wind anomalies evolve from northerly at t = −1.5 day to southerly at t = 0 back to northerly at t = +1.5 day. At upper levels the wind anomalies are generally from the west. The synoptic signal (Fig. 9c) is similar to that of the subsynoptic signal at low levels. These wind anomalies, however, vary on a longer (approximately 5 day) timescale and vary coherently well into the upper troposphere. The substantial low-level southerly wind anomalies (greater than 4 m s−1) last 2–3 days. The supersynoptic wind anomalies (Fig. 9d) are characterized by relatively long periods of low-level southerlies (greater than 4 m s−1) extending up to about 700 mb.

We next examine the spatial maps of various composited fields associated with the Ft. Worth (32°N, 97.5°W, σ = 0.97) moisture flux indices described above. For each frequency band two figures are presented showing 1) the evolution of the low-level moisture flux superimposed on the surface pressure and precipitation anomalies, and 2) the evolution of the midtropospheric and boundary layer wind anomalies. The latter figure also contains, for comparison with the precipitation anomalies from the GEOS-1 assimilation, the precipitation anomalies based on station observations. This is done to help establish the credibility of this quantity, which is not directly assimilated but computed from the model parameterizations during the course of the assimilation. The only exception to this format is for the diurnal signal for which (as noted below) we show only the observed precipitation anomalies.

a. Diurnal composites

Figure 10 shows the evolution of the diurnal low-level moisture transport, anomalous surface pressure and precipitation anomalies for −12, −6, 0, +6 h relative to the time of maximum northward low-level moisture flux at Ft. Worth. The members of the lag 0 composite occur primarily at local midnight or early morning at Ft. Worth (see previous section), and in the following we refer to the time of the composites accordingly. We show here (for the diurnal timescale only) the precipitation based on hourly station observations (Higgins et al. 1996a) since the diurnal cycle of the precipitation from the assimilation is unrealistic (Schubert and Chang 1996; Higgins et al. 1996b). In particular, the diurnal cycle consists of unrealistically strong afternoon and weak nocturnal rainfall. The weak diurnal rainfall occurs despite a realistic diurnal evolution of the moisture flux patterns in the assimilation (see below; also Min and Schubert 1997), suggesting a problem with one or more of the model’s physical parameterizations. For example, it may be that the excessive afternoon precipitation depletes the boundary layer moisture, thereby inhibiting nocturnal convection in the assimilation, or there may be insufficient nighttime radiational cooling at the cloud tops due to errors in the radiation and/or cloud parameterizations. In any event, the deficiencies in the summertime diurnal cycle of precipitation appears to be a problem common to other reanalysis data (e.g., Higgins et al. 1996b). We shall see in the following that, despite this problem, the precipitation at the longer timescales is quite realistic.

The surface pressure anomalies are dominated by a low pressure anomaly of more than 1.5 mb extending over much of the Great Plains region throughout the day. This low pressure anomaly reflects the tendency for the enhanced southerly flux at the diurnal timescale to occur during those times when midlatitude weather systems are positioned to produce enhanced southerly flow over the Great Plains. This was noted earlier in reference to Fig. 6 and is further quantified in Table 2. The table shows, for example, that more than one-half (88/167) of all the diurnal events occur in the presence of southerly synoptic timescale moisture flux (compared with only 6% for northly synoptic timescale flux). A similar dependence of the diurnal flux is seen for the subsynoptic and supersynoptic timescales. The surface pressure anomaly has a weak diurnal cycle, with the minimum pressure center moving about 10° lat (35° to 45°N) during the course of the day. Somewhat weaker positive surface pressure anomalies form along the east coast during the night.

The dominant signal in the moisture flux is the increasing flux during the course of the day, with maximum southerly flux occurring at local midnight or 0600 LT and extending over much of the Great Plains. The flux turns southwesterly during the night, extending toward the northeast (Fig. 10d). Positive precipitation anomalies develop in the northern Great Plains in the late afternoon (Fig. 10b) coinciding with the region of low-level moisture convergence. The precipitation anomalies increase in strength and move northeastward (following the region of moisture flux convergence): at local midnight the positive precipitation anomalies are most pronounced (maximum values greater than 3 mm day−1 over Nebraska and Iowa) with values exceeding 2 mm day−1 over much of the upper Midwest (Fig. 10c). Substantial negative precipitation anomalies (less than −2 mm day−1) occur over the southeast at this time. By early morning the positive precipitation anomalies have moved farther to the northeast (Minnesota and Wisconsin) and decreased in strength, whereas the negative precipitation anomalies (while also weakening) extend along much of the East Coast (Fig. 10d). The pronounced nocturnal maximum associated with nocturnal jet events is consistent with other studies that have documented a general nocturnal maximum in rainfall over the Great Plains during summer (e.g., Wallace 1975; Winkler et al. 1988), and the work of Higgins et al. (1997), who showed a strong link between the occurrence of the GPLLJ and nocturnal precipitation.

Figure 11 shows the corresponding evolution of the midtropospheric (σ = 0.5) and low-level (σ = 0.97) wind anomalies. The midtropospheric wind anomalies show a continental-scale wave pattern with an anomalous cyclonic circulation over the western United States and an anomalous anticyclonic circulation to the east. This is consistent with the surface pressure anomalies discussed earlier and reflects the predilection for the enhanced southerly flux at diurnal timescales to occur during periods when the longer timescale weather systems are positioned to generate enhanced southerly flux over the Great Plains. The upper levels also exhibit a substantial diurnal cycle with the strongest southerlies over the central part of the continent occurring at local noon (lag = −12, Fig. 11a). At low levels over the western half of the continent, the wind anomalies are opposite (anticyclonic) to the midtropospheric anomalies with relatively strong northerly wind anomalies over the central United States (near 100°W) and south/southeasterlies along the west coast. During the course of the day the midtropospheric anomalies weaken, almost disappearing over the Great Plains at local midnight (Fig. 11c), and building again by the early morning (Fig. 11d). We note that during the night the largest upper-level wind anomalies occur above 300 mb (not shown) and are dominated by an anticyclonic anomaly over the eastern half of the continent. At low levels the west coast wind anomalies turn clockwise, leading to onshore flow in the evening (Fig. 11b) and offshore flow during the early morning (Fig. 11d). Over the Great Plains the low-level wind anomalies also turn clockwise with upslope anomalies in late afternoon (Fig. 11b), enhanced southerly flow at midnight (Fig. 11c), and southwesterly flow anomalies during early morning (Fig. 11d).

The diurnal cycle of the wind anomalies in the lower and midtroposphere described above appear to be in part a rather complicated response to diurnal variations in near-surface forcing (see, e.g., Blackadar 1957), tidal motions associated with the absorption of solar energy by water vapor (Lindzen 1967), and possibly a response to the diurnal variations in latent heating. The low-level wind variations are clearly tied to the development of the GPLLJ as documented from observations by Bonner (1968) and many others, and more recently in simulations with the current GCM by Helfand and Schubert (1995). The exact nature of the forcing of the GPLLJ is still not fully understood, though it is likely that a number of mechanisms including terrain effects and the variation in eddy viscosity play a major role (e.g., Bonner and Paegle 1970; Stensrud 1996). The strong collocation between the observed precipitation anomalies and the low-level moisture flux (Fig. 10c) supports the idea that the nocturnal precipitation is closely tied to the development of the GPLLJ (e.g., Pitchford and London 1962; Wallace 1975; Higgins et al. 1997). This conclusion must be tempered, however, by the fact that the current assimilation system does not produce realistic nocturnal precipitation. It is likely that the role of the jet is primarily to provide the low-level convergence necessary to lift the boundary layer air. In view of the short time the GPLLJ has to advect moisture to the central Great Plains, the moisture and convective instability of the nighttime boundary layer must be determined by the “preconditioning” of the boundary layer from previous LLJs and the daytime history of evaporation and precipitation. As pointed out above, the GEOS-1 GCM generates unrealistically strong precipitation during the daytime, which may be a factor in supressing the nocturnal rainfall.

Above the boundary layer, the diurnal wind variations are consistent with independent station observations presented by Wallace and Hartranft (1969) in their investigation of tidal motions for the period 1950–64. In particular, the time-mean (May–August) 0000–1200 UTC wind anomalies extending up to 100 mb (not shown) are consistent with the annual averages of the 0000–1200 UTC wind anomalies shown in Figs. 1–4 of that study.

b. Subsynoptic composites

Figure 12 shows the evolution of the subsynoptic low-level moisture transport, anomalous surface pressure and precipitation anomalies for −1, −0.5, 0, +0.5 days relative to the time of maximum northward low-level moisture flux at Ft. Worth. The precipitation anomalies are from the assimilation and are quite similar to the observed anomalies (cf. Fig. 13). The main difference is that the positive precipitation anomalies at lag +0.5 are somewhat more intense (exceeding 3 mm day−1) in the observations (cf. Figs. 12d and 13d). The precipitation anomalies from the assimilation provide estimates over the adjacent ocean regions that are not available from the observations.

One of the key features of the surface pressure is a low pressure anomaly over the central Great Plains (Fig. 12a), which moves south (near 35°N, Fig. 12b) and then east, extending from Texas to Wisconsin (Fig. 12c), and eventually covers much of the eastern United States. A high pressure anomaly develops behind the trough (Fig. 12b), moving south and east (Fig. 12c), and eventually extends deep into the south central United States (Fig. 12d). Southerly low-level moisture flux occurs over the Great Plains along the eastern side of the low pressure anomaly (Fig. 12a). The region of southerly flux moves eastward (becoming southwesterly) following the low pressure anomaly, and by lag 0 is situated over a region of positive precipitation anomalies, which extend from northern Texas–Oklahoma to the upper Midwest (Fig. 12c). One-half day later the moisture flux (now predominately southwesterly) and the positive precipitation anomalies (contained within the low pressure anomaly) have moved farther east and weakened (Fig. 12d). Weak negative precipitation anomalies occur over the Gulf of Mexico throughout the subsynoptic evolution, and by the end of the period extend along the southeast coast (Fig. 12d).

Figure 13 shows the corresponding evolution of the midtropospheric (σ = 0.5) and low-level (σ = 0.97) wind anomalies. One day prior to the maximum flux at Ft. Worth, a weak cyclonic anomaly is situated over the Northern Plains, together with an upstream weak anticyclonic wind anomaly along the west coast north of 45°N (Fig. 13a). Much of the United States south of about 40°N is covered by anomalous westerlies associated with both the anomalous northwest trough and a weak east coast trough. The anomalous trough over the northwest moves eastward (Fig. 13b) intensifying on the lee side of the Rocky Mountains and extending deep into the south-central United States (Figs. 13c and 13d). The main low-level wind anomalies develop at a lag of −0.5 day and consist primarily of a northerly wind under the backside of the upper trough, and weak southerly wind anomalies under the leading edge of the upper trough (just ahead of the maximum in the speed of the anomalous wind) in association with positive precipitation anomalies (>2 mm day−1) extending from northern Texas into Kansas. At lag 0 (Fig. 13c), the northerly and southerly wind anomalies clash along a line extending from Texas to Wisconsin. The positive precipitation anomalies lie south of that line in a region characterized by midtropospheric westerly and lower-tropospheric southerly wind anomalies. A half-day later the northerly and northeasterly wind anomalies dominate much of the southern Great Plains, while weak southwesterly anomalies are confined to the east/southeast (Fig. 13d).

The above sequence of composites are consistent with the development and evolution of a warm-season lee cyclone. The composite initial surface development on the lee side of the Rockies occurs in a region characterized by a maximum in initial warm-season cyclone formation (e.g., Roebber 1984). Such systems typically form in association with an upper-level short wave. Other conditions favorable for development include enhanced cross-mountain upper-level westerly flow, enhanced baroclinicity, and reduced static stability (e.g., Hovanec and Horn 1975; Carlson 1991). The presence of a capping inversion or lid (associated with the upper-level advection of warm dry air from the desert southwest) above a region of low-level southerly flow bringing moisture from the Gulf can further act to enhance the latent instability and produce a large-scale environment favorable for the development of severe storms (Carlson 1991). The development of the low-level southerly wind anomalies (directed toward the low pressure anomalies; cf. Figs. 12c and 13c) and enhanced precipitation underneath the region of strong westerlies is consistent with the notion of coupled jet streaks first proposed by Uccellini and Johnson (1979). The proposed coupling mechanism is such that the LLJ is induced via the ageostrophic indirect circulations generated in the upper-level jet exit and entrance regions.

c. Synoptic composites

Figure 14 shows the evolution of the synoptic low-level moisture transport, anomalous surface pressure and precipitation anomalies for −2, −1, 0, +1 days relative to the time of maximum northward low-level moisture flux at Ft. Worth. The precipitation anomalies from the assimilation are again quite similar to the observed anomalies (Fig. 15). The main differences are generally more intense negative anomalies in the assimilation and a somewhat weaker positive anomaly at +1 day in the observations. At lag −2 days (Fig. 14a) the surface pressure anomalies show a weak midcontinental high pressure anomaly sandwiched between two low pressure anomalies on either coast (centered near 50°–55°N). After 1 day (Fig. 14b), the western low pressure anomaly has strengthened (anomalies less than −3.5 mb) and moved southward and slightly east (42°N, 102°W). At this time a region of southerly moisture flux on the leading edge of the surface low pressure anomaly extends from the Gulf of Mexico into the upper Midwest. The primary precipitation anomalies during this time are negative in association with an extensive but weak area of positive pressure anomalies covering much of the eastern United States. During the next 2 days (Figs. 14c and 14d) the main low pressure anomaly moves eastward and widespread positive precipitation anomalies develop along the leading edge in association with strong southwesterly moisture transport. Relatively large positive precipitation anomalies eventually extend from Arkansas to the Great Lakes (Fig. 14d). During the course of the evolution, the moisture flux from the Gulf evolves from primarily southerly over the central United States (Figs. 14a and 14b), to southwesterly, extending over much of the eastern half of the continent (Fig. 14d).

Figure 15 shows the corresponding evolution of the midtropospheric (σ = 0.5) and low-level (σ = 0.97) wind anomalies on the synoptic timescales. The midtropospheric wind anomalies at a lag of −2 day (Fig. 15a) show evidence of a zonally oriented wave train consisting of cyclonic anomalies on either coast and a relatively weak midcontinental anticyclonic anomaly. Over the next three days the wave system moves eastward and the cyclonic anomaly originally situated over the West Coast strengthens and moves eastward over the western (Figs. 15b and c) and then central (Fig. 15d) United States. At low levels, southerly wind anomalies begin to develop ahead of the anomalous upper-level cyclone (Fig. 15b). In contrast with the subsynoptic composite, the low-level southerly wind anomalies are nearly parallel to the upper-wind and surface pressure anomalies, and penetrate much farther north (well into Canada). At the same time north/northwesterly anomalies develop to the west directly underneath the upper-level cyclonic anomalies (Fig. 15c). At 0 lag, the low-level southerly wind anomalies extend from the Gulf Coast into Wisconsin and clash with northerly wind anomalies along a line extending from Texas to the upper plains. One day later (lag +1, Fig. 15d), the northerly anomalies dominate much of the Great Plains and upper Midwest, while the southerlies have been pushed east and south along a line stretching from Lousiana to Michigan.

The above composites appear to capture the canonical propagation and development of a midlatitude synoptic-scale storm system. The system, which was already established offshore at a lag of −4 day (not shown) intensifies as it crosses the Rocky Mountains and taps moisture from the Gulf of Mexico. Very similar warm-season composites have been obtained by Chen et al. (1996) employing a subset (1985–89) of the data used in the present study and a broader-scale index computed from the area-mean divergence of water vapor over the central and eastern part of the continent. That study showed the systems are reminiscent of the classic extratropical cyclonic model with low (upper) level convergence (divergence) centers ahead of the trough coupled by the upward branch of the divergent circulation, which is in part maintained by diabatic heating.

d. Supersynoptic composites

Figure 16 shows the evolution of the supersynoptic low-level moisture transport, anomalous surface pressure and precipitation anomalies for −3, −1, 0, +2 days relative to the time of maximum northward low-level moisture flux at Ft. Worth. The precipitation anomalies are generally in agreement with the observed (Fig. 17), though at lag 0 the observed positive anomalies have a different orientation. The main feature at this timescale is a low pressure anomaly over the northwest (Fig. 16a), which strengthens and moves southward and eastward (Fig. 16b), eventually covering much of the upper Midwest (Fig. 16c). Southerly moisture transport develops on the eastern flank of the cyclonic anomaly, extending at 0 lag (Fig. 16c), over much of the central and northeastern United States. Positive precipitation anomalies extend over much of the northern Great Plains at this time. Negative precipitation anomalies occur in the region of positive surface pressure anomalies, which move from the upper Midwest (Fig. 16a) to the southeast (Fig. 16b) and eventually off the coast (Figs. 16c and 16d). At a lag of 2 days (Fig. 16d), the surface pressure has weakened and is situated over the Great Lakes region just to the north of a weak positive pressure anomaly. At the same time an extensive positive pressure anomaly has developed over the western half of the country (Fig. 16d).

Figure 17 shows the corresponding evolution of the midtropospheric (σ = 0.5) and low-level (σ = 0.97) wind anomalies for the supersynoptic timescales. The primary feature of the midtropospheric wind anomalies is a cyclonic anomaly that, during the course of three days, moves from the West Coast (Fig. 17a) onto the western half of the continent (Figs. 17b and 17c) where it strengthens extending southward well into Mexico. At a lag of 0 day, the cyclonic wind anomaly encompasses the entire western and central United States and northern Mexico (Fig 17c). Low-level southerly wind anomalies develop on the leading edge of the upper-cyclonic anomaly (extending into Minnesota and Wisconsin) accompanied along the southeastern flank by low-level westerly/northwesterly winds (Figs. 17b and 17c). The cyclonic anomaly then moves over the central United states where it weakens (Fig. 17d).

The evolution and structure of the mid- and low-level winds on these (8–16 day) timescales is similar to those of the synoptic-scale composite with, however, somewhat larger zonal scales (e.g., cf. the surface pressure anomalies in Figs. 14c and 16c) and spatially more diffuse southerly moisture flux and precipitation anomalies. These systems also appear to get more “locked in” to the orography in that the anomalous upper-level cyclonic flow over the western United States remains quasi-stationary for more than three days (Figs. 17a–c). A qualitative inspection of the events shows that 20% of the supersynoptic cases are associated with cutoff lows and a few (6%) are clearly associated with blocking events.

4. Discussion and conclusions

This study examined the variations in the moisture entering the continental United States from the Gulf of Mexico, with a particular emphasis on the role of low-level southerly jets and the impact on the continental warm-season precipitation. The results are based on 9 yr of reanalysis data generated by the DAO at Goddard Space Flight Center employing the GEOS-1 DAS. As with any assimilated data, some caution must be used in interpreting the results since model deficiencies and/or poor observational coverage can impact the quality of the data. In the current study, precipitation anomalies from the assimilation are verified against gridded station observations (Higgins et al. 1997). The low-level moisture fluxes, while less model-dependent than precipitation, are only weakly constrained by the observations (especially the diurnal cycle) and may therefore also be impacted by model deficiencies. Recent studies suggest, however, that in spite of substantial bias (Higgins et al. 1996b) the variability in the moisture fluxes appears to be quite reasonable (e.g., Min and Schubert 1997).

It was shown that the moisture flux variations at the primary influx region, while confined to a rather narrow longitude band (between about 92.5° and 102.5°W; referred to here as the Texas Corridor or TC), are closely linked with a number of well-defined subcontinental and larger-scale phenomena, which vary on timescales ranging from diurnal to biweekly. The diurnal variation of the TC moisture flux is part of a regular nighttime acceleration of the boundary layer winds over northern Mexico and the Great Plains associated with the development of the Great Plains low-level jet (e.g., Bonner 1968; Helfand and Schubert 1995). The diurnal variation occurs throughout the spring and summer seasons examined here (May–August) with some tendency for a preference for the two warmest summer months. The composite nighttime anomalies are characterized by a strong southerly moisture flux covering northeast Mexico and the southern Great Plains, and enhanced boundary layer convergence and precipitation over much of the upper Great Plains. The large-scale environment during which the diurnally varying northward flux is most intense consists of an anomalous surface low over the Great Plains and mid- and upper-tropospheric cyclonic/anticyclonic wind anomalies (with the leading edge of the cyclonic anomaly over the Great Plains) spanning the continent. This reflects the predilection for the enhanced southerly flux at diurnal timescales to occur during periods when midlatitude weather systems are positioned to generate enhanced southerly flux over the Great Plains. The mid- and upper-level wind anomalies also have a substantial diurnal component, which is consistent with previous observational evidence of atmospheric tidal motion documented by Wallace and Hartranft (1969).

On subsynoptic (2–4 day) timescales the TC moisture flux variations are associated with the development and evolution of a warm-season lee cyclone. These systems, which are most prevalent during the early part of the warm season (May and June) form over the central Great Plains in association with an upper-level shortwave in the presence of enhanced upper-tropospheric cross-mountain westerly flow and baroclinicity. A low-level southerly wind maximum or jet develops underneath and perpendicular to the advancing edge of enhanced midtropospheric westerlies. These characteristics as well as reduced static stability are well-known conditions favoring severe weather outbreaks (e.g., Newton 1967; Hovanec and Horn 1975; Carlson 1991). The developing anomalous southerly moisture flux, while initially limited in its northward extent to the southern Great Plains, clashes with a deep intrusion of anomalous northerly low-level winds, resulting in enhanced precipitation beginning over eastern Colorado and eventually stretching from Texas to the Great Lakes.

The synoptic timescale (4–8 day) composites capture the propagation and intensification of a warm-season midlatitude cyclone. This system, which also occurs preferentially during May and June, develops offshore and intensifies as it crosses the Rocky Mountains and taps moisture from the Gulf of Mexico. Similar warm-season composites have been obtained by Chen et al. (1996) employing a subset (1985–89) of the data used in the present study, and Chen and Kpaeyeh (1993) employing NCEP operational analyses. The latter used the strength of the 850-mb meridional wind over the Great Plains to show a similar coupling between the LLJ and upper-level circulation. Compared with the subsynoptic timescales, the upper-level system is meridionally and vertically more extensive. The anomalous low-level southerly wind maximum (or LLJ) develops in phase with and parallel to the winds on the lee side of the middle-level trough. At the time of maximum TC moisture influx, the central United States is dominated by an anomalous surface low and southerly moisture flux extends across the Midwest into Canada. The associated precipitation anomalies in the central and eastern United States move with the propagating system with reduced rainfall occurring over the anomalous surface high, and enhanced rainfall occurring over the anomalous surface low. On longer timescales (8–16 day) the variations in the TC moisture transport are tied to slower eastward moving systems. The evolution and structure of the mid- and low-level winds is similar to those of the synoptic-scale composite with, however, somewhat larger zonal scales of motion and spatially more diffuse southerly moisture flux and precipitation anomalies.

The physical distinctions between the three types of synoptic systems summarized above is somewhat unclear. Understanding these differences is complicated by the fact that there is some overlap in the time periods when the systems are active. For example, about one-third (one-fourth) of the days of strong southerly moisture flux at the subsynoptic timescales coincide with periods when there is also at least a moderate southerly (northerly) moisture flux at the synoptic timescales. As already mentioned, the differences in the composites are partly a matter of scale, with the lower-frequency systems having somewhat larger horizontal scales and greater vertical extent. Further differences are associated with the structure of the associated LLJs (see below), and with the time evolution of the systems. In particular, the subsynoptic composite suggests a development on the lee side of the Rocky Mountains while the synoptic system is already well developed several days before it reaches the central United States, where it then experiences further growth as it taps moisture from the Gulf. The evolution of the supersynoptic systems suggests a greater coupling with the orography than the synoptic systems, in that the anomalous upper-level flow over the Rocky Mountains remains quasi-stationary for several days. We note that 20% of the supersynoptic cases are associated with cutoff lows and a few (6%) are clearly associated with blocking events.

The various low-level southerly wind maxima or LLJs appear to play distinctly different roles in the development of continental warm-season precipitation anomalies. While the GPLLJ is the dominant contributor to the time-mean moisture influx, it appears to contribute primarily indirectly to the occurrence of the Great Plains nocturnal rainfall as a convergence and lifting mechanism: the relatively short (diurnal) timescale allows insufficient time for a substantial northward transport of water to the precipitating region. The strong contribution to the time-mean moisture flux reflects the other key role of the GPLLJ to precondition the boundary layer for deep convection by night after night transporting moisture from the Gulf to the Great Plains. On subsynoptic timescales the associated LLJ appears to play a pivotal role in the initial development of precipitation over the southern Great Plains. Further development of the upper-level system and heavy precipitation is, however, marked less by a deep northward penetration of Gulf moisture, than a deep southward intrusion of northerly wind anomalies. Only on the synoptic and longer timescales does the precipitation appear to be directly tied to vertically extensive and deep northward intrusions of moist Gulf air.

The relationship between the GPLLJ and the other low-level southerly wind maxima appears to be primarily one in which the lower-frequency variations provide an enhanced or reduced basic southerly flow upon which the GPLLJ develops. Thus, depending on the geographic phasing of these systems with respect to the Great Plains, they alternatively suppress and enhance the occurrence of the jet. The mechanism for the development of the GPLLJ appears to be substantially different from that involved in the development of the low-level subsynoptic wind maxima, which in turn is different from the development of the low-level longer timescale wind maxima. The GPLLJ is primarily a boundary layer phenomena. While the exact nature of the mechanism generating the jet is still unclear, terrain effects and the diurnal variation in eddy viscosity appear to play a major role (e.g., Bonner and Paegle 1970; Stensrud 1996). On subsynoptic timescales, the low-level southerly wind anomalies develop underneath and perpendicular to the “exit” region of anomalous middle tropospheric westerlies in a manner consistent with the mechanism of coupled jet streaks first proposed by Uccellini and Johnson (1979). The LLJ in that case has a substantial ageostrophic component and appears to form as a result of the ageostrophic indirect circulations generated in the upper-level jet exit region. On the synoptic and longer timescales, the associated low-level southerly wind anomalies have a strong geostrophic component developing in phase with and parallel to the mid- and upper-level flow along the leading edge of the trough. These jets appear to be similar to those identified by Browning and Pardoe (1973) in association with midlatitude cold fronts.

Acknowledgments

This work was supported by NASA’s Earth Observing Systems projects on 4D Data Assimilation and Computing.

REFERENCES

  • Benton, G. S., and M. A. Estoque, 1954: Water-vapor transfer over the North American continent. J. Meteor.,11, 462–477.

  • Blackadar, A. K., 1957: Boundary-layer wind maxima and their significance for the growth of nocturnal inversions. Bull. Amer. Meteor. Soc.,38, 283–290.

  • Blackmon, M. L., V.-H. Lee, and J. M. Wallace, 1984: Horizontal structure of 500 mb height fluctuations with long, intermediate and short time scales. J. Atmos. Sci.,41, 961–979.

  • Bloom, S. C., L. L. Takacs, A. M. da Silva, and D. Ledvina, 1996: Data assimilation using incremental analysis updates. Mon. Wea. Rev.,124, 1256–1271.

  • Bonner, W. D., 1966: Case study of thunderstorm activity relation to the low-level jet. Mon. Wea. Rev.,94, 167–178.

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

  • ——, and J. Paegle, 1970: Diurnal variations in the boundary-layer winds over the south central United States in summer. Mon. Wea. Rev.,98, 735–744.

  • Bowen, B. M., 1996: Rainfall and climate variation over a sloping New Mexico plateau during the North American monsoon. J. Climate,9, 3432–3442.

  • Browning, K. A., and C. W. Pardoe, 1973: Structure of low-level jet streams ahead of midlatitude cold fronts. Quart. J. Roy. Meteor. Soc.,99, 619–638.

  • Carlson, T. N., 1991: Mid-Latitude Weather Systems. Harper Collins Academic, 507 pp.

  • Chang, F.-C., and J. M. Wallace, 1987: Meteorological conditions during heat waves and droughts in the United States Great Plains. Mon. Wea. Rev.,115, 1253–1269.

  • Chen, T.-C., and J. A. Kpaeyeh, 1993: The synoptic-scale environment associated with the low-level jet of the Great Plains. Mon. Wea. Rev.,121, 416–420.

  • ——, M.-C. Yen, and S. Schubert, 1996: Hydrological processes associated with cyclonic systems over the United States. Bull. Amer. Meteor. Soc.,77, 1557–1567.

  • Harrold, T. W., 1973: Mechanisms influencing the distribution of precipitation within baroclinic disturbances. Quart. J. Roy. Meteor. Soc.,99, 232–251.

  • Harshvardhan, R. Davies, D. A. Randall, and T. G. Corsetti, 1987: A fast radiation parameterization for atmospheric circulation models. J. Geophys. Res.,92, 1009–1016.

  • Helfand, H. M., and J. C. Labraga, 1988: Design of a non-singular level 2.5 second-order closure model for the prediction of atmospheric turbulence. J. Atmos. Sci.,45, 113–132.

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

  • Higgins, R. W., J. E. Janowiak, and Y. Yao, 1996a: A gridded hourly precipitation data base for the United States (1963–93). Atlas No. 1., NCEP/Climate Prediction Center, 47 pp.

  • ——, K. C. Mo, and S. D. Schubert, 1996b: The moisture budget of the central United States in spring as evaluated in the NCEP/NCAR and the NASA/DAO reanalyses. Mon. Wea. Rev.,124, 939–963.

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

  • Hoecker, W. J., 1963: Three southerly low-level jet systems delineated by the Weather Bureau special pibal network of 1961. Mon. Wea. Rev.,91, 573–582.

  • Hovanec, R. D., and L. H. Horn, 1975: Static stability and the 300 mb isotach field in the Colorado cyclonetic area. Mon. Wea. Rev.,103, 628–638.

  • Lindzen, R. S., 1967: Thermally driven diurnal tide in the atmosphere. Quart. J. Roy. Meteor. Soc.,93, 18–42.

  • Lyon, B., and R. Dole, 1995: A diagnostic comparison of the 1980 and 1988 U.S. summer heat wave–droughts. J. Climate,8, 1658–1675.

  • Maddox, R. A., 1983: Large-scale meteorological conditions associated with midlatitude, mesoscale convective complexes. Mon. Wea. Rev.,111, 1475–1493.

  • Meyers, S. D., B. G. Kelly, and J. J. O’Brien, 1993: An introduction to wavelet analysis in oceanography and meteorology: With application to the dispersion of Yanai waves. Mon. Wea. Rev.,121, 2858–2878.

  • Min, W., and S. Schubert, 1997: The climate signal in regional moisture fluxes: A comparison of three global data assimilation products. J. Climate,10, 2623–2642.

  • Molod, A., H. M. Helfand, and L. L. Takacs, 1996: The climatology of parameterized physical processes in the GEOS-1 GCM and their impact on the GEOS-1 Data Assimilation System. J. Climate,9, 764–785.

  • Moorthi, S., and M. J. Suarez, 1992: Relaxed Arakawa–Schubert: A parameterization of moist convection for general circulation models. Mon. Wea. Rev.,120, 978–1002.

  • Namias, J., 1955: Some meteorological aspects of drought with special reference to the summers of 1952–1954 over the United States. Mon. Wea. Rev.,83, 199–205.

  • ——, 1982: Anatomy of Great Plains protracted heat waves (especially the 1980 U.S. summer drought). Mon. Wea. Rev.,110, 824–838.

  • Newton, C. W., 1967: Severe convective storms. Advances in Geophysics, Vol. 12, Academic Press, 257–308.

  • Oglesby, R. J., 1991: Springtime soil moisture, natural climatic variability, and North American drought as simulated by the NCAR Community Climate Model 1. J. Climate,4, 890–897.

  • ——, K. A. Maasch, and B. Saltzman, 1989: Glacial meltwater cooling of the Gulf of Mexico: GCM implications for Holocene and present-day climate. Climate Dyn.,3, 115–133.

  • Paegle, J., 1984: Topographically bound low-level circulations. Riv. Meteor. Aeronaut.,44, 113–125.

  • Pfaendtner, J., S. Bloom, D. Lamich, M. Seablom, M. Sienkiewicz, J. Stobie, and A. da Silva, 1995: Documentation of the Goddard Earth Observing System (GEOS) Data Assimilation System-Version 1. NASA Tech. Memo. 104606, Vol. 4, 44 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • Pitchford, K. L., and J. London, 1962: The low-level jet as related to nocturnal thunderstorms over midwest United States. J. Appl. Meteor.,1, 43–47.

  • Rasmusson, E. M., 1967: Atmospheric water vapor transport and the water balance of North America: Part I. Characteristics of the water vapor flux field. Mon. Wea. Rev.,95, 403–426.

  • Reiter, E. R., 1969: Tropopause circulation and jet streams. World Survey of Climatology, Climate of the Free Atmosphere, D. F. Rex, Ed., Vol. 4, Elsevier, 85–193.

  • Roads, J. O., S.-C. Chen, A. K. Guetter, and K.-P. Georgakakos, 1994:Large-scale aspects of the United States hydrological cycle. Bull. Amer. Meteor. Soc.,75, 1589–1610.

  • Roebber, P. J., 1984: Statistical analysis and updated climatology of explosive cyclones. Mon. Wea. Rev.,112, 1577–1589.

  • Schemm, J.-K., S. Schubert, J. Terry, S. Bloom, and Y. Sud, 1992: Estimates of monthly mean soil moisture for 1979–89. NASA Tech. Memo. 104571, 252 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • Schubert, S., and Y. Chang, 1996: An objective method for inferring sources of model error. Mon. Wea. Rev.,124, 325–340.

  • ——, J. Pfaendtner, and R. Rood, 1993: An assimilated dataset for earth science applications. Bull. Amer. Meteor. Soc.,74, 2331–2342.

  • ——, and Coauthors, 1995: A multiyear assimilation with the GEOS-1 system: Overview and results. NASA Tech. Memo. 104606, Vol. 6, 183 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • Stensrud, D. J., 1996: Importance of low-level jets to climate: A review. J. Climate,9, 1698–1711.

  • Suarez, M. J., and L. L. Takacs, 1995: Documentation of the Aries-GEOS dynamical core: Version 2. NASA Tech. Memo. 104606, Vol. 5, 45 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • Sud, Y., and A. Molod, 1988: The roles of dry convection, cloud-radiation feedback processes and the influence of recent improvements in the parameterization of convection in the GLA GCM. Mon. Wea. Rev.,116, 2366–2387.

  • Takacs, L. L., and M. J. Suarez, 1996: Dynamical aspects of climate simulations using the GEOS General Circulation Model. NASA Tech. Memo. 104606, Vol. 10, 56 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • ——, A. Molod, and T. Wang, 1994: Goddard Earth Observing System (GEOS) General Circulation Model (GCM) Version 1. NASA Tech. Memo. 104606, Vol. 1, 100 pp. [Available from Goddard Space Flight Center, Greenbelt, MD 20771.].

  • Uccellini, L. W., and D. R. Johnson, 1979: The coupling of upper and lower tropospheric jet streams and implications for the development of severe convective storms. Mon. Wea. Rev.,107, 682–703.

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

  • ——, and F. R. Hartranft, 1969: Diurnal wind variations, surface to 30 kilometers. Mon. Wea. Rev.,97, 446–455.

  • Weng, H., and K.-M. Lau, 1994: Wavelets, period doubling, and time-frequency localization with application to organization of convection over the tropical western Pacific. J. Atmos. Sci.,51, 2523–2541.

  • Winkler, J. A., B. R. Skeeter, and P. D. Yamamoto, 1988: Seasonal variations in the diurnal characteristics of heavy hourly precipitation across the United States. Mon. Wea. Rev.,116, 1641–1658.

APPENDIX

The GEOS-1 DAS

The main components of the GEOS-1 DAS are the GEOS-1 atmospheric general circulation model (AGCM; Takacs et al. 1994) and a 3D, multivariate optimal interpolation (OI) scheme (Pfaendtner et al. 1995). The OI scheme is multivariate in geopotential height and winds and employs a damped cosine function for the horizontal correlation of model prediction error. The height-wind cross-correlation model is geostrophic and scaled to zero at the equator. The multivariate surface analysis scheme over the oceans adopts an Ekman balance for the pressure-wind analysis. The moisture analysis for mixing ratio employs only rawinsonde moisture data.

All gridpoint analyses are done using up to 75 nearby observations from within a circular data selection cylinder of 1600-km radius. The assimilation is carried out at a horizontal resolution of 2° lat × 2.5° long at 14 upper-air pressure levels (20, 30, 50, 70, 100, 150, 200, 250, 300, 400, 500, 700, 850, 1000 mb) and at sea level. The analysis increments are computed every 6 h using observations from a ±3-h data window centered on the analysis times (0000, 0600, 1200, and 1800 UTC). For the global sea level pressure and near-surface wind analysis over the oceans, data from surface land synoptic reports (sea level pressure only), ships, and buoys are used. The upper-air analyses of height, wind, and moisture incorporate the data from rawinsondes, dropwindsondes, rocketsondes, aircraft winds, cloud-tracked winds, and thicknesses from the historical TIROS-N Operational Verticle Sounder soundings produced by the National Oceanic and Atmospheric Administration/National Environmental Satellite Data and Information Service. The satellite heights are computed using a reference level that depends on the analyzed sea level pressure.

The assimilation system does not include an initialization scheme and relies on the damping properties of a Matsuno time-differencing scheme to control initial imbalances generated by the insertion of observations. However, the initial imbalances and spinup have been greatly reduced over earlier versions by the introduction of an incremental analysis update (IAU) procedure (Bloom et al. 1996). In the IAU procedure, standard OI analysis increments are computed at the analysis times (0000, 0600, 1200, 1800 UTC). The increments are then inserted gradually into the AGCM by rerunning the forecast and adding a fraction of the increment at each model time step. Over the 6-h period centered at the analysis time, the full effect of the increment is realized. The assimilation thus effectively consists of a continuous AGCM forecast with additional heat, momentum, moisture, and mass source terms updated every 6 h from observations.

The GEOS-1 AGCM is a gridpoint model employing the Aries/GEOS dynamical core described in Suarez and Takacs (1995). The climate characteristics of the GCM are documented in Molod et al. (1996) and Takacs and Suarez (1996). The infrared and solar radiation parameterizations follow closely those described by Harshvardhan et al. (1987). The penetrative convection originating in the boundary layer is parameterized using the Relaxed Arakawa–Schubert (RAS) scheme (Moorthi and Suarez 1992). The AGCM also includes a parameterization that models the evaporation of falling convective rain as described in Sud and Molod (1988). Negative values of specific humidity produced by the finite-differenced advection are filled by borrowing from below. The planetary boundary layer uses the second-order closure model of Helfand and Labraga (1988). The performance of the AGCM in simulating the GPLLJ is documented in Helfand and Schubert (1995). For the assimilation, GEOS-1 was integrated at a resolution of 2° lat × 2.5° lon with 20 sigma levels. Since GEOS-1 is run without a land surface model, soil moisture is computed offline based on a simple bucket model using monthly mean observed surface air temperature and precipitation (Schemm et al. 1992). The snow line and surface albedo are prescribed and vary with the season. The sea surface temperature is updated according to the observed monthly mean values provided by the Climate Analysis Center at NCEP and the Center for Ocean, Land and Atmosphere (COLA) at the University of Maryland.

Fig. 1.
Fig. 1.

The time-mean (May–June 1985–93) moisture transport in (a) the lower troposphere (surface to σ = 0.84) and (b) the middle and upper troposphere σ < 0.84. Units are m s−1 g kg−1.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 2.
Fig. 2.

Time series of northward moisture transport (υq) profile at 32°N, 97.5°W. Vertical scale is approximate pressure level assuming 1000-mb surface pressure. Units are m s−1 g kg−1.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 3.
Fig. 3.

(a) Power spectrum of υ (m s−1)2, (b) power spectrum of q (g kg−1)2, (c) power spectrum of υq (m s−1 g kg−1)2, and (d) q2 times the power spectrum of υ (m s−1 g kg−1)2 at 32°N, 97.5°W, σ = 0.97 for May–June 1985–93.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 4.
Fig. 4.

(a) Profile of the standard deviation of υq (solid line) and approximate form q2 Var(υ) (dashed line) at 32°N, 97.5°W. Units are m s−1 g kg−1. (b) The vertical correlation with σ = 0.97 as a reference level at 32°N, 97.5°W.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 5.
Fig. 5.

The standard deviation at σ = 0.97 of (a) υq and (b) uq for May–June 1985–93. Units are m s−1 g kg−1. The two components of the correlation (displayed as a vector) of the moisture flux (uq, υq) at σ = 0.97 with (c) υq at a base point at 32°N, 97.5°W and with (d) uq at base point at 38°N, 77.5°W for May/June 1985–93.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 6.
Fig. 6.

Wavelet analysis of υq at 32°N, 97.5°W shown for the period May–August of 1993. The top panel shows the time series of υq. The bottom panel shows the real part of the wavelet transform for each frequency. Units are (m s−1 g kg−1)2. The wavelet analysis was performed on the entire 9-yr record.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 7.
Fig. 7.

Longitude–pressure cross sections at 34°N of the composite (time lag = 0, see text) mean horizontal wind anomalies for (a) diurnal, (b) subsynoptic (2 < τ < 4 day), (c) synoptic (4 < τ < 8 days), and (d) supersynoptic (8 < τ < 16 day) timescales. Vectors pointing down denote northerly wind anomalies. Shading shows the full composite northward wind component for each frequency band. Units: m s−1.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 8.
Fig. 8.

Same as Fig. 7 except for latitude–pressure cross sections at 97.5°W. Vectors pointing down denote westerly wind anomalies.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 9.
Fig. 9.

Same as Fig. 7 except for profiles at 32°N, 97.5°W at different time lags. Vectors pointing down (to the right) are northerly (westerly) wind anomalies. The absisca in panel (a) is given in hours, and in days for the other panels.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 10.
Fig. 10.

Composite mean diurnal surface pressure anomalies (contours, mb), moisture flux at σ = 0.97 (m s−1 g kg−1), and observed precipitation anomalies (shading, mm day−1) for lags (a) −12 h, (b) −6 h, (c) 0 h, and (d) + 6 h. Moisture flux vectors with magnitudes less than 20 m s−1 g kg−1 are omitted.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 11.
Fig. 11.

Composite mean diurnal wind anomalies at σ = 0.97 (dark vectors) and σ = 0.5 (light vectors) for lags (a) −12 h, (b) −6 h, c) 0 h, and d) + 6 h. Units are m s−1. Wind vectors at σ = 0.97 with magnitudes less than 1 m s−1 are omitted.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 12.
Fig. 12.

Composite mean subsynoptic surface pressure anomalies (contours, mb), moisture flux at σ = 0.97 (m s−1 g kg−1), and GEOS-1 precipitation anomalies (shading, mm day−1) for lags (a) −1 days, (b) −0.5 days, (c) 0 days, and (d) + 0.5 days. Moisture flux vectors with magnitudes less than 20 m s−1 g kg−1 are omitted.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 13.
Fig. 13.

Composite mean subsynoptic wind anomalies at σ = 0.97 (dark vectors), σ = 0.5 (light vectors), and observed precipitation anomalies (shading, mm day−1) for lags (a) −1 day, (b) −0.5 day, (c) 0 day, and (d) + 0.5 day. Units are m s−1. In (b) the light contours indicate the speed of the wind anomalies at σ = 0.5. Wind vectors at σ = 0.97 with magnitudes less than 1 m s−1 are omitted.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 14.
Fig. 14.

Composite mean synoptic surface pressure anomalies (contours, mb), moisture flux at σ = 0.97 (m s−1 g kg−1), and GEOS-1 precipitation anomalies (shading, mm day−1) for lags (a) −2 day, (b) −1 day, (c) 0 day, and (d) + 1 day. Moisture flux vectors with magnitudes less than 20 m s−1 g kg−1 are omitted.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 15.
Fig. 15.

Composite mean synoptic wind anomalies at σ = 0.97 (dark vectors), σ = 0.5 (light vectors), and observed precipitation anomalies (shading, mm day−1) for lags (a) −2 day, (b) −1 day, (c) 0 day, and (d) +1 day. Units are m s−1. Wind vectors at σ = 0.97 with magnitudes less than 1 m s−1 are omitted.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 16.
Fig. 16.

Composite mean supersynoptic surface pressure anomalies (contours, mb), moisture flux at σ = 0.97 (m s−1 g kg−1), and GEOS-1 precipitation anomalies (shading, mm day−1) for lags (a) −3 day, (b) −1 day, (c) 0 day, and (d) + 2 day. Moisture flux vectors with magnitudes less than 20 m s−1 g kg−1 are omitted.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Fig. 17.
Fig. 17.

Composite mean supersynoptic wind anomalies at σ = 0.97 (dark vectors) and σ = 0.5 (light vectors), and observed precipitation anomalies (shading, mm day−1) for lags (a) −3 day, (b) −1 day, (c) 0 day, and (d) + 2 day. Units are m s−1. Wind vectors at σ = 0.97 with magnitudes less than 1 m s−1 are omitted.

Citation: Journal of Climate 11, 10; 10.1175/1520-0442(1998)011<2530:SVIWSM>2.0.CO;2

Table 1.

The number of cases for each composite by month and time of day for 1985–93. The numbers in parentheses are the counts expressed as a percentage of the total counts within each frequency band.

Table 1.
Table 2.

Counts of the number of cases in the diurnal composite according to the phase of the lower-frequency composites. The second column indicates the subset of the diurnal cases, which occurred when the subsynoptic moisture flux (exceeding one-half standard deviation) was in a southerly/northerly phase. Similarly for the synoptic and supersynoptic fluxes in columns three and four, respectively.

Table 2.

1

We note that the southeasterly flow from the Gulf of Mexico is also known to impact the southwestern United States as part of the North American monsoon system (e.g., Bowen 1996).

2

The vertical integral for this layer extends to the model top at 10 hPa; however, radiosonde moisture observations are only assimilated up to 300 hPa.

Save