• Amador, J. A., 1998: A climatic feature of the tropical Americas: The trade wind easterly jet. Top. Meteor. Oceanogr., 5 , 2,. 113.

  • Amador, J. A., and Magaña V. , 1999: Dynamics of the low level jet over the Caribbean Sea. Preprints, Third Conf. on Hurricanes and Tropical Meteorology, Dallas, TX, Amer. Meteor. Soc., 868–869.

  • Berbery, E. H., 2001: Mesoscale moisture analysis of the North American monsoon. J. Climate, 14 , 121137.

  • Berbery, E. H., and Rasmusson E. M. , 1999: Mississippi moisture budgets on regional scales. Mon. Wea. Rev., 127 , 26542673.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Berbery, E. H., and Fox-Rabinovitz M. S. , 2003: Multiscale diagnosis of the North American monsoon system using a variable resolution GCM. J. Climate, 16 , 16291647.

    • Search Google Scholar
    • Export Citation
  • Berbery, E. H., Rasmusson E. M. , and Mitchell K. E. , 1996: Studies of North American continental scale hydrology using Eta Model forecast products. J. Geophys. Res., 101 , 73057319.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chu, L. S., Chang A. T. C. , and Janowiak J. , 1993: Comparison of monthly mean rates derived from GPI and SSM/I using probability distribution functions. J. Appl. Meteor., 32 , 323334.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cressman, G. P., 1959: An operational objective analysis system. Mon. Wea. Rev., 87 , 367374.

  • Dettinger, M. D., Cayan D. R. , Diaz H. F. , and Meko D. M. , 1998: North–south precipitation patterns in western North America on interannual and decadal time scales. J. Climate, 11 , 30953111.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Douglas, M. W., 1995: The summertime low-level jet over the Gulf of California. Mon. Wea. Rev., 123 , 23342347.

  • Douglas, M. W., Maddox R. A. , Howard K. , and Reyes S. , 1993: The Mexican monsoon. J. Climate, 6 , 16651677.

  • Douglas, M. W., Valdez-Manzanilla A. , and Cueto R. G. , 1998: Diurnal variation and horizontal extent of the low level jet over the northern Gulf of California. Mon. Wea. Rev., 126 , 20172025.

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

    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., Yao Y. , and Wang X. , 1997a: Influence of the North American monsoon system on the U.S. summer precipitation regime. J. Climate, 10 , 26002622.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., Yao Y. , Yarosh E. S. , Janowiak J. E. , and Mo K. C. , 1997b: Influence of the Great Plains low-level jet on summer precipitation and moisture transport over the central United States. J. Climate, 10 , 481507.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., Chen Y. , and Douglas A. V. , 1999: Interannual variability of the North American warm season precipitation regime. J. Climate, 12 , 653680.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., Shi W. , Yarosh E. , and Joyce R. , 2000: Improved United States Precipitation Quality Control System and Analysis. NCEP/Climate Prediction Center Atlas 7, U.S. Department of Commerce, 45 pp. [Available online at http://www.cpc.ncep.noaa.gov/research_papers/ncep_cpc_atlas/7/index.html.].

  • Hsu, C. F., and Wallace J. M. , 1976: The global distribution of the annual and semiannual cycles in precipitation. Mon. Wea. Rev., 104 , 10931101.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Janowiak, J. E., Xue P. , Morrissey M. L. , and Legates D. R. , 1995: An examination of the east Pacific ITCZ rainfall distribution. J. Climate, 8 , 28102823.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Joyce, R. J., Janowiak J. E. , Arkin P. A. , and Xie P. , 2004: CMORPH: A method that produces global precipitation estimates from passive microwave and infrared data at high spatial and temporal resolution. J. Hydrometeor., 5 , 487503.

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

  • Kanamitsu, M., Ebisuzaki W. , Wollen J. , Yang S. K. , Hnilo J. J. , Fiorino M. , and Potter G. L. , 2001: NCEP/DOE AMIP-II reanalysis R-2. Bull. Amer. Meteor. Soc., 82 , 16311643.

    • Search Google Scholar
    • Export Citation
  • Keables, M., 1989: A synoptic climatology of the bimodal precipitation distribution in the upper Midwest. J. Climate, 2 , 12891294.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kendrew, W. G., 1922: The Climates of the Continent. Clarendon Press, 308 pp.

  • Magaña, V., Amador J. A. , and Medina S. , 1999: The midsummer drought over Mexico and Central America. J. Climate, 12 , 15771588.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mesinger, F., and Coauthors, 2004: North American regional reanalysis. Preprints, 15th Symp. on Global Change and Climate Variations, Seattle, WA, Amer. Meteor. Soc., CD-ROM, P1.1.

  • Mitchell, K. E., and Coauthors, 2004: The multi-institution North American Land Data Assimilation System (NLDAS): Utilizing multiple GCIP products and partners in a continental distributed hydrological modeling system. J. Geophys. Res., 109 .D07S90, doi:10.1029/2003JD003823.

    • Search Google Scholar
    • Export Citation
  • Mo, K. C., and Higgins R. W. , 1996: Large scale atmospheric moisture transport as evaluated in the NCEP/NCAR and the NASA/DAO reanalysis. J. Climate, 9 , 15311545.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mo, K. C., and Higgins R. W. , 1998: Tropical influence on California precipitation. J. Climate, 11 , 412430.

  • Mo, K. C., and Berbery E. H. , 2004: The low level jets and the summer precipitation regimes over North America. J. Geophys. Res., 109 .D06117, doi:10.1029/2003JD004106.

    • Search Google Scholar
    • Export Citation
  • Peixoto, J. P., and Oort A. H. , 1992: Physics of Climate. American Institute of Physics, 520 pp.

  • Rasmusson, E. M., 1967: Atmospheric water vapor transport and the water balance of North America. I: Characteristics of the water vapor flux field. Mon. Wea. Rev., 95 , 403426.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rasmusson, E. M., 1968: Atmospheric water vapor transport and the water balance of North America. II: Large scale water balance investigations. Mon. Wea. Rev., 96 , 720734.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roads, J. O., Chen S. C. , Guetter A. K. , and Georgakakos P. K. , 1994: Large scale aspects of the United States hydrological cycle. Bull. Amer. Meteor. Soc., 75 , 15891610.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roads, J. O., Chen S. C. , Kamamitsu M. , and Juang H. M. , 1997: GDAS’s GCIP energy budgets. J. Atmos. Sci., 54 , 17761794.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roads, J. O., Kanamitsu M. , and Stewart R. , 2002: CSE water and energy budgets in the NCEP/DOE reanalysis 2. J. Hydrometeor., 3 , 227248.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schmitz, J. T., and Mullen S. L. , 1996: Water vapor transport associated with the summertime North American monsoon as depicted by ECMWF analyses. J. Climate, 9 , 16211634.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shafran, P., Woollen J. , Ebisuzaki W. , Shi W. , Fan Y. , Grumbine R. , and Fennessy M. , 2004: Observational data used for assimilation in the NCEP North American regional reanalysis. Preprints, 14th Conf. on Applied Climatology, Seattle, WA, Amer. Meteor. Soc., CD-ROM, 1.4.

  • Stensrud, D. J., Gall R. L. , Mullen S. L. , and Howard K. W. , 1995: Model climatology of the Mexico monsoon. J. Climate, 8 , 17751794.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stensrud, D. J., Gall R. L. , and Nordquist M. K. , 1997: Surge over the Gulf of California during the Mexico monsoon. Mon. Wea. Rev, 125 , 417437.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tang, W., and Liu W. T. , 1996: Objective interpolation of scatterometer winds. JPL Publication 96-19, Jet Propulsion Laboratory, Pasadena, CA, 16 pp.

  • Trenberth, K. E., and Guillemot C. J. , 1995: Evaluation of the atmospheric moisture budget as seen from analyses. J. Climate, 8 , 22252272.

    • Search Google Scholar
    • Export Citation
  • Xie, P. P., Janowiak J. E. , Arkin P. A. , Alder R. , Gruber A. , Ferraro R. , Huffmann G. J. , and Curtis S. , 2003: GPCP pentad precipitation analyses: An experimental dataset based on gauge observations and satellite estimates. J. Climate, 16 , 21972214.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery

    Monthly mean precipitation climatology from the RR for the period from 1979 to 2001 for (a) Jan, (b) Feb, (c) Mar, (d) Apr, (e) May, (f) Jun, (g) Jul, (h) Aug, (i) Sep, (j) Oct, (k) Nov, and (l) Dec. Contour intervals are 1, 2, 3, 4, 6, 8, and 10 mm day−1.

  • View in gallery

    Seasonal cycle from daily mean P climatology from the RR (dark line) and from the URD (crosses) for (a) the Pacific Northwest (40°–50°N, 118°–125°W), (b) California (30°–40°N, 118°–125°W, over land), (c) northern Mexico (25°–32°N, 100°–110°W), (d) Southeast (30°–33°N, 80°–90°W), (e) Midwest (43°–48°N, 80°–90°W), (f) northern plains (38°–48°N, 90°–100°W), (g) southern plains (32°–38°N, 90°–100°W), and (h) southern Mexico (17°–23°N, 95°–100°W). A 16-day running mean was applied to the time series before plotting.

  • View in gallery

    Same as in Fig. 1, but for the monthly mean difference between the RR and the observed precipitation from the URD averaged from 1979 to 2001. Contour interval is 0.5 mm day−1. Zero contours are omitted.

  • View in gallery

    (a) Mean P for Dec–Feb averaged from 1979 to 2001. Contour interval 2 mm day−1. Contours 0.5 and 1 mm day−1 are added. Values greater than 4 mm day−1 are shaded. (b) As in (a), but for E. (c) As in (a), but for mean PE. Values less than −2 mm day−1 are shaded. (d) Same as (c), but for the residual analysis increment term Winc

  • View in gallery

    (a) Mean vertically integrated moisture flux divergence for Dec–Feb (DJF) averaged from 1979 to 2001. Contour interval is 1 mm day−1. Values less than −4 mm day−1 are shaded. (b) The mean vertically integrated moisture flux (qu; ). The unit vector is 200 kg (ms)−1.

  • View in gallery

    (a)–(c) Same as Figs. 4a–c, but for the R1. (d) Same as Fig. 5a, but for R1.

  • View in gallery

    (a) Composite P difference between the eight wettest and eight driest months in California (30°–40°N, 118°–125°W over land) for Dec–Mar 1979–2001 from the RR. Contour interval is 2 mm day−1. Zero contours are omitted. Contours −1 and 1 mm day−1 are added. Areas with positive (negative) values that are statistically significant at the 5% level are shaded dark (light). (b) As in (a), but for the vertically integrated moisture flux and divergence. The unit vector is 200 kg (ms)−1.

  • View in gallery

    Time–longitude cross section of (a) P averaged from 32° to 36°N for the daily mean climatology from the RR. A 16-day running mean was applied to the field before plotting. Contour interval is 1 mm day−1. Values greater than 3 mm day−1 are shaded. (b) As in (a), but for E. (c) As in (a), but P averaged from 38° to 48°N. (d) Same as (a), but for the vertically integrated moisture flux convergence. Values greater than 1 mm day−1 are shaded. Contours of −0.5 mm day−1 are added. Zero contours are omitted.

  • View in gallery

    Same as Fig. 4, but for Jul–Sep averaged from 1979 to 2001.

  • View in gallery

    Same as Fig. 5, but for Jul–Sep.

  • View in gallery

    Same as Fig. 6, but for Jul–Sep.

  • View in gallery

    (a) Surface zonal winds averaged for Jun–Jul 2000 from the satellite QuikSCAT winds. Contour interval is 1 m s−1. Values less than −10 m s−1 are shaded. (b) Same as in (a), but for the RR. (c), (d) Same as in (a), (b), respectively, but for 2001.

  • View in gallery

    Monthly mean vertically integrated meridional moisture flux () for (a) Jun, (b) Jul, (c) Aug, and (d) Sep averaged from 1995 to 2000 from the RR. The contour interval is 30 kg (ms)−1. Values greater than 60 kg (ms)−1 [90 kg (ms)−1] are shaded light (dark). (e)–(h) Same as (a)–(d), but for the difference of () between the RR and the operational EDAS. Contour interval is 30 kg (ms)−1. Values greater than 60 kg (ms)−1 are shaded.

  • View in gallery

    (a) Vertical cross section of wind speed at 36°N, 97.5°W from the RR from 17 to 31 Aug 1994. Contour interval is 2 m s−1. Values greater than 8 m s−1 are shaded. (b) Vertical profile of the meridional wind at 36°N, 97.5°W from the RR for Jun–Aug (JJA) 1994 at each synoptic time. Contour interval is 1 m s−1. Positive values are shaded.

  • View in gallery

    (a) Time series of 3-hourly meridional winds at 975 hPa at a grid point near Puerto Peñasco (open circles). The dark line is the eight-point running mean. (b) Vertical profile of meridional wind (m s−1) at Puerto Peñasco for each synoptic time: 0000 UTC (1600 LT, dark squares), 0300 UTC (1900 LT, crosses), 0600 UTC (2200 LT, triangles), 0900 UTC (0100 LT, dark line), 1200 UTC (0400 LT, pluses), 1500 UTC (0700 LT, open circles), 1800 UTC (1000 LT, dark circles), and 2100 UTC (1300 LT, open squares) averaged from 31 Jul–14 Aug 1995.

  • View in gallery

    (a) Same as Fig. 15a, but from observations; (b) same as Fig. 15b, but from observations [reproduced from Fig. 3 of Douglas et al. (1998)].

  • View in gallery

    Vertical cross section of meridional moisture flux () at 30°N for the period 1998–2000 from the RR for (a) Jul, (b) Aug, and (c) Sep. Contour interval is 10 g kg−1 m s−1. Zero contours are omitted. Contours of −5 and 5 g kg−1 m s−1 are added. (d)–(f) Same as (a)–(c), but for the operational EDAS.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 158 152 12
PDF Downloads 94 91 4

Atmospheric Moisture Transport over the United States and Mexico as Evaluated in the NCEP Regional Reanalysis

View More View Less
  • 1 Climate Prediction Center, NOAA/NWS/NCEP, Camp Springs, Maryland
© Get Permissions
Full access

Abstract

The large-scale atmospheric hydrologic cycle over the United States and Mexico derived from the 23-yr NCEP regional reanalysis (RR) was evaluated by comparing the RR products with satellite estimates, independent sounding data, and the operational Eta Model three-dimensional variational data assimilation (3DVAR) system (EDAS).

In general, the winter atmospheric transport and precipitation are realistic. The climatology and interannual variability of the Pacific, subtropical jet streams, and low-tropospheric moisture transport are well captured. During the summer season, the basic features and the evolution of the North American monsoon (NAM) revealed by the RR compare favorably with observations. The RR also captures the out-of-phase relationship of precipitation as well as the moisture flux convergence between the central United States and the Southwest. The RR is able to capture the zonal easterly Caribbean low-level jet (CALLJ) and the meridional southerly Great Plains low-level jet (GPLLJ). Together, they transport copious moisture from the Caribbean to the Gulf of Mexico and from the Gulf of Mexico to the Great Plains, respectively. The RR systematically overestimates the meridional southerly Gulf of California low-level jet (GCLLJ). A comparison with observations suggests that the meridional winds from the RR are too strong, with the largest differences centered over the northern Gulf of California. The strongest winds over the Gulf in the RR extend above 700 hPa, while the operational EDAS and station soundings indicate that the GCLLJ is confined to the boundary layer.

Corresponding author address: Dr. K. C. Mo, Climate Prediction Center, NOAA/NWS/NCEP, 5200 Auth Road, Rm 605, Camp Springs, MD 20746. Email: Kingtse.Mo@noaa.gov

Abstract

The large-scale atmospheric hydrologic cycle over the United States and Mexico derived from the 23-yr NCEP regional reanalysis (RR) was evaluated by comparing the RR products with satellite estimates, independent sounding data, and the operational Eta Model three-dimensional variational data assimilation (3DVAR) system (EDAS).

In general, the winter atmospheric transport and precipitation are realistic. The climatology and interannual variability of the Pacific, subtropical jet streams, and low-tropospheric moisture transport are well captured. During the summer season, the basic features and the evolution of the North American monsoon (NAM) revealed by the RR compare favorably with observations. The RR also captures the out-of-phase relationship of precipitation as well as the moisture flux convergence between the central United States and the Southwest. The RR is able to capture the zonal easterly Caribbean low-level jet (CALLJ) and the meridional southerly Great Plains low-level jet (GPLLJ). Together, they transport copious moisture from the Caribbean to the Gulf of Mexico and from the Gulf of Mexico to the Great Plains, respectively. The RR systematically overestimates the meridional southerly Gulf of California low-level jet (GCLLJ). A comparison with observations suggests that the meridional winds from the RR are too strong, with the largest differences centered over the northern Gulf of California. The strongest winds over the Gulf in the RR extend above 700 hPa, while the operational EDAS and station soundings indicate that the GCLLJ is confined to the boundary layer.

Corresponding author address: Dr. K. C. Mo, Climate Prediction Center, NOAA/NWS/NCEP, 5200 Auth Road, Rm 605, Camp Springs, MD 20746. Email: Kingtse.Mo@noaa.gov

1. Introduction

The National Centers for Environmental Prediction (NCEP) Environmental Modeling Center (EMC) has completed a regional reanalysis (RR) for the period 1979 to 2002 (Mesinger et al. 2004). The RR is a long-term homogeneous mesoscale regional analysis performed with a frozen state-of-the-art model and data assimilation system. The domain covers North America and the adjacent oceans. The model used is the NCEP operational Eta Model of 2003, which has a horizontal resolution of 32 km and 45 layers in the vertical. A selected subset of the RR data was postprocessed and archived at the Climate Prediction Center. The data were archived on the Eta Advanced Weather Interactive Processing System (AWIPS) grid with a horizontal resolution of 0.3°. The temporal resolution is eight times daily (i.e., 3 hourly). Daily and monthly means of selected variables were computed along with climatological means for the period 1979 to 2001. The climatology datasets are available from the University Corporation for Atmospheric Research (UCAR) Joint Office for Science Support (JOSS; http://www.joss.ucar.edu).

Input data for the RR are listed in Mesinger et al. (2004) and Shafran et al. (2004). Precipitation (P) data sources for the RR include Parameter-elevation Regressions on Independent Slopes Model (PRISM) analysis of rain gauge station precipitation over the United States, daily rain gauge data over Mexico and Canada, the NCEP/Department of Energy (DOE) global analysis (R2; Kanamitsu et al. 2001), and the Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP; Xie et al. 2003) pentad data over oceans south of 27.5°N. There is no assimilation of precipitation north of 42.5°N because no reliable precipitation data were available. There is a transition band over a 15° latitude belt centered at 35°N. The P from R2 was also used to derive hourly P from daily P over Canada and Mexico. The boundary conditions were taken from R2. The rawinsondes, dropsondes, pibals, and cloud drift winds from the geostationary satellite and aircraft data were obtained from the global reanalysis R1 (Kalnay et al. 1996) input data archive. In addition to global reanalysis inputs, the Television Infrared Observation Satellite (TIROS) Operational Vertical Sounder (TOVS)-1B radiances data were also assimilated (Tables 1 and 2 of Shafran et al. 2004). A detailed comparison of the input data between the R2 and the RR is given in Shafran et al. (2004). (More details on the RR system, input data, and model can be found on the RR Web site at http://www.emc.ncep.noaa.gov.)

In comparison to the global reanalysis, the RR has much higher vertical and horizontal resolution. Both the R1 and R2 models have low horizontal resolutions of T62 with 28 levels in the vertical. The RR also has the North American land data assimilation system (NLDAS) (Mitchell et al. 2003; Ek et al. 2003) as a land surface subcomponent. Therefore, the surface fluxes and hydrological variables from the RR are more reliable than those from the global reanalyses R1 and R2.

In comparison with the recent operational Eta Model three-dimensional variational data assimilation (3DVAR) system (EDAS), the EDAS assimilates more satellite data than the RR. Currently the EDAS has the horizontal resolution of 12 km and also has the NLDAS subcomponent. The EDAS assimilates the Special Sensor Microwave Imager (SSM/I) wind speed, SSM/I and Tropical Rainfall Measuring Mission (TRMM) P estimates, and National Oceanic and Atmospheric Administration (NOAA) satellites NOAA-14, NOAA-16, and NOAA-17 High Resolution Infrared Radiation Sounder (HIRS) 1b radiances, and Advanced Microwave Sounding Unit (AMSU) -A 1b and -B 1b radiances. These data are not included in the RR. Both the EDAS and the RR assimilate observed P, but the EDAS includes only radar precipitation over the continental United States.

Earlier studies of the hydrologic cycle over North America were largely based on rawinsondes (Rasmusson 1967, 1968; Peixoto and Oort 1992) and satellite data (Chu et al. 1993; Janowiak et al. 1995). More recently, improvements in the model and data assimilation procedures have provided a more realistic description of the hydrologic cycle globally. Many budget and hydrologic cycle studies used the R1 or R2 reanalyses (Roads et al. 1997, 2002). Because the T62 model used in the R1 and R2 cannot adequately resolve mesoscale features (Roads et al. 1994; Trenberth and Guillemot 1995; Mo and Higgins 1996; Schmitz and Mullen 1996), global reanalyses based on low-resolution models are not well suited for studies of the continental hydrologic cycle over North America.

Previous studies of mesoscale moisture transport were based mostly on station data (Douglas et al. 1993; Douglas 1995) and regional model simulations (Stensrud et al. 1995, 1997; Mo and Berbery 2004). Many studies used the operational EDAS to examine the hydrologic cycle over the United States (Berbery et al. 1996; Berbery 2001), but the EDAS is operational, so there have been many changes in physics, resolution, and data assimilation procedures over the years that have affected the homogeneity of the dataset. The RR dataset offers a unique opportunity to examine continental moisture transport and the hydrologic cycle over North America.

The major objective of this paper is to evaluate the atmospheric moisture transport revealed by the RR by comparing them with observations, satellite estimates, and the operational EDAS. The study period is 1979 to 2001. The comparison with the R1 shows the differences between the global and regional reanalyses. The hydrologic variables P, and evaporation (E) are accumulations from 3-h forecasts in the assimilation cycle. All other variables presented here are from the analysis. Because of the injection of observed data and the model errors, the long-term mean of 〈EP〉 is not necessarily balanced by the mean vertically integrated moisture flux divergence 〈D(Q)〉. This inconsistency is corrected by adding an analysis increment term Winc, which is the residual term needed to close the budget cycle (Mo and Higgins 1996):
i1525-7541-6-5-710-eq1
where the 〈〉 denotes the long-term time mean.

The observed P dataset used for verification is the Unified Rain Gauge Dataset (URD), which covers the United States and Mexico (Higgins et al. 2000). The URD data were interpolated from daily rain gauge observations to a 1°-resolution grid using a Cressman (1959) scheme. The seasonal cycle is examined in section 2. Results are compared with previous seasonal cycle studies. Because of the fine-resolution model, many detailed features are revealed by the RR. The winter precipitation regimes are discussed in section 3. The summer precipitation regimes and low-level jets are evaluated in section 4, and conclusions are given in section 5.

2. Seasonal cycle of precipitation

Both the monthly mean RR climatology (Fig. 1) and the daily P climatology for selected regions (Fig. 2) for the period 1979–2001 indicate that there are two P regimes over the United States and Mexico: the winter Mediterranean regime and the summer continental regime (Kendrew 1922; Hsu and Wallace 1976).

The winter rainfall regime covers western North America and the Southeast. Over the Pacific Northwest, rainfall begins in October and peaks in November (Fig. 2a). As the season progresses, precipitation extends southward to California. Precipitation over California peaks in February (Fig. 2b) and then retreats northward in March. At that time, P over the Pacific Northwest increases. Therefore, P over the Pacific Northwest has one maximum in November and another one in late February. The west region is mostly dry during summer (Fig. 1).

In the Southeast the mean daily rainfall is consistently above 2 mm day–1, with a relatively small seasonal cycle. During spring, precipitation expands westward from the Southeast to the central United States. Over the southern Great Plains, an early summer maximum occurs in late May or early June. At the same time, monsoon rainfall starts over southern Mexico in May and intensifies in June. Rainfall spreads northward through northwestern Mexico along the western slopes of the Sierra Madre Occidental (SMO) and reaches the Southwest in July, while rainfall over the southern Great Plains diminishes and retreats eastward. Monsoon rainfall reaches a maximum over the Southwest in August and then retreats in September. The demise of the monsoon is slower than the onset. The evolution of the monsoon depicted by the RR is consistent with observations (Higgins et al. 1997a).

Rainfall migrates from the Southeast to the southern plains in spring and retreats back to the Southeast in autumn. This results in a bimodal distribution of P over the southern plains (Fig. 2g). The first maximum is in May–June and the second maximum is in October–November. Precipitation also shows a bimodal annual cycle in southern Mexico, where the midsummer drought has been documented by Magaña et al. (1999). Both the RR and URD capture the bimodal distribution over southern Mexico, but the RR shows a weaker magnitude (Fig. 2h). Over the upper Midwest, Keables (1989) documented the bimodal distribution of rainfall. This feature does not occur every year and does not appear in the daily climatology depicted by the RR or the URD (Fig. 2e)

The RR climatology compares favorably with the URD (Figs. 2 and 3). To compute the difference between two P datasets, the RR precipitation was coarsened to 1° resolution. The RR assimilated the P gauge data after applying the PRISM procedures over the United States. Over Mexico, the Cressman scheme was used to analyze P, but no PRISM procedure was applied. No smoothing was applied when these two datasets were merged, which explains the discontinuity at the international border. Other major differences are found over southern Mexico, where P from the RR is about 1–2 mm day−1 less than the URD.

3. Winter regime

The western region and the Southeast experience a winter raining season as indicated by Fig. 1. Over the western region, rainfall extends northward from California to the Pacific Northwest, Canada, and Alaska (Fig. 4a). Precipitation P over the Atlantic coincides with the locations of the storm tracks, and P over the tropical eastern Pacific is associated with the ITCZ. The contribution to P over the western region is largely from the vertically integrated moisture flux convergence −D(Q) (Fig. 5a). Evaporation E (Fig. 4b) is small over land and larger over the oceans in comparison to divergence D(Q) (Fig. 5a). Over the Southeast, both E and −D(Q) contribute to P. Two branches of moisture fluxes from the North Pacific bring needed moisture for rainfall over the western region (Fig. 5b). One branch turns northward to supply moisture to the Pacific Northwest, Canada, and Alaska while the southern branch supplies moisture to California. A subtropical branch of moisture flux comes from the North Pacific through northern Mexico to supply moisture to the southern plains and the Southeast. This branch sometimes merges with one from the Gulf of Mexico to increase the total moisture supply to the eastern United States.

In comparison with the RR (Fig. 4), the R1 captures most large-scale features of the atmospheric transport (Fig. 6) except that the magnitudes of P and E over the western region are smaller. The 32-km Eta Model is able to capture many detailed features driven by orography over the western region, while the R1 misses. The largest differences are located over northern Mexico and the Gulf of California. The RR shows that large E extends from the eastern Pacific to the Gulf of California, but the R1 misses these features entirely because the R1 model does not resolve the Gulf of California.

Because of the RR model errors and the assimilation of P, the increment residual term Winc (Fig. 4d) is very large over the oceans in comparison to the global analyses (Mo and Higgins 1996). The values of Winc are more than 2 mm day−1 along the path of storm track in the North Atlantic, the ITCZ, and over the Gulf of Mexico. For the R1, the magnitudes of Winc are less than 1 mm day−1 over the oceans. The reason may be that there are not many observations over the oceans. The R1 does not assimilate P so the differences between the first guess and the analysis are small. For the RR, the large residual is in part due to the model errors, but also to the difference between the CMAP data and the model P.

The dominant mode for winter precipitation over North America is a dipole pattern of P anomalies between California and the Pacific Northwest. This mode is found on the decadal, interannual, and intraseasonal time scales (Dettinger et al. 1998; Mo and Higgins 1998). This pattern is robust and is also captured by R1 (Mo and Higgins 1998). Precipitation difference maps between the eight wettest months and the eight driest months over California (30°–40°N, 118°–125°W over land) for winter (December–March) for the 1979–2001 period from the RR (Fig. 7a) are similar to those from the URD (not shown). Both show the dipole pattern. In addition to the dipole, there is another negative center in the eastern Pacific that represents the weakening of the ITCZ during wet California winter. The contribution from E is small. Therefore, the dipole also appears in the moisture flux divergence difference. The dipole is supported by the cyclonic anomalies near the West Coast. The weakening of the northern branch of moisture transport is accompanied by a strengthening of the southern branch (Fig. 7b). Typically, when more moisture is transported into California, less moisture is transported to the Pacific Northwest and vice versa.

4. Summer regime

a. Hydrologic cycle in summer

The summer precipitation regime is dominated by the North American monsoon system (Higgins et al. 1997a). The evolution of P shows a bimodal distribution over the southern plains. To examine the bimodal distribution in detail, we plotted the time–longitude diagrams of daily P, E, and moisture flux convergence averaged from 32° to 36°N for the period from 1979 to 2001 (Fig. 8). A 16-day running mean was applied to all fields before plotting.

The evolution of P is consistent with discussions in section 2 (Fig. 1). When the monsoon rainfall starts, rainfall over the southern plains including Texas and Oklahoma diminishes. There is a phase reversal between P over the southern plains (32°–36°N, 85°–100°W) and the Southwest (Higgins et al.1999; Mo and Berbery 2004). For the southern plains, the winter (November–March) contributions to P are from the moisture flux convergence (Fig. 8d). The contribution from E is small. Evaporation starts to increase in mid-March and reaches a maximum in June. Evaporation dominates the contribution to P in early summer, as noticed by Berbery and Rasmusson (1999) and Berbery et al. (1996). As first reported by Rasmusson (1967, 1968), E is larger than P over the central United States in summer; E starts to decrease in July. The decrease of E and the moisture flux divergence contribute to the P minimum in late summer. In October, the moisture flux convergence starts to increase and dominates the contribution to P when E decreases to less than 2 mm day−1. The bimodal distribution of precipitation is confined to the southern plains. Over the northern plains (38°–48°N, 85°–100°W), there is only one P maximum in June (Fig. 8c). The contributions to the P maximum come almost entirely from E.

The moisture budget terms for July–September (JAS) are presented in Figs. 9 and 10. Over land, P is about 3 mm day−1 over the Great Plains, but only 1–2 mm day−1 over the Southwest. Large P is also located over the Southeast, with a maximum over Florida. A rainfall band extends from the ITCZ in the Tropics to southern Mexico and the western slopes of the SMO (Fig. 9a), which is an important feature of the monsoon system. Over land, large E values are located over the central and eastern United States. Both E (Fig. 9b) and the vertically integrated moisture flux convergence (Fig. 10a) contribute to P over southern Mexico, the western slopes of the SMO, and the Southeast in summer (Figs. 9b, 9c and 10a). Over the central United States, the contribution from the vertically integrated moisture flux convergence −D(Q)) is small, but over the Southwest, both E and D(Q) contribute to P. The inverse relationship between D(Q) over the central United States and the Southwest is consistent with the findings of Berbery and Fox-Rabinovitz (2003). The residual term Winc is large over the oceans for all seasons. There are also large negative Winc values over northern Mexico and the Southwest. Since there are many soundings located in the Southwest, large Winc indicates discrepancies between the RR forecasts and observations.

The summer moisture transport over the United States is regulated by two meridional low-level jets (LLJs) from the Great Plains (GPLLJ) and from the Gulf of California (GCLLJ), and one zonal LLJ from the Caribbean (CALLJ) (Amador 1998; Amador and Magaña 1999) (Fig. 10b). The major branch of moisture transport is carried by the CALLJ from the tropical Atlantic to the Gulf of Mexico and then is transported by the GPLLJ from the Gulf of Mexico to the central United States. The path then turns northeastward and the jet brings moisture to the Ohio Valley and the eastern United States. There is another branch of moisture transport extending directly from the Gulf of Mexico to northern Mexico and eastern New Mexico. Over the Southwest, the moisture comes from the North Pacific and from the eastern Pacific. That branch is carried by the GCLLJ through the Gulf of California to the Southwest. Both the GPLLJ and the GCLLJ can be represented by the meridional component of the vertically integrated moisture flux, while the CALLJ is represented by the zonal component of the moisture flux.

In comparison to the RR, the R1 captures the large-scale features of the moisture transport (Fig. 11). For example, the R1 also shows more E than P over the central United States in summer. The R1 has wet biases over the Southeast and dry biases over the Southwest and the Great Plains. The R1 also overestimates P in the Atlantic and over the Caribbean and the ITCZ. The coarse R1 model does not capture the monsoon rainfall well. This is apparent from the mean vertically integrated moisture flux (Fig. 11d). It captures the Great Plains low-level jet extending from the Gulf of Mexico, which brings moisture to the Plains. The R1 also captures the location, intensity, and profile of the CALLJ (Amador 1998). However, it does not show the low-level jet extending from the Gulf of California to the Southwest. The R1 shows that one branch of moisture fluxes extends from the Gulf of Mexico through northern Mexico to the Southwest and another branch of moisture from the North Pacific. The T62 model is not able to capture the GCLLJ. Findings here are consistent with Schmitz and Mullen (1996) that a model higher than T107 is needed to capture moisture transport associated with monsoon rainfall.

b. Low-level jets

1) The Caribbean low-level jet

The CALLJ is an easterly jet that is most pronounced during the late spring and early summer months (Amador 1998; Amador and Magaña 1999). To assess the ability of the RR to capture the CALLJ, we compare the RR surface winds with satellite scatterometer low-level wind measurements over the Caribbean region. The scatterometer measures the ocean surface wind speed and direction from the scattering of microwaves at the ocean surface (Tang and Liu 1996). The winds were interpolated to the 0.5° grid twice daily by Tang and Liu (1996). The data presented here were obtained from the Jet Propulsion Laboratory. The scatterometer winds were not assimilated by the RR so this serves as an independent verification. The surface winds are easterlies over the Atlantic during June and July. The horizontal structure of the surface zonal wind field from the satellite estimates (Figs. 12a and 12c) averaged over June and July for 2000 and 2001 indicates a maximum in the northern tropical Atlantic centered at 12°–14°N, 70°–80°W. The location and magnitude of the jet maximum are consistent with studies by Amador (1998) and Amador and Magaña (1999). The RR captures the location of the maximum while the magnitude is only about 1 m s−1 weaker than the satellite measurements (Figs. 12b and 12d). The RR does not capture the wind structures near the coastline, as revealed by the Quick Scatterometer (QuikSCAT).

2) Great Plains low-level jet

A comparison between the monthly mean vertically integrated meridional flux () averaged from the RR and the operational EDAS for the period 1995–2000 is shown in Fig. 13. During this period, the two models have comparable horizontal resolution. The GPLLJ is strongest in June, when P over the central United States reaches a maximum (Fig. 1). In July, as the monsoon rainfall reaches the Southwest, transport associated with the GCLLJ strengthens. At the same time, both rainfall over the Great Plains and the GPLLJ diminish. In September, the GPLLJ is weaker, with a maximum moisture flux of 60 kg (ms)–1, less than half that of the June maximum. The evolution of the two LLJs is consistent with the P seasonal evolution (Fig. 1).

The GPLLJ generated by the RR compares well with the EDAS, but the GCLLJ is systematically stronger. The difference in moisture flux can be as large as 90 kg (ms)−1. The largest differences are located over the northern Gulf of California. The strong GCLLJ in the RR is present in all summer months and is not limited to these 6 yr presented here (i.e., 1995–2000). Consistent with the overestimated GCLLJ, more moisture is transported to the Southwest. The RR shows more precipitable water along the western slopes of the SMO and in Arizona and New Mexico (not shown), and that may contribute to large Winc over the region (Fig. 9d). A comparison with the independent observations in the next section indicates that the GCLLJ in the RR is too strong.

The GPLLJ generated by the RR compares favorably with observations. For example, the GPLLJ was strong and regular for two weeks in August 1994 (17–31 August 1994). A time series of the 3-hourly wind speeds at a grid point (36°N, 97.5°W) from the RR (Fig. 14) is compared with the hourly wind profiler data at the same grid point from Higgins et al. (1997b, their Fig. 5). The wind profiler data were not assimilated in the RR so this serves as an independent confirmation. The wind profiler data show a low-level regular nocturnal peak of 10–16 m s−1 centered about 250–500 m above the surface (or roughly the 950–975-hPa level). The LLJ depicted by the RR is weaker and less organized at the beginning of the time series, but it shows the regular diurnal oscillations from 23 to 31 August 1994. The maximum is located between 930 and 960 hPa, which is at a slightly higher altitude than the maximum depicted by the wind profiler data. This may be due to the vertical resolution of the RR, which is 25 hPa, while the wind profiler is essentially continuous. The RR shows that the LLJ starts to develop about 0300 UTC and reaches a maximum between 0600 and 0900 UTC and persists through 1200 UTC and then diminishes (Fig. 14b). The diurnal cycle of the GPLLJ compares well with the wind profiler data (Higgins at el. 1997b).

3) Gulf of California low-level jet

The largest differences in the vertically integrated meridional moisture transport between the RR and the EDAS are found over the northern Gulf of California (Fig. 13). To verify, we compared the meridional winds from the RR with ones from the pilot balloon and rawindsonde data at Puerto Peñasco located at the northern end of the Gulf of California (31.3°N, 113.5°W) reported by Douglas et al. (1998, their Fig. 3). The rawinsonde data were not used by the RR so it serves as an independent confirmation. A time series of the 975-hPa meridional wind from the RR at the nearest location to Puerto Peñasco from 31 July to 14 August 1995 (Fig. 15a) indicates that the meridional winds are southerly during this period. The RR seems to capture the low-frequency components of the wind, but the mean magnitude is about 3 m s−1 stronger than the observations in comparison with Fig. 16 [reproduced from Fig. 3 of Douglas et al. (1998)].

Vertical profiles of the meridional wind averaged over this 15-day period (Fig. 15b) show that the RR captures the diurnal cycle quite well, with maximum winds at 0100 LT and minimum winds at 1600 LT, similar to observations. The maximum magnitude is about 10.5 m s−1, which is again about 3–4 m s−1 higher than the observed maximum. The stronger meridional wind speeds occur at all levels. At 700 hPa, the RR winds are 1–3 m s−1 (Fig. 15b). The station data indicate that winds above 700 hPa are less than 1 m s−1 (Fig. 16b). This indicates that the jet is shallow and is confined to the boundary layer. The strong meridional winds above the boundary layer occur in every summer.

Vertical cross sections of the monthly mean vertically integrated meridional flux at 30°N for July–September 1997–2000 from the RR and the EDAS (Fig. 17) show that the meridional flux associated with the GPLLJ extends to 700 hPa and fluxes are strongest in July and weakest in September. The maximum is located along the eastern slopes of the Rockies near 925 hPa. The intensity at the core of the jet is also similar. For the GCLLJ, the comparison is less favorable. Both models show that the moisture flux is centered along the Gulf of California near 114°W for all summer months. For July and August, moisture flux is confined to the boundary layer below 850 hPa in the EDAS, while it extends above 850 hPa in the RR, with fluxes nearly double those in the EDAS.

5. Conclusions

This study examines the atmospheric moisture transport depicted by the RR. The focus is over the United States and Mexico where the observed P data were assimilated. The RR is a mesoscale model with 32-km horizontal resolution and is able to resolve some of mesoscale features over North America. The RR has the NLDAS component so the surface properties are more reliable than the global reanalysis. Both the EDAS and the RR are based on the Eta Model, but the EDAS is an operational system so there are many changes in model physics, resolution, and procedures, while the RR was produced with a frozen model and data assimilation system.

The atmospheric moisture transport as estimated in the RR is overall realistic. Because of precipitation assimilation and the differences between the first guess and analysis, a residual term Winc is needed to balance the budget. Over the oceans, there are not many observations available; the Winc reflects in part the differences between the model P forecasts and the P analyses. For summer, Winc is also large over the monsoon core region and the Southwest. That may be related to the strong GCLLJ in the region.

For the winter hydrological cycle, the RR captures the two branches of moisture transport from the North Pacific to the Pacific Northwest and California as well as the out-of-phase relationship between these two regions. During summer the dominant feature is the North American monsoon and rainfall over the Great Plains. The RR captures the inverse relationship of P as well as the vertically integrated moisture flux divergence between the southern plains and the Southwest. The most important branch of moisture transport is from the Caribbean carried by the zonal easterly CALLJ to the Gulf of Mexico and then by the meridional GPLLJ to the Great Plains. The winds associated with both the CALLJ and the GPLLJ depicted by the RR compare favorably with the independent wind profiler data, and the satellite estimates. The comparison with the EDAS is also good.

The RR systematically overestimates the GCLLJ water vapor transport. A comparison with independent soundings and satellite wind estimates (not shown) indicates that the surface meridional wind from the RR is systematically too strong. The largest differences are centered over the northern Gulf of California. The strong winds are not limited to the surface. Large winds associated with the GCLLJ extend above 700 hPa, while the operational EDAS and soundings show that the GCLLJ is confined to the boundary layer. The strong GCLLJ occurs every summer and that limits the variability of the jet. Therefore, the RR may not be suited for studies of the GCLLJ and its relationship to monsoon rainfall.

The reason for the strong GCLLJ is not clear and is still under investigation. The RR and EDAS both are based on the Eta Model. The major differences between two systems are that the EDAS assimilates more satellite data than the RR, but it only assimilates the radar P data over the continental United States. We are in the process of performing data assimilation with different input data to examine the reasons for the strong GCLLJ. A big part of the problem is that observations are rather sparse in this region. More soundings along the Gulf of California may improve the low-level winds associated with the GCLLJ. Enhanced observations gathered during the North American Monsoon Experiment (NAME) 2004 field campaign will help us to better understand the spatial structure and temporal evolution of the GCLLJ, and hence improve our ability to diagnose the GCLLJ in the RR.

In comparison with the global reanalyses R1 and R2, the RR has higher horizontal and vertical resolution. As a result, the RR is able to resolve some of mesoscale features that the T62 global reanalyses are not able to capture. In comparison to the RR, the R1 is able to capture the large-scale features of the hydrological cycle. It also captures the CALLJ and the GPLLJ well. The coarse-resolution R1 does not resolve the Gulf of California. As a result, the moisture transport from the R1 over the monsoon core region is not realistic. The R1 does not assimilate P. It has wet biases over the Atlantic near the Florida coast and over the ITCZ region.

The RR will continue as near real-time Regional Climate Data Assimilation System (R_CDAS) and will be used by the Climate Prediction Center for climate monitoring. The P data inputs will be different. The URD and the CMAP data will not be available in real time. Instead, the real-time precipitation analysis for the United States and Mexico along with the Climate Prediction Center morphing method (CMORPH; Joyce et al. 2004) data will be used.

Acknowledgments

This project is partially supported by NOAA Grant GC04-072. The authors thank Dr. Hugo Berbery for supplying the operational EDAS data.

REFERENCES

  • Amador, J. A., 1998: A climatic feature of the tropical Americas: The trade wind easterly jet. Top. Meteor. Oceanogr., 5 , 2,. 113.

  • Amador, J. A., and Magaña V. , 1999: Dynamics of the low level jet over the Caribbean Sea. Preprints, Third Conf. on Hurricanes and Tropical Meteorology, Dallas, TX, Amer. Meteor. Soc., 868–869.

  • Berbery, E. H., 2001: Mesoscale moisture analysis of the North American monsoon. J. Climate, 14 , 121137.

  • Berbery, E. H., and Rasmusson E. M. , 1999: Mississippi moisture budgets on regional scales. Mon. Wea. Rev., 127 , 26542673.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Berbery, E. H., and Fox-Rabinovitz M. S. , 2003: Multiscale diagnosis of the North American monsoon system using a variable resolution GCM. J. Climate, 16 , 16291647.

    • Search Google Scholar
    • Export Citation
  • Berbery, E. H., Rasmusson E. M. , and Mitchell K. E. , 1996: Studies of North American continental scale hydrology using Eta Model forecast products. J. Geophys. Res., 101 , 73057319.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chu, L. S., Chang A. T. C. , and Janowiak J. , 1993: Comparison of monthly mean rates derived from GPI and SSM/I using probability distribution functions. J. Appl. Meteor., 32 , 323334.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cressman, G. P., 1959: An operational objective analysis system. Mon. Wea. Rev., 87 , 367374.

  • Dettinger, M. D., Cayan D. R. , Diaz H. F. , and Meko D. M. , 1998: North–south precipitation patterns in western North America on interannual and decadal time scales. J. Climate, 11 , 30953111.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Douglas, M. W., 1995: The summertime low-level jet over the Gulf of California. Mon. Wea. Rev., 123 , 23342347.

  • Douglas, M. W., Maddox R. A. , Howard K. , and Reyes S. , 1993: The Mexican monsoon. J. Climate, 6 , 16651677.

  • Douglas, M. W., Valdez-Manzanilla A. , and Cueto R. G. , 1998: Diurnal variation and horizontal extent of the low level jet over the northern Gulf of California. Mon. Wea. Rev., 126 , 20172025.

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

    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., Yao Y. , and Wang X. , 1997a: Influence of the North American monsoon system on the U.S. summer precipitation regime. J. Climate, 10 , 26002622.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., Yao Y. , Yarosh E. S. , Janowiak J. E. , and Mo K. C. , 1997b: Influence of the Great Plains low-level jet on summer precipitation and moisture transport over the central United States. J. Climate, 10 , 481507.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., Chen Y. , and Douglas A. V. , 1999: Interannual variability of the North American warm season precipitation regime. J. Climate, 12 , 653680.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., Shi W. , Yarosh E. , and Joyce R. , 2000: Improved United States Precipitation Quality Control System and Analysis. NCEP/Climate Prediction Center Atlas 7, U.S. Department of Commerce, 45 pp. [Available online at http://www.cpc.ncep.noaa.gov/research_papers/ncep_cpc_atlas/7/index.html.].

  • Hsu, C. F., and Wallace J. M. , 1976: The global distribution of the annual and semiannual cycles in precipitation. Mon. Wea. Rev., 104 , 10931101.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Janowiak, J. E., Xue P. , Morrissey M. L. , and Legates D. R. , 1995: An examination of the east Pacific ITCZ rainfall distribution. J. Climate, 8 , 28102823.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Joyce, R. J., Janowiak J. E. , Arkin P. A. , and Xie P. , 2004: CMORPH: A method that produces global precipitation estimates from passive microwave and infrared data at high spatial and temporal resolution. J. Hydrometeor., 5 , 487503.

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

  • Kanamitsu, M., Ebisuzaki W. , Wollen J. , Yang S. K. , Hnilo J. J. , Fiorino M. , and Potter G. L. , 2001: NCEP/DOE AMIP-II reanalysis R-2. Bull. Amer. Meteor. Soc., 82 , 16311643.

    • Search Google Scholar
    • Export Citation
  • Keables, M., 1989: A synoptic climatology of the bimodal precipitation distribution in the upper Midwest. J. Climate, 2 , 12891294.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kendrew, W. G., 1922: The Climates of the Continent. Clarendon Press, 308 pp.

  • Magaña, V., Amador J. A. , and Medina S. , 1999: The midsummer drought over Mexico and Central America. J. Climate, 12 , 15771588.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mesinger, F., and Coauthors, 2004: North American regional reanalysis. Preprints, 15th Symp. on Global Change and Climate Variations, Seattle, WA, Amer. Meteor. Soc., CD-ROM, P1.1.

  • Mitchell, K. E., and Coauthors, 2004: The multi-institution North American Land Data Assimilation System (NLDAS): Utilizing multiple GCIP products and partners in a continental distributed hydrological modeling system. J. Geophys. Res., 109 .D07S90, doi:10.1029/2003JD003823.

    • Search Google Scholar
    • Export Citation
  • Mo, K. C., and Higgins R. W. , 1996: Large scale atmospheric moisture transport as evaluated in the NCEP/NCAR and the NASA/DAO reanalysis. J. Climate, 9 , 15311545.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mo, K. C., and Higgins R. W. , 1998: Tropical influence on California precipitation. J. Climate, 11 , 412430.

  • Mo, K. C., and Berbery E. H. , 2004: The low level jets and the summer precipitation regimes over North America. J. Geophys. Res., 109 .D06117, doi:10.1029/2003JD004106.

    • Search Google Scholar
    • Export Citation
  • Peixoto, J. P., and Oort A. H. , 1992: Physics of Climate. American Institute of Physics, 520 pp.

  • Rasmusson, E. M., 1967: Atmospheric water vapor transport and the water balance of North America. I: Characteristics of the water vapor flux field. Mon. Wea. Rev., 95 , 403426.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rasmusson, E. M., 1968: Atmospheric water vapor transport and the water balance of North America. II: Large scale water balance investigations. Mon. Wea. Rev., 96 , 720734.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roads, J. O., Chen S. C. , Guetter A. K. , and Georgakakos P. K. , 1994: Large scale aspects of the United States hydrological cycle. Bull. Amer. Meteor. Soc., 75 , 15891610.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roads, J. O., Chen S. C. , Kamamitsu M. , and Juang H. M. , 1997: GDAS’s GCIP energy budgets. J. Atmos. Sci., 54 , 17761794.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roads, J. O., Kanamitsu M. , and Stewart R. , 2002: CSE water and energy budgets in the NCEP/DOE reanalysis 2. J. Hydrometeor., 3 , 227248.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schmitz, J. T., and Mullen S. L. , 1996: Water vapor transport associated with the summertime North American monsoon as depicted by ECMWF analyses. J. Climate, 9 , 16211634.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shafran, P., Woollen J. , Ebisuzaki W. , Shi W. , Fan Y. , Grumbine R. , and Fennessy M. , 2004: Observational data used for assimilation in the NCEP North American regional reanalysis. Preprints, 14th Conf. on Applied Climatology, Seattle, WA, Amer. Meteor. Soc., CD-ROM, 1.4.

  • Stensrud, D. J., Gall R. L. , Mullen S. L. , and Howard K. W. , 1995: Model climatology of the Mexico monsoon. J. Climate, 8 , 17751794.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stensrud, D. J., Gall R. L. , and Nordquist M. K. , 1997: Surge over the Gulf of California during the Mexico monsoon. Mon. Wea. Rev, 125 , 417437.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tang, W., and Liu W. T. , 1996: Objective interpolation of scatterometer winds. JPL Publication 96-19, Jet Propulsion Laboratory, Pasadena, CA, 16 pp.

  • Trenberth, K. E., and Guillemot C. J. , 1995: Evaluation of the atmospheric moisture budget as seen from analyses. J. Climate, 8 , 22252272.

    • Search Google Scholar
    • Export Citation
  • Xie, P. P., Janowiak J. E. , Arkin P. A. , Alder R. , Gruber A. , Ferraro R. , Huffmann G. J. , and Curtis S. , 2003: GPCP pentad precipitation analyses: An experimental dataset based on gauge observations and satellite estimates. J. Climate, 16 , 21972214.

    • Crossref
    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Monthly mean precipitation climatology from the RR for the period from 1979 to 2001 for (a) Jan, (b) Feb, (c) Mar, (d) Apr, (e) May, (f) Jun, (g) Jul, (h) Aug, (i) Sep, (j) Oct, (k) Nov, and (l) Dec. Contour intervals are 1, 2, 3, 4, 6, 8, and 10 mm day−1.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 2.
Fig. 2.

Seasonal cycle from daily mean P climatology from the RR (dark line) and from the URD (crosses) for (a) the Pacific Northwest (40°–50°N, 118°–125°W), (b) California (30°–40°N, 118°–125°W, over land), (c) northern Mexico (25°–32°N, 100°–110°W), (d) Southeast (30°–33°N, 80°–90°W), (e) Midwest (43°–48°N, 80°–90°W), (f) northern plains (38°–48°N, 90°–100°W), (g) southern plains (32°–38°N, 90°–100°W), and (h) southern Mexico (17°–23°N, 95°–100°W). A 16-day running mean was applied to the time series before plotting.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 3.
Fig. 3.

Same as in Fig. 1, but for the monthly mean difference between the RR and the observed precipitation from the URD averaged from 1979 to 2001. Contour interval is 0.5 mm day−1. Zero contours are omitted.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 4.
Fig. 4.

(a) Mean P for Dec–Feb averaged from 1979 to 2001. Contour interval 2 mm day−1. Contours 0.5 and 1 mm day−1 are added. Values greater than 4 mm day−1 are shaded. (b) As in (a), but for E. (c) As in (a), but for mean PE. Values less than −2 mm day−1 are shaded. (d) Same as (c), but for the residual analysis increment term Winc

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 5.
Fig. 5.

(a) Mean vertically integrated moisture flux divergence for Dec–Feb (DJF) averaged from 1979 to 2001. Contour interval is 1 mm day−1. Values less than −4 mm day−1 are shaded. (b) The mean vertically integrated moisture flux (qu; ). The unit vector is 200 kg (ms)−1.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 6.
Fig. 6.

(a)–(c) Same as Figs. 4a–c, but for the R1. (d) Same as Fig. 5a, but for R1.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 7.
Fig. 7.

(a) Composite P difference between the eight wettest and eight driest months in California (30°–40°N, 118°–125°W over land) for Dec–Mar 1979–2001 from the RR. Contour interval is 2 mm day−1. Zero contours are omitted. Contours −1 and 1 mm day−1 are added. Areas with positive (negative) values that are statistically significant at the 5% level are shaded dark (light). (b) As in (a), but for the vertically integrated moisture flux and divergence. The unit vector is 200 kg (ms)−1.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 8.
Fig. 8.

Time–longitude cross section of (a) P averaged from 32° to 36°N for the daily mean climatology from the RR. A 16-day running mean was applied to the field before plotting. Contour interval is 1 mm day−1. Values greater than 3 mm day−1 are shaded. (b) As in (a), but for E. (c) As in (a), but P averaged from 38° to 48°N. (d) Same as (a), but for the vertically integrated moisture flux convergence. Values greater than 1 mm day−1 are shaded. Contours of −0.5 mm day−1 are added. Zero contours are omitted.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 9.
Fig. 9.

Same as Fig. 4, but for Jul–Sep averaged from 1979 to 2001.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 10.
Fig. 10.

Same as Fig. 5, but for Jul–Sep.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 11.
Fig. 11.

Same as Fig. 6, but for Jul–Sep.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 12.
Fig. 12.

(a) Surface zonal winds averaged for Jun–Jul 2000 from the satellite QuikSCAT winds. Contour interval is 1 m s−1. Values less than −10 m s−1 are shaded. (b) Same as in (a), but for the RR. (c), (d) Same as in (a), (b), respectively, but for 2001.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 13.
Fig. 13.

Monthly mean vertically integrated meridional moisture flux () for (a) Jun, (b) Jul, (c) Aug, and (d) Sep averaged from 1995 to 2000 from the RR. The contour interval is 30 kg (ms)−1. Values greater than 60 kg (ms)−1 [90 kg (ms)−1] are shaded light (dark). (e)–(h) Same as (a)–(d), but for the difference of () between the RR and the operational EDAS. Contour interval is 30 kg (ms)−1. Values greater than 60 kg (ms)−1 are shaded.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 14.
Fig. 14.

(a) Vertical cross section of wind speed at 36°N, 97.5°W from the RR from 17 to 31 Aug 1994. Contour interval is 2 m s−1. Values greater than 8 m s−1 are shaded. (b) Vertical profile of the meridional wind at 36°N, 97.5°W from the RR for Jun–Aug (JJA) 1994 at each synoptic time. Contour interval is 1 m s−1. Positive values are shaded.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 15.
Fig. 15.

(a) Time series of 3-hourly meridional winds at 975 hPa at a grid point near Puerto Peñasco (open circles). The dark line is the eight-point running mean. (b) Vertical profile of meridional wind (m s−1) at Puerto Peñasco for each synoptic time: 0000 UTC (1600 LT, dark squares), 0300 UTC (1900 LT, crosses), 0600 UTC (2200 LT, triangles), 0900 UTC (0100 LT, dark line), 1200 UTC (0400 LT, pluses), 1500 UTC (0700 LT, open circles), 1800 UTC (1000 LT, dark circles), and 2100 UTC (1300 LT, open squares) averaged from 31 Jul–14 Aug 1995.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 16.
Fig. 16.

(a) Same as Fig. 15a, but from observations; (b) same as Fig. 15b, but from observations [reproduced from Fig. 3 of Douglas et al. (1998)].

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Fig. 17.
Fig. 17.

Vertical cross section of meridional moisture flux () at 30°N for the period 1998–2000 from the RR for (a) Jul, (b) Aug, and (c) Sep. Contour interval is 10 g kg−1 m s−1. Zero contours are omitted. Contours of −5 and 5 g kg−1 m s−1 are added. (d)–(f) Same as (a)–(c), but for the operational EDAS.

Citation: Journal of Hydrometeorology 6, 5; 10.1175/JHM452.1

Save