• Auad, G., , Miller A. J. , , Roads J. O. , , and Cayan D. , 2001: Pacific Ocean wind stress and surface heat flux anomalies from NCEP reanalysis and observation: Cross-statistics and ocean model responses. J. Geophys. Res., 106, 22 24922 265, doi:10.1029/2000JC000264.

    • Search Google Scholar
    • Export Citation
  • Baldwin, M. E., , and Mitchell K. E. , 1997: The NCEP hourly multi-sensor U.S. precipitation analysis for operations and GCIP research. Preprints, 13th Conf. on Hydrology, Long Beach, CA, Amer. Meteor. Soc., 5455.

  • Barlage, M., and et al. , 2010: Noah land surface model modifications to improve snowpack prediction in the Colorado Rocky Mountains. J. Geophys. Res., 115, D22101, doi:10.1029/2009JD013470.

    • Search Google Scholar
    • Export Citation
  • Barlage, M., , Tewari M. , , Chen F. , , Manning K. , , and Miguez-Macho G. , 2013: North American regional climate simulations with WRF/Noah-MP validation and the effect of ground water interaction. Proc. 14th WRF User's Workshop, Boulder, CO, WRF. [Available online at www2.mmm.ucar.edu/wrf/users/workshops/WS2013/ppts/5B.2.pdf.]

    • Search Google Scholar
    • Export Citation
  • Berbery, E. H., , and Rasmusson E. M. , 1999: Mississippi moisture budgets on regional scales. Mon. Wea. Rev.,127, 2654–2673, doi:10.1175/1520-0493(1999)127<2654:MMBORS>2.0.CO;2.

  • Berbery, E. H., , Luo Y. , , Mitchell K. , , and Betts A. K. , 2003: Eta model–estimated land surface processes and the hydrologic cycle of the Mississippi basin. J. Geophys. Res., 108, 8852, doi:10.1029/2002JD003192.

    • Search Google Scholar
    • Export Citation
  • Betts, A., , Hong S.-Y. , , and Pan H.-L. , 1996: Comparison of NCEP–NCAR reanalysis with 1987 FIFE data. Mon. Wea. Rev., 124, 14801498, doi:10.1175/1520-0493(1996)124<1480:CONNRW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Betts, A., , Viterbo P. , , and Wood E. , 1998: Surface energy and water balance for the Arkansas–Red River basin from the ECMWF reanalysis. J. Climate, 11, 28812897, doi:10.1175/1520-0442(1998)011<2881:SEAWBF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Betts, A., , Ball J. H. , , and Viterbo P. , 1999: Basin-scale surface water and energy budgets for the Mississippi from the ECMWF reanalysis. J. Geophys. Res., 104, 19 29319 306, doi:10.1029/1999JD900056.

    • Search Google Scholar
    • Export Citation
  • Black, T., 1994: The new NMC mesoscale Eta model: Description and forecast examples. Wea. Forecasting, 9, 265278, doi:10.1175/1520-0434(1994)009<0265:TNNMEM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Brunke, M. A., , Wang Z. , , and Zeng X. , 2011: An assessment of the uncertainties in ocean surface turbulent fluxes in 11 reanalysis, satellite-derived, and combined global datasets. J. Climate,24, 5469–5493, doi:10.1175/2011JCLI4223.1.

  • Campana, K., , and Caplan P. , 2005: Technical procedures bulletin for the T382 Global Forecast System. NOAA/NCEP/Environmental Modeling Center. [Available online at www.emc.ncep.noaa.gov/gc_wmb/Documentation/TPBoct05/T382.TPB.FINAL.htm.]

  • Chen, F., , Mitchell K. , , Schaake J. , , Xue Y. , , Pan H.-L. , , Koren V. , , Duan Q. , , and Betts A. , 1996: Modeling of land surface evaporation by four schemes and comparison with FIFE observations. J. Geophys. Res., 101, 7251–7268, doi:10.1029/95JD02165.

    • Search Google Scholar
    • Export Citation
  • Chen, M., , Shi W. , , Xie P. , , Silva V. B. S. , , Kousky V. E. , , Wayne Higgins R. , , and Janowiak J. E. , 2008: Assessing objective techniques for gauge-based analyses of global daily precipitation. J. Geophys. Res., 113, D04110, doi:10.1029/2007JD009132.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and et al. , 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597, doi:10.1002/qj.828.

    • Search Google Scholar
    • Export Citation
  • Dole, R., , Hoerling M. , , and Schubert S. , Eds., 2008: Reanalysis of historical climate data for key atmospheric features: Implications for attribution of causes of observed change. U.S. Climate Change Science Program Synthesis and Assessment Product 1.3, NOAA, Asheville, NC, 156 pp.

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

    • Search Google Scholar
    • Export Citation
  • Gochis, D. J., , and Chen F. , 2003: Implementation and testing of a grid-based routing scheme in the Noah land surface model. EGS–AGU–EUG Joint Assembly, Nice, France, EGU, Abstract 799.

  • Helfrich, S. R., , McNamara D. , , Ramsay B. H. , , Baldwin T. , , and Kasheta T. , 2007: Enhancements to and forthcoming developments in the Interactive Multisensor Snow and Ice Mapping System (IMS). Hydrol. Process., 21, 1576–1586, doi:10.1002/hyp.6720.

  • Hollinger, S. E., , and Isard S. A. , 1994: A soil moisture climatology of Illinois. J. Climate, 7, 822827, 833, doi:10.1175/1520-0442(1994)007<0822:ASMCOI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Holtslag, A. A. M., , and Ek M. , 1996: Simulation of surface fluxes and boundary layer development over the pine forest in HAPEX-MOBILHY. J. Appl. Meteor., 35, 202213, doi:10.1175/1520-0450(1996)035<0202:SOSFAB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Jacquemin, B., , and Noilhan J. , 1990: Sensitivity study and validation of a land surface parameterization using the HAPEX-MOBILHY dataset. Bound.-Layer Meteor., 52, 93134, doi:10.1007/BF00123180.

    • Search Google Scholar
    • Export Citation
  • Jarvis, P. G., 1976: The interpretation of the variations in leaf water potential and stomatal conductance found in canopies in the field. Philos. Trans. Roy. Soc. London.,B273, 593–610, doi:10.1098/rstb.1976.0035.

    • Search Google Scholar
    • Export Citation
  • Ji, M., , Leetmaa A. , , and Derber J. , 1995: An ocean analysis system for seasonal to interannual climate studies. Mon. Wea. Rev., 123, 460481, doi:10.1175/1520-0493(1995)123<0460:AOASFS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Josey, S. A., 2001: A Comparison of ECMWF, NCEP–NCAR, and SOC surface heat fluxes with moored buoy measurements in the subduction region of the northeast Atlantic. J. Climate, 14, 17801789, doi:10.1175/1520-0442(2001)014<1780:ACOENN>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., 2000: The US reanalysis program: Climatology for the new millennium. Rep. of the First Planning Workshop for the Next US Global Reanalysis, University of Maryland, College Park, College Park, MD, 19 pp. [Available online at http://old.usclivar.org/Pubs/ReanalysisWorkshop_Rep.pdf.]

  • Kalnay, E., and et al. , 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437470, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kanamitsu, M., and et al. , 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 16311643, doi:10.1175/BAMS-83-11-1631.

    • Search Google Scholar
    • Export Citation
  • Kistler, R., and et al. , 2001: The NCEP–NCAR 50-Year Reanalysis: Monthly means CD-ROM and documentation. Bull. Amer. Meteor. Soc., 82, 247267, doi:10.1175/1520-0477(2001)082<0247:TNNYRM>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kopp, T. J., , and Kiess R. B. , 1996: The Air Force Global Weather Central snow analysis model. Preprints, 15th Conf. on Weather Analysis and Forecasting, Norfolk, Virginia, Amer. Meteor. Soc., 220222.

  • Koren, V., , Schaake J. , , Mitchell K. , , Duan Q.-Y. , , Chen F. , , and Baker J. M. , 1999: A parameterization of snowpack and frozen ground intended for NCEP weather and climate models. J. Geophys. Res., 104, 19 56919 585, doi:10.1029/1999JD900232.

    • Search Google Scholar
    • Export Citation
  • Koster, R., and et al. , 2004: Regions of strong coupling between soil moisture and precipitation. Science,305, 1138–1140, doi:10.1126/science.1100217.

  • Koster, R., , Guo Z. , , Yang R. , , Dirmeyer P. A. , , Mitchell K. E. , , and Puma M. J. , 2009: On the nature of soil moisture in land surface models. J. Climate, 22, 43224335, doi:10.1175/2009JCLI2832.1.

    • Search Google Scholar
    • Export Citation
  • Liang, X., , Lettenmaier D. P. , , Wood E. F. , , and Burges S. J. , 1994: A simple hydrologically based model of land surface water and energy fluxes for GSMs. J. Geophys. Res., 99, 14 41514 428, doi:10.1029/94JD00483.

    • Search Google Scholar
    • Export Citation
  • Lin, Y., , Mitchell K. E. , , Rogers E. , , and DiMego G. J. , 2005: Using hourly and daily precipitation analyses to improve model water budget. Preprints, Ninth Symp. on Integrated Observing and Assimilation Systems for the Atmosphere, Oceans, and Land Surface, San Diego, CA, Amer. Meteor. Soc., 3.3. [Available online at http://ams.confex.com/ams/pdfpapers/84484.pdf.]

  • Livneh, B., , Xia Y. , , Mitchell K. E. , , Ek M. B. , , and Lettenmaier D. P. , 2010: Noah LSM snow model diagnostics and enhancements. J. Hydrometeor., 11, 721738, doi:10.1175/2009JHM1174.1.

    • Search Google Scholar
    • Export Citation
  • Lohmann, D., and et al. , 2004: Streamflow and water balance intercomparison of four land surface models in the North American Land Data Assimilation Systems project. J. Geophys. Res.,109, D07S91, doi:10.1029/2003JD003517.

  • Luo, Y., , Berbery H. , , Mitchell K. , , and Betts A. , 2007: Relationships between land surface and near-surface atmospheric variables in the NCEP North American Regional Reanalysis. J. Hydrometeor., 8, 11841203, doi:10.1175/2007JHM844.1.

    • Search Google Scholar
    • Export Citation
  • Mahrt, L., , and Ek M. , 1984: The influence of atmospheric stability on potential evaporation. J. Climate Appl. Meteor., 23, 222234, doi:10.1175/1520-0450(1984)023<0222:TIOASO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Mahrt, L., , and Pan H.-L. , 1984: A two-layer model of soil hydrology. Bound.-Layer Meteor., 29, 120, doi:10.1007/BF00119116.

  • Matson, M., , Ropelewski C. F. , , and Varnadore M. S. , 1986: An Atlas of Satellite-Derived Northern Hemisphere Snow Cover Frequency. National Weather Service, 75 pp.

  • Maurer, E. P., , O’Donnell G. , , Lettenmaier D. , , and Roads J. , 2001: Evaluation of the land surface water budget in NCEP/NCAR and NCEP/DOE reanalyses using an off-line hydrologic model. J. Geophys. Res., 106, 17 84117 862, doi:10.1029/2000JD900828.

    • Search Google Scholar
    • Export Citation
  • McNally, A. P., , Derber J. C. , , Wu W.-S. , , and Katz B. B. , 2006: The use of TOVS level-1 radiances in the NCEP SSI analysis system. Quart. J. Roy. Metor. Soc, 126, 689724, doi:10.1002/qj.49712656315.

    • Search Google Scholar
    • Export Citation
  • Meng, J., , Yang R. , , Wei H. , , Ek M. , , Gayno G. , , Xie P. , , and Mitchell K. , 2012: The land surface analysis in the NCEP Climate Forecast System Reanalysis. J. Hydrometeor.,13, 1621–1630, doi:10.1175/JHM-D-11-090.1.

  • Mesinger, F., , Janjić Z. I. , , Nickovic S. , , Gavrilov D. , , and Deaven D. G. , 1988: The step-mountain coordinate: Model description and performance for cases of alpine lee cyclogenesis and for a case of an Appalachian redevelopment. Mon. Wea. Rev., 116, 1493–1518, doi:10.1175/1520-0493(1988)116<1493:TSMCMD>2.0.CO;2.

  • Mesinger, F., and et al. , 2006: North American Regional Reanalysis. Bull. Amer. Meteor. Soc., 87, 343360, doi:10.1175/BAMS-87-3-343.

    • Search Google Scholar
    • Export Citation
  • Milly, P. C. D., , and Dunne K. A. , 2001: Trends in evaporation and surface cooling is the Mississippi River basin. Geophys. Res. Lett.,28, 1219–1222, doi:10.1029/2000GL012321.

  • Mitchell, K., and et al. , 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.

  • Music, B., , and Caya D. , 2007: Evaluation of the hydrological cycle over the Mississippi River basin as simulated by the Canadian Regional Climate Model (CRCM). J. Hydrometeor., 8, 969988, doi:10.1175/JHM627.1.

    • Search Google Scholar
    • Export Citation
  • Niu, G.-Y., , Yang Z.-L. , , Dickinson R. E. , , Gulden L. E. , , and Su H. , 2007: Development of a simple groundwater model for use in climate models and evaluation with Gravity Recovery and Climate Experiment data. J. Geophys. Res., 112, D07103, doi:10.1029/2006JD007522.

    • Search Google Scholar
    • Export Citation
  • Niu, G.-Y., and et al. , 2011: The community Noah land surface model with multi-parameterization options (Noah-MP): 1. Model description and evaluation with local-scale measurements. J. Geophys. Res., 116, D12109, doi:10.1029/2010JD015139.

    • Search Google Scholar
    • Export Citation
  • Noilhan, J., , and Planton S. , 1989: A simple parameterization of land surface processes for meteorological models. Mon. Wea. Rev., 117, 536549, doi:10.1175/1520-0493(1989)117<0536:ASPOLS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Onogi, K., and et al. , 2007: The JRA-25 Reanalysis. J. Meteor. Soc. Japan, 85, 369432, doi:10.2151/jmsj.85.369.

  • Pan, H.-L., , and Mahrt L. , 1987: Interaction between soil hydrology and boundary-layer development. Bound.-Layer Meteor., 38, 185202, doi:10.1007/BF00121563.

    • Search Google Scholar
    • Export Citation
  • Peixóto, J. P., , and Oort A. H. , 1983: The atmospheric branch of the hydrological cycle and climate. Variations of the Global Water Budget, D. Reidel, 5–65, doi:10.1007/978-94-009-6954-4_2.

  • Rienecker, M., and et al. , 2011: MERRA: NASA’s Modern-Era Retrospective Analysis for Research and Applications. J. Climate, 24, 36243648, doi:10.1175/JCLI-D-11-00015.1.

    • Search Google Scholar
    • Export Citation
  • Roads, J., , and Betts A. , 2000: NCEP–NCAR and ECMWF reanalysis surface water and energy budgets for the Mississippi River basin. J. Hydrometeor., 1, 8894, doi:10.1175/1525-7541(2000)001<0088:NNAERS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Roads, J., , Chen S. , , Kanamitsu M. , , and Juang H. , 1999: Surface water characteristics in the NCEP global spectral model and reanalysis. J. Geophys. Res., 104, 19 30719 327, doi:10.1029/98JD01166.

    • Search Google Scholar
    • Export Citation
  • Roads, J., , Kanamitsu M. , , and Stewart R. , 2002: CSE water and energy budgets in the NCEP–DOE Reanalysis II. J. Hydrometeor., 3, 227248, doi:10.1175/1525-7541(2002)003<0227:CWAEBI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Roads, J., and et al. , 2003: GCIP water and energy synthesis (WEBS). J. Geophys. Res.,108, 8609, doi:10.1029/2002JD002583.

  • Robock, A., and et al. , 2000: The Global Soil Moisture Data Bank. Bull. Amer. Meteor. Soc.,81, 1281–1299, doi:10.1175/1520-0477(2000)081<1281:TGSMDB>2.3.CO;2.

  • Rodell, M., and et al. , 2004: The Global Land Data Assimilation System. Bull. Amer. Meteor. Soc., 85, 381394, doi:10.1175/BAMS-85-3-381.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and et al. , 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc.,91, 1015–1057, doi:10.1175/2010BAMS3001.1.

  • Saha, S., and et al. , 2014: The NCEP Climate Forecast System version 2. J. Climate, 27, 21852208, doi:10.1175/JCLI-D-12-00823.1.

  • Schaake, J. C., , Koren V. I. , , Duan Q.-Y. , , Mitchell K. , , and Chen F. , 1996: Simple water balance model for estimating runoff at different spatial and temporal scales. J. Geophys. Res., 101, 74617475, doi:10.1029/95JD02892.

    • Search Google Scholar
    • Export Citation
  • Seager, R., , and Henderson N. , 2013: Diagnostic computation of moisture budgets in the ERA-Interim Reanalysis with reference to analysis of CMIP-archived atmospheric model data. J. Climate, 26, 78767901, doi:10.1175/JCLI-D-13-00018.1.

    • Search Google Scholar
    • Export Citation
  • Sheffield, J., , Livneh B. , , and Wood E. F. , 2012: Representation of terrestrial hydrology and large-scale drought of the continental United States from the North American Regional Reanalysis. J. Hydrometeor., 13, 856876, doi:10.1175/JHM-D-11-065.1.

    • Search Google Scholar
    • Export Citation
  • Srinivasan, G., and et al. , 2000: Soil moisture simulations in revised AMIP models. J. Geophys. Res., 105, 26 63526 644, doi:10.1029/2000JD900443.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , Fasullo J. T. , , and Mackaro J. , 2011: Atmospheric moisture transports from ocean to land and global energy flows in reanalyses. J. Climate,24, 4907–4924, doi:10.1175/2011JCLI4171.1.

  • Uppala, S., and et al. , 2005: The ERA-40 re-analysis. Quart. J. Roy. Meteor., 131, 29613012, doi:10.1256/qj.04.176.

  • Wang, F., , Wang L. , , Koike T. , , Zhou H. , , Yang K. , , Wang A. , , and Li W. , 2011: Evaluation and application of a fine-resolution global data set in a semiarid mesoscale river basin with a distributed biosphere hydrological model. J. Geophys. Res., 116, D21108, doi:10.1029/2011JD015990.

    • Search Google Scholar
    • Export Citation
  • Wang, L., and et al. , 2009: Development of a distributed biosphere hydrological model and its evaluation with the Southern Great Plains Experiments (SGP97 and SGP99). J. Geophys. Res., 114, D08107, doi:10.1029/2008JD010800.

    • Search Google Scholar
    • Export Citation
  • Xie, P., , and Arkin P. A. , 1997: Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor. Soc., 78, 2539–2558, doi:10.1175/1520-0477(1997)078<2539:GPAYMA>2.0.CO;2.

  • Xie, P., , Chen M. , , Yatagai A. , , Hayasaka T. , , Fukushima Y. , , and Yang S. , 2007: A gauge based analysis of daily precipitation over East Asia. J. Hydrometeor., 8, 607626, doi:10.1175/JHM583.1.

    • Search Google Scholar
    • Export Citation
  • Yang, R., , Mitchell K. E. , , Meng J. , , and Ek M. , 2011: Summer-season forecast experiments with the NCEP Climate Forecast System using different land models and different initial land states. J. Climate,24, 2319–2334, doi:10.1175/2010JCLI3797.1.

  • View in gallery

    Seasonal water budget terms averaged over the Mississippi River basin from CFSR, NARR, and GR2 and comparison with observations and GLDAS: (a) P, (b) R, (c) E, (d) surface water tendency [−d(W + S)/dt; squares are surface water increment and triangles are soil moisture increment in CFSR from the analysis cycle], (e) surface water (W + S), and (f) unbalanced surface water [Res(W + S)]. In (a) and (b), the observed seasonal precipitation is based on the offline GLDAS forcing from Jan 1979 to Dec 2009, and the observed runoff climatology is based on the streamflow at Vicksburg, MS, from Oct 1979 to Sep 1999.

  • View in gallery

    (a) Seasonal soil moisture averaged over the state of Illinois from the three reanalyses, observations, and GLDAS (mm); and (b) the corresponding seasonal precipitation (mm day−1) from Jan 1981 to Dec 2004.

  • View in gallery

    As in Fig. 1, but for monthly anomalies of surface water budget terms from Jan 1984 to Dec 1993 (3-month running means for the presentation only) for CFSR, NARR, GR2, and observations [in (a) and (b) only].

  • View in gallery

    Seasonal surface energy budget terms averaged over the Mississippi River basin from CFSR, NARR, and GR2 and comparison with observations [in (a),(b), and (e) only]: (a) SW, (b) LW, (c) −SH, (d) −LH, (e) Ts, and (f) residual (the observed SW, LW, and Ts are from Jul 1983 to Dec 2007). The units are W m−2 for SW, LW, SH, LH, and residual, and kelvins for Ts.

  • View in gallery

    As in Fig. 4, but for monthly anomalies of surface energy budget terms from Jan 1984 to Dec 1993 (3-month running means for presentation only).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 47 47 5
PDF Downloads 29 29 0

Surface Water and Energy Budgets for the Mississippi River Basin in Three NCEP Reanalyses

View More View Less
  • 1 Environmental Modeling Center, NOAA/NWS/NCEP, College Park, and I.M. Systems Group, Inc., Rockville, Maryland
  • | 2 Environmental Modeling Center, NOAA/NWS/NCEP, College Park, Maryland
  • | 3 Environmental Modeling Center, NOAA/NWS/NCEP, College Park, and I.M. Systems Group, Inc., Rockville, Maryland
© Get Permissions
Full access

Abstract

Surface water and energy budgets from the National Centers for Environmental Prediction–U.S. Department of Energy (NCEP–DOE) Atmospheric Model Intercomparison Project (AMIP-II) Global Reanalysis 2 (GR2), the North American Regional Reanalysis (NARR), and the NCEP Climate Forecast System Reanalysis (CFSR) are compared here with each other and with available observations over the Mississippi River basin. The comparisons in seasonal cycle, interannual variation, and annual mean over a 31-yr period show that there are a number of noticeable differences and similarities in the large-scale basin averages. Warm season precipitation and runoff in the GR2 are too large compared to the observations, and seasonal surface water variation is small. By contrast, the precipitation in both NARR and CFSR is more reasonable and in better agreement with the observation, although the corresponding seasonal runoff is very small. The main causes of the differences in both surface parameterization and approach used in assimilating the observed precipitation datasets and snow analyses are then discussed. Despite the discrepancies in seasonal water budget components, seasonal energy budget terms in the three reanalyses are close to each other and to available observations. The interannual variations in both water and energy budgets are comparable. This study shows that the CFSR achieves a large improvement over the GR2, indicating that the CFSR dataset can be used in climate variability studies. Nonetheless, improved land surface parameterization schemes and data assimilation techniques are needed to depict the surface water and energy climates better, in particular, the variation in seasonal runoff.

Corresponding author address: Dr. Rongqian Yang, Environmental Modeling Center, National Centers for Environmental Prediction, NOAA Center for Weather and Climate Prediction, 5830 University Research Ct., College Park, MD 20740. E-mail: rongqian.yang@noaa.gov

Abstract

Surface water and energy budgets from the National Centers for Environmental Prediction–U.S. Department of Energy (NCEP–DOE) Atmospheric Model Intercomparison Project (AMIP-II) Global Reanalysis 2 (GR2), the North American Regional Reanalysis (NARR), and the NCEP Climate Forecast System Reanalysis (CFSR) are compared here with each other and with available observations over the Mississippi River basin. The comparisons in seasonal cycle, interannual variation, and annual mean over a 31-yr period show that there are a number of noticeable differences and similarities in the large-scale basin averages. Warm season precipitation and runoff in the GR2 are too large compared to the observations, and seasonal surface water variation is small. By contrast, the precipitation in both NARR and CFSR is more reasonable and in better agreement with the observation, although the corresponding seasonal runoff is very small. The main causes of the differences in both surface parameterization and approach used in assimilating the observed precipitation datasets and snow analyses are then discussed. Despite the discrepancies in seasonal water budget components, seasonal energy budget terms in the three reanalyses are close to each other and to available observations. The interannual variations in both water and energy budgets are comparable. This study shows that the CFSR achieves a large improvement over the GR2, indicating that the CFSR dataset can be used in climate variability studies. Nonetheless, improved land surface parameterization schemes and data assimilation techniques are needed to depict the surface water and energy climates better, in particular, the variation in seasonal runoff.

Corresponding author address: Dr. Rongqian Yang, Environmental Modeling Center, National Centers for Environmental Prediction, NOAA Center for Weather and Climate Prediction, 5830 University Research Ct., College Park, MD 20740. E-mail: rongqian.yang@noaa.gov

1. Introduction

The process of optimal combination of short-term model predictions with observations, known as four-dimensional data assimilation (FDDA), is a critical element of weather prediction and climate analysis systems. The quality of a given data assimilation system depends upon how the physical processes are represented in the assimilation model, how accurate the uncertainty about the model physical parameterizations is characterized, the availability and quality of observational data, and the approach in which the observations are utilized. For these reasons, advances in the model physics, more available observations, and improvements on the analysis schemes almost always prompt a necessity that data from the recent past are reanalyzed using the state-of-the-art “frozen” data assimilation system to develop a climate record and understand climate variations.

With differences in geographic coverage, the reanalysis can be performed over the entire earth or over a specific region to make use of more quality observations. With differences in the components of the earth’s climate system involved, the reanalysis can be done in a fully coupled mode, to better understand the nature of ocean–land–atmosphere coupling; in a partially coupled mode, where the ocean is prescribed to study the characteristics of land–atmosphere interaction; or in an offline mode, where the land component is driven by surface observations to provide a useful benchmark for evaluation of the reanalysis surface fluxes and to derive realistic land states to initialize climate models.

Focusing on the land and atmosphere components of the earth system, there are a few atmospheric reanalyses developed over the past. Examples of these reanalyses include the 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40; Uppala et al. 2005), the ECMWF interim reanalysis (ERA-Interim; Dee et al. 2011), the NASA Modern-Era Retrospective Analysis for Research and Applications (MERRA; Rienecker et al. 2011), the Japanese 25-year Reanalysis Project (JRA-25; Onogi et al. 2007), the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) Global Reanalysis 1 (referred to as GR1; Kalnay et al. 1996), the NCEP–U.S. Department of Energy (NCEP–DOE) Atmospheric Model Intercomparison Project (AMIP-II) Global Reanalysis 2 (referred to as GR2; Kanamitsu et al. 2002), and the North America Regional Reanalysis (NARR; Mesinger et al. 2006). From the land perspective, with varying geographic coverage, there are also several existing Land Data Assimilation Systems (LDASs), such as the North American Land Data Assimilation System (NLDAS; Mitchell et al. 2004) and the Global Land Data Assimilation System (GLDAS; Rodell et al. 2004).

Reanalyses are imperfect because they must rely on imperfect models to augment scarce observations. The deficiencies found in the existing global reanalyses have limited the usefulness of these reanalysis products. For instance, spurious seasonal and interannual variations in soil moisture found in the GR1 (Roads et al. 1999; Srinivasan et al. 2000; Maurer et al. 2001) and summertime precipitation bias found in the GR2 (Betts et al. 1996; Roads et al. 2002; Music and Caya 2007) prevent the datasets from being used in climate change study and land surface hydrologic research. The problems of the atmospheric model in estimating ocean surface forcing, such as precipitation, sensible heat, and evaporation (Auad et al. 2001; Josey 2001; Kanamitsu et al. 2002; Brunke et al. 2011) make it unsuitable to use these products to drive ocean models. The problems have been partially addressed by limiting the reanalysis domain (e.g., NARR) and decreasing the components of the earth system involved (e.g., LDAS). However “climate change,” an important application of a reanalysis (Dole et al. 2008), includes many complex ocean–land–atmosphere interactions in a global setting. The coupled nature of the interactions itself requires that the reanalysis should cover all components of the earth system (Kalnay 2000).

With these issues and demands in mind, the NCEP has, for the first time, developed the Climate Forecast System Reanalysis (CFSR; Saha et al. 2010). The CFSR is designed and executed as a global, high-resolution, coupled ocean–land–atmosphere system to provide the best estimate of the state of these coupled components over a long period. The CFSR atmospheric model contains observed solar variation and changes in carbon dioxide, aerosols, and trace gases, which account for the most of climate variability. CFSR also incorporates the operational NCEP Global Ocean Data Assimilation System (GODAS; Ji et al. 1995) and an advanced GLDAS (Meng et al. 2012). With release of the CFSR dataset, it is desirable to conduct an analysis to examine how well the new generation of global reanalysis performs in depicting the water and energy budgets. An examination of the reanalysis is necessary to understand the characteristics of water and energy climate and to point out the limitations and issues that still affect the ability to develop adequate budgets.

Assessing surface energy and water balances for the Mississippi River basin has been a key objective and a focused region of the past Global Energy and Water Cycle Experiment (GEWEX) Continental Scale International Project (GCIP). Using data from a variety of sources, the basin budgets have been studied extensively, for example, by Betts et al. (1998, 1999) using the ECMWF reanalysis, Roads et al. (1999) using the GR2, Berbery and Rasmusson (1999) and Berbery et al. (2003) using Eta Data Assimilation System (EDAS) products (Mesinger et al. 1988; Black 1994), Luo et al. (2007) using the NARR, and Milly and Dunne (2001) using synthetic observational data. Roads and Betts (2000) also carried out a comparison study using the GR2 and the ECMWF datasets. Based on the data from 1996 to 1999, Roads et al. (2003) provided a comprehensive description of the GCIP Water and Energy Budget Synthesis (WEBS) by summarizing the estimates of several models as well as data from global and regional reanalyses. They concluded that, in spite of qualitative agreement between the modeled and observed water budgets, there is still much quantitative uncertainty.

Concentrating on the three major NCEP reanalyses (GR2, NARR, and CFSR), in this study, these efforts are combined and the large-scale averages of surface water and energy budget components from the reanalyses are compared explicitly with each other and with available observations to assess the ability of forecast models under (appropriate) observational constraints to estimate the energy and water balances for the large Mississippi River basin. Comparisons between the reanalyses and with measurements can yield new insights. The comparison between CFSR and GR2 will reveal its general improvement over the previous global reanalysis, and the comparison between CFSR and NARR will give us an idea of how the new reanalysis is the best dataset to study water climate over the North America, whereas the comparisons with the observations will show how the three reanalyses are close to the reality, which can serve as a basis for future reanalysis enhancements.

Surface water and energy processes are physically coupled. Observational constraints applied by data assimilation techniques will propagate throughout the system via physical processes. As such, surface water and energy budget terms in a reanalysis are not only a function of model physical parameterizations, but also a function of the quality of assimilated observations and data assimilation technique used in the data assimilation system. The differences in land model physical parameterizations (section 2), ingested observational data, and assimilation approach (section 3) lead to notable differences in the surface water and energy budgets (section 5). A description of the reanalysis and verification datasets used in this study is given in section 4. The main findings and a brief discussion are provided in section 6.

2. Land surface model

a. OSU LSM

There are two land surface models (LSMs) used in the three reanalyses. The first is the Oregon State University (OSU) LSM developed in the 1980s (Mahrt and Pan 1984; Pan and Mahrt 1987). The vertical configuration of the OSU LSM in GR2 has two layers (at depths of 10 and 200 cm). It uses thermal conduction equations for soil temperature and a form of Richardson’s equation for soil moisture. The surface skin temperature is a diagnostic quantity from the surface energy balance. The effect of stomatal control is represented via a “plant coefficient” and the Penman–Monteith potential evaporation (Mahrt and Ek 1984). The plant coefficient that accounts for stomatal control is related to canopy conductance using the common “big leaf” approach (Jarvis 1976; Noilhan and Planton 1989; Holtslag and Ek 1996), where the canopy conductance is modeled as a function of soil moisture availability and atmospheric conditions (solar insolation, temperature, and humidity). The evaporation at the surface has three components: the direct evaporation from the bare soil, transpiration from vegetation, and evaporation from the canopy partitioned by the vegetation cover that varies geographically as well as seasonally. Total runoff consists of surface storm overflow and base flow at the bottom of the soil layer. The surface runoff occurs when soil moisture exceeds a determined maximum value (porosity of 0.47; in volumetric). Some other land surface characteristics of the OSU LSM in GR2 such as vegetation cover (70%), soil wilting point (0.12; in volumetric), and vegetation root zone depth (2.0 m) are fixed at all grid cells.

b. Noah LSM

The second LSM is Noah (Chen et al. 1996; Ek et al. 2003). It originates from the OSU LSM, but it is equipped with many physical improvements and updates added by researchers. The Noah LSM uses an explicit canopy resistance formulation used by Jacquemin and Noilhan (1990) for a more realistic representation of canopy transpiration; the simple soil–water balance (SWB) model of Schaake et al. (1996) to account for subgrid spatial variability in soil moisture, precipitation, and runoff; and a frozen soil physics package (Koren et al. 1999). Other updates include improved soil and snow thermal conductivity, spatially varying root depth (vegetation class dependent), improved seasonal vegetation cover, and snowpack physics, among others. The Noah LSM has the same soil thickness of 200 cm, but it is configured with four vertical soil layers (at depths of 10, 30, 60, and 100 cm). Compared to the OSU LSM, these physical changes generally lead to higher canopy resistance, better representation of runoff, and realistic soil moisture. The Noah LSM is used in both CFSR and NARR.

3. Assimilation technique

There are two approaches used to assimilate surface observations in the three reanalyses, namely, the nudging method and the initialization method. The basic idea of the nudging technique is to force the model-predicted quantity toward available measurement or a predetermined climatology with some relaxation terms. The initialization approach, on the other hand, focuses on the model physical processes where necessary adjustments are made to the closed fields in background model until the predicted quantity is in agreement with the observation. In essence, the initialization approach is another type of nudging scheme. The only difference is that it is done from the atmospheric part and involves more than one field. The choice of a nudging technique mainly stems from the fact that it is computationally much cheaper and easily allows for operational implementation.

The most important variables to be assimilated over the land surface are precipitation and snow, as both have a great impact on soil moisture, which plays a critical role in regulating land surface water and energy processes. It is common to control soil moisture so that it does not drift to an unrealistic state. Depending on the availability and quality of ingested data, the corrections made to both soil moisture and snow depth are done quite differently in the three reanalyses.

a. GR2

The nudging method is adopted by the GR2. In GR2, the precipitation forcing at land surface is from its parent atmospheric model. To prevent long-term climate drift of soil moisture, the GR2 first computes 5-day mean precipitation. The average is then compared with the corresponding observed “pentad” precipitation dataset of the Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP; Xie and Arkin 1997) that merges satellite and gauge measurements on a 2.5° × 2.5° latitude–longitude grid. The difference is used as a soil moisture correction term that is applied evenly throughout the subsequent 5-day period. In case of no model runoff, such a correction of either sign is generally applied to the topmost soil layer (0–10 cm), except for the case where it is necessary to adjust the second soil layer (10–200 cm) when the top layer saturates or dries. In case of model runoff, when the water seeping into the soil (i.e., model precipitation minus runoff) is less than observed precipitation, no correction is necessary, as any model precipitation errors are assumed to affect the runoff rather than the soil moisture. When the observed precipitation is less than the model precipitation minus runoff, then the soil moisture is reduced by the difference of the two quantities. This correction method is equivalent to using observed precipitation in the hydrological calculations, except that the correction is delayed by 5 days. This delay could be reduced to 1 day by using daily precipitation. However, only the pentad-mean precipitation was available for the entire period of the reanalysis when the GR2 project was started in 1998.

Water equivalent snow depth is handled on a weekly basis, as only a weekly Northern Hemisphere analysis of snow cover (without snow depth; Matson et al. 1986) was available for ingest. In the GR2, the model snow depth was used (including model-predicted snow accumulation) where it was consistent with the input snow cover analysis. When averaged model snow cover (1 cm is used as a threshold for a complete cover) disagrees with observation, the model snow depth was adjusted to the snow cover analysis by either removing or adding the model snow without affecting the soil moisture. The GR2 is run on T62 Gaussian grid (~200 km) with 28 vertical levels.

b. NARR

Unlike the direct use of observed precipitation in GR2, the soil moisture control in NARR is achieved via an initialization approach, where the difference between modeled and observed precipitation is used to adjust (scale) the background atmospheric model’s latent heat profile, water vapor mixing ratio, and cloud-related fields on an hourly basis [see Lin et al. (2005) for details]. The observed hourly precipitation is mainly developed from the real-time, 4-km U.S. precipitation analysis (Baldwin and Mitchell 1997); hourly precipitation estimates from the Weather Surveillance Radar-1988 Doppler (WSR-88D) network; and the hourly reports from the approximately 2500 automated rain gauges. The resulting hourly model precipitation is used to drive the land. In addition to precipitation, the NARR also assimilates near-surface observations (e.g., 2-m temperatures and 10-m winds). Another important feature of NARR compared to GR2 is the direct (instead of preprocessed) use of unprocessed (or level 1b) TOVS radiance data, which has shown a significant positive impact on the quality of the analysis and a significant improvement in forecast skill (McNally et al. 2006). The U.S. Air Force Weather Agency’s (AFWA) 47-km snow analysis (Kopp and Kiess 1996) is used to replace the model snow depth daily at 0000 UTC. The NARR uses the operational NCEP Eta model as the background model to better resolve orographic features and runs on the Eta grid at horizontal space of 32 km with 45 vertical levels.

c. CFSR

The CFS uses a so-called “semi coupled” nudging approach (Saha et al. 2010; Meng et al. 2012) to correct soil moisture. In CFSR, in addition to the online Noah LSM in the parent model, there is an offline parallel run where the Noah LSM is driven by the observed precipitation, but with near-surface variables from the coupled atmospheric analysis. The simulated states of soil moisture and soil temperature are then used to replace their corresponding prognostic fields in the online Noah LSM without impacting surface fluxes. The update/coupling is done daily at 0000 UTC. To avoid any discrepancies, the offline Noah LSM is configured with the identical settings as the land surface component in the fully coupled CFSR. Two observational sets of global precipitation analyses are used in the CFSR land surface analysis. One is the CMAP dataset used in GR2, and the other is the CPC unified global daily gauge analysis (Xie et al. 2007; Chen et al. 2008). At high latitudes, the model precipitation is used because of uncertainties existing in both products. The datasets are then reconstructed over the global land using a “blending” technique (Meng et al. 2012). The synergic precipitation product is used to drive the offline Noah LSM. An external snow analysis with a combined use of the AFWA product used in NARR and the product from the National Environmental Satellite, Data, and Information Service (NESDIS) Interactive Multisensor Snow and Ice Mapping System (IMS; Helfrich et al. 2007) is used to update the snow depth in such a way that if model-guessed snow depth is greater than twice (or less than half) the analyzed value, then the model snow depth is set to twice (half) the analyzed value; otherwise, the model snow depth is not affected. In contrast to the direct replacement of model snow depth, this treatment tends to let the model have a relatively smooth evolution of the snowpack and reduce the artificial addition of snow water into the system if the land model has a quick snow meltdown problem. The CFSR is run on a T382 Gaussian (~38 km) grid with 64 vertical levels. The high horizontal resolution makes a direct comparison with the NARR possible.

4. Reanalysis and verification data

a. Reanalysis

For the GR2 and NARR, monthly means covering a 31-yr period from 1979 to 2009 (available at http://nomads.ncdc.noaa.gov/) are used in the computation. For the CFSR, both monthly means and daily guess and analysis fields covering the same period are available to this study, which make it possible to compute the water balance residual (artificially added or removed) resulting from the analysis cycle. The grids of the datasets are T62 (~200 km) for the GR2, the operational NCEP Eta/North American Model (NAM) 212 grid (~40 km, interpolated from the native Eta grid) for the NARR, and T382 (~38 km) for the CFSR.

Also included in this study (for comparison purposes) is monthly data (T126; ~100 km) obtained from the traditional offline GLDAS–Noah (Rodell et al. 2004) simulation for the same period. The inclusion mainly stems from the fact that the measurement and data acquisition for many surface variables, such as surface evapotranspiration ET, a shared component in the surface water and energy budget, is difficult and expensive, especially at the global scale. The common modeling is taken as an alternative approach. It is believed that an advanced LSM forced with satellite- and ground-based observations can provide the optimal land states (soil moisture, soil temperature, and snow) and the best estimates of other surface quantities (e.g., latent heat and runoff). Since there is only the land component involved, the GLDAS provides a useful benchmark for evaluation of the reanalysis fluxes and realistic land states, especially soil moisture, to initialize climate models.

To avoid any discrepancies, the GLDAS is constructed and executed using the same terrain field and land mask (on the T126 Gaussian grid), as well as all the same specifications of land surface characteristics (soil class, vegetation class, etc.) and land physical parameters and the same version of the Noah LSM (2.7.1) as employed in the CFSR. The GLDAS simulation is driven by the precipitation (described next in section 4b) and surface meteorological observations from a combination of the datasets from various sources, including the NCEP Global Data Assimilation System (GDAS), the NASA Goddard Earth Observing System Model (GEOS) data assimilation system, the ECMWF reanalysis, and the Princeton global meteorological forcing datasets, among others. While the GLDAS–Noah simulation is subject to the quality of forcing data and model physical parameterizations, Wang et al. (2011) demonstrate that with a combined use of GLDAS–Noah input and output, their water and energy budget–based distributed hydrological model (Wang et al. 2009) performs well in simulating basin-integrated discharge and evapotranspiration over a semiarid basin in northeastern China.

b. Observed precipitation and runoff

The precipitation verification data are taken from the product used to drive the offline Noah LSM in CFSR. One reason for the choice is its global coverage, which makes it possible to extend the present study to other parts of the globe, and another reason is the good quality over the contiguous United States, where there are lots of rain gauge observations available.

The observed runoff is computed from the U.S. Geological Survey (USGS) observed daily discharge at the gauge at Vicksburg, Mississippi (available from http://nwis.waterdata.usgs.gov/nwis), covering a 21-yr period from 1979 to 1999. Although additional observed discharge data are intermittently available on this site after 2000, only these 21 years are chosen to compute the climatology for quality-control purposes.

c. Radiation flux and skin temperature

The radiation verification datasets come from the National Aeronautics and Space Administration (NASA) World Climate Research Programme (WCRP)/GEWEX Surface Radiation Budget (SRB) project. Based on satellite observations, the project developed a 24.5-yr surface shortwave and longwave flux dataset on a regular global grid at 1° resolution, and it is currently available from July 1983 to December 2007 at the NASA Langley web page (http://eosweb.larc.nasa.gov).The latest release 3.0 is used in this study. According to the data document, the quality-controlled aggregate dataset for all monthly averaged sites and years, mean biases for the shortwave and longwave fluxes are −4.2 and −0.1 W m−2, respectively. Given the good quality of the longwave flux, the input skin temperature, a critical variable to surface longwave flux retrieval, is used for comparison with the output from the reanalyses even though it has known limitations (values at the poles).

d. Soil moisture

The soil moisture observation is taken from the Global Soil Moisture Data Bank (Hollinger and Isard 1994; Robock et al. 2000; available at www.ipf.tuwien.ac.at). The soil moisture is collected for the top 10 cm of soil, and then for 20-cm layers down to a depth of 2 m. The dataset used in the present study covers from January 1981 to June 2004. The spatial coordinates are from 37° to 43°N and 269° to 273°W. The data are averages of the 19 stations over the state of Illinois.

5. Surface water and energy budgets

a. Surface water budget

The equation for computing surface water budget in a reanalysis can be written as
e1
The temporal change in surface water d(W + S)/dt is equal to precipitation P minus evaporation E and runoff R, plus residual Res(W + S). The surface water includes total soil water W in various soil layers and snow water S. The evaporation includes evaporation from bare soil, transpiration from vegetation, direct evaporation from canopy surface water, and snow sublimation. The residual term has two sources. The first results from the explicit or implicit insertions of the observations. The insertions include direct use of the observed “pentad” data over a 5-day period and weekly snow update in the GR2, daily snowpack update in the NARR, and daily replacement of modeled soil water and snow update in the CFSR. The second comes from model spinup and postprocessing (such as roundoff and interpolation of a variable from one grid to another). Note that, without adequate frequencies of outputs from the reanalyses (e.g., 5-day and weekly guess and analysis outputs from the GR2 and daily snow guess output from the NARR), it is not possible to separate the two residual sources. In addition, information of surface water from the first and last days of each month is needed to account for actual monthly storage changes. The annual and seasonal residuals computed from GR2 and NARR monthly data come from both sources. However, for the CFSR, both monthly and daily (guess and analysis) data are available to this study, so the residuals reflect only the imbalance introduced from the assimilation technique; presumably the other errors are neglected. Figure 1 shows the mean annual cycle of the terms in Eq. (1) in the surface water budget of the three analyses.
Fig. 1.
Fig. 1.

Seasonal water budget terms averaged over the Mississippi River basin from CFSR, NARR, and GR2 and comparison with observations and GLDAS: (a) P, (b) R, (c) E, (d) surface water tendency [−d(W + S)/dt; squares are surface water increment and triangles are soil moisture increment in CFSR from the analysis cycle], (e) surface water (W + S), and (f) unbalanced surface water [Res(W + S)]. In (a) and (b), the observed seasonal precipitation is based on the offline GLDAS forcing from Jan 1979 to Dec 2009, and the observed runoff climatology is based on the streamflow at Vicksburg, MS, from Oct 1979 to Sep 1999.

Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0056.1

1) Precipitation and soil moisture control

The mean seasonal cycle of precipitation (Fig. 1a) is accurately described (albeit a small low bias) in NARR. It is not surprising because the NARR assimilates observed hourly precipitation (section 3b) via modifying a lower boundary moisture profile that plays a critical role in controlling precipitation production, especially in the summer months when convective clouds become a major source of precipitation.

Compared to the NARR, the CFSR has a high bias in the cold months (October–April) and a low bias in the warm months (June–August). This is associated with systematic biases in the background atmospheric model and the assimilation technique. The atmospheric model in CFSR tends to have a high bias in both surface evaporation and precipitation over nonarid land regions (e.g., the eastern United States; Campana and Caplan 2005; www.nws.noaa.gov/ost/climate/STIP/GFS_news_110305.htm) during months of nonsparse green vegetation. Changes made to correct the bias by increasing canopy resistance were implemented in mid-2005. This change is expected because the key components in surface moisture flux, especially during warm season, are bare soil evaporation and plant transpiration; however, this modification does not help too much in the cold season, as the evaporation from surface is mostly energy limited. The direct insertion of soil moisture from the offline parallel run to correct the model precipitation and consequently soil moisture is not effective in wintertime when the evaporation is at its minimum since the soil could be frozen. Figure 1d indicates that the extra surface water in CFSR is removed from the system via soil moisture (triangles) at the analysis cycle over the cold months. In addition, to overcome the early snowmelt bias found in the Noah LSM (Barlage et al. 2010; Livneh et al. 2010), snow is added into the system (less than half of the observation) through the daily land analysis. As a result, the surface water removal (including both soil water and snow water; squares) reaches its maximum in April.

In the warm season, the efficiency of the soil moisture correction is mainly determined by the atmospheric sensitivity to soil moisture variations. The Mississippi River basin has a number of distinct climate zones; for instance, the central Great Plains, covering most of the Missouri River basin (the largest Mississippi subbasin), has a semiarid climate and is identified as one of the “hot spots” (Koster et al. 2004) where the summertime precipitation is sensitive to soil moisture availability. The canopy resistance in the online Noah LSM was tuned to offset the positive bias in precipitation from the atmospheric model. However, the inserted soil moisture is obtained from the offline parallel GLDAS run driven by observed precipitation. The soil moisture correction is to remove soil moisture, which is the main source of surface evaporation and convective rainfall. As a result, the precipitation in CFSR is the lowest of the three reanalyses from June to August. In contrast, east of the Mississippi River has a dominant humid continental climate (www.ncdc.noaa.gov/cag/), the convective precipitation is mainly controlled by net radiative fluxes, and it is less sensitive to soil moisture variations. An examination of the difference in June–August (JJA) precipitation climatology between the CFSR and observation indicates that the CFSR precipitation has a low bias of around 1.5 mm day−1 (not shown) averaged over the Missouri and the upper Mississippi subbasins, while maintaining a high bias of 2.5 mm day−1 (not shown) over the relatively small Ohio subbasin. These results are in good agreement with the findings by Berbery and Rasmusson (1999), who examined the mesoscale features of the moisture budgets of the Mississippi River basin and its subbasins using 2 years of regional analyses based on the EDAS data (Noah LSM based). They found a dry bias over the central United States and an excess of precipitation over the southeastern United States during the summertime, which are similar to what is seen in the global model. Note that total soil moisture is mainly determined by the wet climates, and even with the precipitation low bias, the large basin is still wet. As shown in Fig. 1d, soil moisture (triangles) is removed from the system at the analysis cycle all season long, which leads to a damped seasonal cycle.

Finally, the summertime high bias in GR2 is a known problem and could be attributed to the convection and planetary boundary parameterizations (Betts et al. 1996).

2) Runoff

The runoff of CFSR, NARR, and GR2 is persistently lower, except during the summertime in GR2, and has a smaller annual cycle compared to the observed streamflow, even though the peak month in both CFSR and NARR agrees reasonably well with the observation. In addition to the assimilation technique problem used in GR2 and CFSR discussed above, there are two physical reasons for the underestimation. The first is the absence of lateral flow representations (both overland surface and saturated subsurface) in the two land models. The Noah and OSU LSMs are single-column models that treat the earth as a uniform surface with no geography and assume spatially continuous soil moisture values. However, elevations within the Mississippi River basin range from sea level at the mouth of the Mississippi to some of the highest peaks in North America. The topography varies from low-lying swampland to undulating hills to craggy mountain peaks. Under these circumstances, the lateral movement of surface and sustained subsurface saturation flow becomes an important source of total runoff, especially in the early spring (March) when the lateral overflows due to snowmelt are not small. To better account for the runoff from these redistribution processes, Gochis and Chen (2003) proposed to add routing schemes of overland and saturated subsurface lateral flows and achieved some success. Lohmann et al. (2004) attached a streamflow routing model to the standard Noah LSM. Unfortunately, however, these enhancements were not implemented in the Noah LSM used in both CFSR and NARR.

The second reason is associated with subsurface hydrology. The two land models treat base flow as a linear function of bottom soil-layer drainage (Schaake et al. 1996). The two land models assume that the water table is shallow enough to be within the model soil layers. However, the water table can vary dramatically depending on a number of factors, including soil property, soil moisture, and precipitation rate. The water table rises to the surface in response to a storm where the lateral redistribution of both overland and subsurface flows contributes more to the total runoff and drops to below the model bottom because of an insufficient amount of precipitation, where upward water flow through capillary action becomes an important process to charge soil moisture in order to prevent the soil from further dryness. The absence of representation of these processes in both land models leads to the soil getting wet and dry quickly, especially in the cold months when the soil moisture is charged following the summertime dryness. As shown in Fig. 1b, the runoff of the three reanalyses continues to drop after September and has only a trivial increase from November to December. To address these shortcomings, Niu et al. (2007) developed a simple groundwater model by representing recharge and discharge processes of the water storage in an unconfined aquifer. The recent Noah multiparameterization (Noah-MP) physics (Niu et al. 2011) now includes soil moisture–groundwater interaction and related runoff production. Barlage et al. (2013) demonstrated simulations using the Noah-MP with free drainage always lose water out of the soil bottom, and the simulations with an interactive aquifer can dramatically improve the simulation results in their 6-month integration with the Weather Research and Forecasting (WRF) Model. Unfortunately, these developments were too late to be tested and incorporated into the Noah LSM when the CFSR project started in 2007.

To examine the behavior of runoff in the Noah LSM, an offline GLDAS–Noah run was conducted, where the Noah LSM is driven by surface observations. Figure 1b shows that the GLDAS–Noah runoff is generally larger than that in both NARR and CFSR, but still exhibits low bias (about 0.3 mm day−1) compared to the observation. Note that, compared to NARR, the offline GLDAS–Noah runoff does show a noticeable increase during the cold months (from November to February), indicating that the Noah LSM can predict the correct seasonal variation and suggesting that some important atmospheric feedbacks may be missing as a result of changes made to near-surface meteorological fields.

The wrong seasonal cycle (peaks in June) in GR2 is attributed to the summertime high bias in precipitation, and runoff in the OSU LSM comes mainly from the top layer (resulting from the infiltration-excess process) and the soil moisture adjustment policy used in GR2 (section 3a).

It is worth mentioning that in most reanalysis water budget studies, the atmospheric moisture convergence −divQ (integrated from surface to the top of model layer) is commonly used as an estimate for total runoff (or vice versa) over a large-scale area on an annual basis (Peixóto and Oort 1983; Trenberth et al. 2011; Seager and Henderson 2013). However, Table 1 indicates that this assumption cannot hold true at this basin scale without accounting for the atmospheric moisture changes in both perceptible water and moisture convergence. This is especially true for the NARR where the initialization assimilation scheme is used. It is also problematic to derive the increment in atmospheric moisture from the surface runoff when a bias exits and some important physical processes are missing (such as the aforementioned lateral water flow and groundwater hydrology).

Table 1.

Annual means and correlations of water balance–related terms from CFSR, NARR, GR2, and observations (Obs). Correlations are calculated from time series that have the climatological mean removed. The annual means and correlations are computed based on data from Jan 1979 to Dec 2009 for precipitation and from Oct 1979 to Sep 1999 for runoff.

Table 1.

It should also be stressed that soil moisture serves as a slowly varying forcing in climate models. Lateral water flow and subsurface hydrological processes affect soil moisture redistribution and associated soil moisture memory, which is essential to climate predictions (Koster et al. 2009; Yang et al. 2011). Of particular importance is the representation of these processes in CFSR as its land states are being used to initialize the NCEP Climate Forecast System, version 2 (CFSv2; Saha et al. 2014), for seasonal predictions. Therefore, an enhanced version of the Noah LSM that can address these issues is desirable for use in the next generation of NCEP reanalysis.

3) Evaporation

Evaporation is a major component of the water balance. Surface evaporation is not only the main sink of precipitation, but also has a direct impact on the efficiency of the data assimilation techniques as corrections are commonly applied to soil moisture, which is the main source of surface evaporation. As such, surface evaporation is a function of both model evaporation parameterizations and the data assimilation techniques.

Surface evaporation is modeled via the same Penman approach, and the vegetation effect is represented using the common big leaf scheme in both land models (see references in section 2). The factors affecting surface evaporation include both atmospheric conditions and surface parameters. Based on the observation data, Jacquemin and Noilhan (1990) identified that the order of importance for surface parameters is soil moisture, vegetation cover, minimum stomatal resistance, leaf area index, and surface roughness. Therefore, for a given atmospheric condition, the differences in evaporation between the two land models mainly come from soil moisture availability (precipitation from the atmospheric model and soil moisture correction) and surface characteristics (vegetation fraction and canopy stomatal control).

Given the great similarity in the evaporation parameterization, it is not surprising to see that the seasonal cycle of evaporation is similar to each other with a peak in summer in the three reanalyses. It is expected because key components in the surface moisture flux are plant transpiration and bare soil evaporation over nonarid regions in the warm season. The only difference in evaporation control between the two land models is that the control in Noah LSM is tight and more realistic. The positive precipitation bias during the warm season, large fractional vegetation cover (fixed at 70% for all vegetation classes), loose canopy evaporation control (plants can draw soil moisture from deep soil because of the fixed root zone depth), and the inefficient 5-day soil moisture nudging all contribute to the highest evaporation of the three reanalyses with minimal impact on soil moisture in GR2. In contrast, the low precipitation bias in CFSR, the realistic seasonal greenness vegetation fraction (GVF; based on the 5-yr climatology derived from the satellite retrieval), vegetation class–dependent root zone depth, and the tuned canopy resistance (to compensate for the high bias in precipitation from the atmospheric model) in the Noah LSM make it difficult for the atmosphere to draw water from soil in CFSR. The NARR evaporation generally lies between GR2 and CFSR in warm months. In the winter months, the difference in evaporation between NARR and CFSR follows closely with that in precipitation because the evaporation is energy limited. Note that compared to NARR, the offline GLDAS evaporation is lower over most of the months, even when the precipitation is close to each other (Fig. 1a). It is likely linked to a higher heat exchange coefficient from the coupled atmospheric model since the coefficient is computed solely by the input atmospheric forcing in offline simulation and there is no atmospheric feedback involved.

4) Surface water and tendency

It is not surprising to see the good agreement in both seasonal surface water (Fig. 1e) and tendency terms (Fig. 1d) between CFSR and NARR, as both reanalyses use the same land model. The success of NARR in assimilating precipitation (Mesinger et al. 2006) with observed surface meteorological variables is equivalent to the direct use of observed precipitation in the offline parallel run with near-surface forcing from the atmospheric analysis in CFSR, except that the precipitation control is tighter in NARR. To check how soil moisture from the reanalyses is close to the observation, Fig. 2 presents a comparison of the seasonal soil moisture climatology and the companioned precipitation with the observations over the state of Illinois. Figure 2 illustrates that the NARR soil moisture and precipitation are in closer agreement with the observations than the other two reanalyses, which are consistent with the large basin averages. The amplified seasonal cycle in both CFSR and NARR is tied to the absence of the subsurface hydrological processes (discussed earlier), which can cause unrealistically dry soil moisture in the cold months because the surface overflow is at its minimum and the geographical relief is small. The simulated soil moisture from GLDAS is higher than that from both CFSR and NARR, but still exhibits a low bias, demonstrating again the importance of the subsurface hydrology and associated base flow because there is no feedback from the atmosphere and the surface water is solely determined by the model physics.

Fig. 2.
Fig. 2.

(a) Seasonal soil moisture averaged over the state of Illinois from the three reanalyses, observations, and GLDAS (mm); and (b) the corresponding seasonal precipitation (mm day−1) from Jan 1981 to Dec 2004.

Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0056.1

In comparison with both CFSR and NARR, the surface water in GR2 is considerably lower and has a small seasonal cycle, even though the two land surface models have the same soil depth of 2 m. This is because the two land models have a different soil moisture climatology, which is determined by several factors, including soil layers, soil state variables, and evaporation controls. The high summertime precipitation is almost counteracted by both higher evaporation and runoff adjustments with little changes in bottom soil moisture content, which contributes to the suppressed seasonal variation (Srinivasan et al. 2000; Kistler et al. 2001).

5) Residual

Residuals in balancing the budgets from the analysis datasets are generally expected because the conservation law cannot be enforced at an analysis cycle. The magnitude of nudging residual is dependent on how often the adjustments are made, how much effort is needed to correct the biases, and how long the model takes to spin up a field. The residual of CFSR and GR2 (Fig. 1f) has a seasonal cycle similar to the surface water change (high in summer and low in winter), therefore the efforts made to correct soil moisture biases mainly occur in the summertime when the incoming solar radiation is high. The daily replacement of soil moisture and soil temperature in CFSR and the 5-day consecutive adjustments to the top-layer soil moisture without modifying their closely related surface fluxes are responsible for the residuals, and the snow water equivalent adjustments are attributed for the winter residuals. Generally, the residuals in CFSR are small. As shown in Table 1, on average, the water balance in CFSR is closed on an annual basis. Unfortunately, the artificial forcing in GR2 cannot be removed from the monthly data. It is not possible to make a direct comparison of the residual term between the two reanalyses. The interpolation from the Eta grid to the AWIP grid may also contribute to the high summertime residual in NARR. The insertion of observed snow depth is also responsible for the residuals in the winter months. Note that the residuals are also impacted by the number of grid cells in the study domain. The smaller residual in GR2 compared to NARR during the summer months may be partially due to its coarse resolution and may be more reflective of sampling errors. More analysis is necessary to confirm the hypothesis, but it is beyond the scope of this study.

Figure 3 shows the variations of the monthly anomalies (with monthly means subtracted) in the surface hydrological cycle. Despite the disparities in the seasonal means, the anomaly fields show greater similarities in amplitude and variation, which provide a great confidence in the value of the reanalyses for climate studies. In particular, the precipitation (Fig. 3a) in both NARR and CFSR has similar interannual variability and is closer to the observation (Table 1). In comparison, the GR2 precipitation appears to have more extreme years and relatively larger interannual variation, which is consistent with its seasonal cycle, and have lower overall correlation with the observation (Table 1).

Fig. 3.
Fig. 3.

As in Fig. 1, but for monthly anomalies of surface water budget terms from Jan 1984 to Dec 1993 (3-month running means for the presentation only) for CFSR, NARR, GR2, and observations [in (a) and (b) only].

Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0056.1

In spite of the low bias of mean runoff in both NARR and CFSR, the anomaly (Fig. 3b) correlates well with the observation (Table 1). Similar to precipitation, the GR2 runoff has a larger variation in the summer months. The evaporation anomaly (Fig. 3c) closely follows the precipitation anomaly in the three analyses and agrees with each other over the most years, for example, the drought of 1988 (April–June 1988 from the Midwest to the Great Plains) and the flood of 1993 (April–October 1993 along the Mississippi and Missouri Rivers and their tributaries). Note that, because of the high evaporation bias existing in both LSMs, the surface water anomaly is less pronounced in the flood of 1993 than the drought of 1988. Despite the lower total surface water amount in GR2, the interannual variations in both surface water tendency (Fig. 3d) and storage (Fig. 3e) are similar to what are shown in both CFSR and NARR. The soil moisture anomalies (Fig. 3e) arising from the 1988 drought and the 1993 flood also agree with each other in the three reanalyses. The water residual in CFSR (Fig. 3f) has the smallest interannual variability. Unfortunately, as mentioned earlier, the artificial forcing in GR2 and the interpolation error in NARR cannot be removed from the calculations because of the absence of guess fields in the monthly averages and output on the native Eta grid, which make it hard to quantify how much they contribute to the larger tendency terms.

Table 1 provides the annual means for the water balance–related terms from the three reanalyses and the correlations of monthly anomaly time series between the three analyses, and with their corresponding available observations (covering the same periods). Table 1 shows the CFSR mean annual precipitation over the large basin goes mainly to evaporation and soil moisture (and removed at the analysis cycle) with the minimum allocation to runoff (8.4%), whereas the NARR precipitation is mainly balanced by surface evaporation and runoff (11.4%) with minimum impact on soil moisture. The partitioning of precipitation to runoff in CFSR and NARR is much lower than the observation (20.7%). Given the success in precipitation assimilation, the direct use of solar radiance, and assimilation of surface winds and temperature in NARR, it is reasonable to believe that the evaporation in NARR has a high bias and that the bias may come from the model’s evaporation parameterization.

The results are similar to the findings by Sheffield et al. (2012), who show that the annual average of runoff over the continental United States for 1979–2003 in both NARR and offline NLDAS (Noah LSM based) is much lower than the observation and the simulation from the Variable Infiltration Capacity model (VIC) (Liang et al. 1994), which has been calibrated and validated against measured streamflow data. They attribute the bias to the low canopy resistance used in the Noah LSM.

Table 1 also shows that the anomaly correlations of all the water budget terms between CFSR and NARR are higher than those between GR2 and NARR and closer to the available observations (P and R).

b. Surface energy budget

The surface energy budget can be written as
e2
The net shortwave heating SW is balanced by the net longwave cooling LW, the sensible heat SH, the latent heat LH, the ground heat G, and the snowmelt heat SNOHF. The unbalanced energy term is Res(e). The reanalyses compute surface skin temperature Ts diagnostically from the surface energy balance.

Figure 4 shows that there is almost a complete seasonal energy balance, where solar radiation provides large positive input. This solar flux is balanced by LW, SH, LH, G, and SNOFH (Table 2). All act to cool the surface during the summer. Sensible heat also helps to warm the surface in GR2. Compared to the satellite observations, to a different degree, the net shortwave heating (Fig. 4a) and the closely followed net longwave cooling (Fig. 4b) are overestimated in all the three reanalyses. The SW in GR2, which is the highest of the three reanalyses, is dissipated mainly through high LH (Fig. 4d) and low SH (Fig. 4c) because the GR2 has a positive precipitation bias and the OSU LSM exerts weaker controls on the surface evaporation, which is the most efficient way to consume energy without raising the surface temperature. As shown in Fig. 4b, the GR2 net longwave cooling is relatively stable from April to October, which leads to a damped seasonal cycle in LW.

Fig. 4.
Fig. 4.

Seasonal surface energy budget terms averaged over the Mississippi River basin from CFSR, NARR, and GR2 and comparison with observations [in (a),(b), and (e) only]: (a) SW, (b) LW, (c) −SH, (d) −LH, (e) Ts, and (f) residual (the observed SW, LW, and Ts are from Jul 1983 to Dec 2007). The units are W m−2 for SW, LW, SH, LH, and residual, and kelvins for Ts.

Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0056.1

Table 2.

Annual means and correlations of energy balance–related terms from CFSR, NARR, GR2, and Obs. Rnet is total radiation forcing. Correlations are calculated from time series that have the climatological mean removed. The annual mean and correlation are computed based on data from Jul 1983 to Dec 2007 for SW, LW, Rnet, and Ts.

Table 2.

In comparison, the slightly higher SW in both NARR and CFSR (Fig. 4a) is dissipated through the higher LH, SH, and LW (Figs. 4b–d) by raising the Ts (Fig. 4e) because the Noah LSM imposes stronger controls on the surface evaporation. The lower P during the warm months in CFSR also contributes to the lower LH. In spite of the differences in the energy terms, Ts in the three reanalyses are close to each other and to the observation, except a small high bias in the NARR in the warm season, which is also consistent with its higher SW during that period.

The variations in surface energy residuals in the early months (Fig. 4f) are similar to what is seen in surface water content (Fig. 1e) in the three reanalyses, indicating snow assimilation and its associated snowmelt processes are the main contributors. Particularly in the early spring, the updates to the snow depth in both CFSR and GR2 and the insertion of the observed snow depth in NARR without modifying the snowmelt and snow sublimation–related heat transfer are believed to make the largest contributions. The relatively small values during the cold months in both CFSR and GR2 are likely linked to the looser controls in the data assimilation techniques since the soil could be frozen. Certainly, all the causes for the water budget residual affect the energy balance closure as well. Note that the replacement of soil temperature (see section 3c) may also contribute to the energy residual in CFSR. As indicated in Table 2, the CFSR has the largest nonzero annual ground heat flux with the smallest annual residual heating of 0.28 W m−2. Considering other possible errors, for example, roundoff [the dataset is in WMO Gridded Binary (GRIB) format], it is assumed that the energy balance is closed in CFSR on an annual basis.

Figure 5 shows the variations of the anomalies in the energy terms, which again have remarkable consistency among the three reanalyses. The anomalies in the energy budget have clear phase relationships; SW (Fig. 5a) is in opposite phase of LW (Fig. 5b), and so is SH (Fig. 5c) of LH (Fig. 5d). They are all consistent with the corresponding climatology. The interannual variations in both SW (Fig. 5a) and LW (Fig. 5b) compare well with the observations. Thus, it is not surprising to see the good agreements in Ts (Fig. 5e), as shown in Table 2; the correlations between the reanalyses and the observation are high. The variation in residual heating is slightly larger in GR2 (Fig. 5f). Table 2 shows that the CFSR energy anomaly fields have high correlations with their corresponding terms in the NARR.

Fig. 5.
Fig. 5.

As in Fig. 4, but for monthly anomalies of surface energy budget terms from Jan 1984 to Dec 1993 (3-month running means for presentation only).

Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0056.1

6. Summary and discussion

In this study, the three reanalyses developed at the NCEP are compared with each other for the period 1979–2009 and with available observations for the Mississippi River basin. There are a number of noticeable differences and similarities. The summertime precipitation and runoff in GR2 are too high compared to the observations, whereas the runoff of both CFSR and NARR is small and the corresponding precipitation is closer to the observation. The discrepancies are caused by the biases from the background forcing, the physical parameterizations used in the two land models, and the approaches in which the observations are ingested in the data assimilation systems. The disagreements between GR2 and the other two reanalyses are mainly caused by the OSU LSM physics and its soil-layer configuration, in particular, the surface evaporation control. In addition, the nudging scheme is not efficient because of the lack of surface observations. These problems have been corrected by using the advanced Noah LSM and alternative assimilation techniques. The success of NARR in assimilating the high-quality observed precipitation through modifying the lower-level moisture profile is a good example. The direct insertion of the observed snow depth may also contribute to a slightly better water climate. Constrained by the data quality and availability, the CFSR assimilates precipitation via changing soil moisture obtained from the semicoupled land data analysis, as it imposes less restriction on the daily precipitation. The efficiency of the technique relies on the feedback between soil moisture and precipitation. This approach is vulnerable to the background model biases, as is the use of confinement of model snow depth within the specified ranges.

In spite of the differences in seasonal water budgets among the three reanalyses, the seasonal energy terms are comparable and highly correlated with each other. Thus, the interannual variability is in both water and energy terms. Unfortunately, the reanalyses, more or less, carry errors, even in CFSR, where the water and energy budgets can only be balanced on an annual basis. A better representation of land surface physics, in particular, runoff parameterization and an improved assimilation methodology that could better close the budgets, are needed in future reanalyses. Nevertheless, the comparison shows that the CFSR achieves a large improvement over the previous GR2, and both water and energy terms compare well with their counterparts in the NARR; thus, the CFSR dataset can be used in global climate studies.

Acknowledgments

This study was supported by the NCEP Core Project component of the NOAA Climate Program Office/Climate Prediction Program for the Americas (CPPA) and by the Modeling, Analysis, Predictions and Projections Program (MAPP) under Grant GC11-589. Thoughtful reviews and constructive suggestions by the editor and reviewers are very much appreciated.

REFERENCES

  • Auad, G., , Miller A. J. , , Roads J. O. , , and Cayan D. , 2001: Pacific Ocean wind stress and surface heat flux anomalies from NCEP reanalysis and observation: Cross-statistics and ocean model responses. J. Geophys. Res., 106, 22 24922 265, doi:10.1029/2000JC000264.

    • Search Google Scholar
    • Export Citation
  • Baldwin, M. E., , and Mitchell K. E. , 1997: The NCEP hourly multi-sensor U.S. precipitation analysis for operations and GCIP research. Preprints, 13th Conf. on Hydrology, Long Beach, CA, Amer. Meteor. Soc., 5455.

  • Barlage, M., and et al. , 2010: Noah land surface model modifications to improve snowpack prediction in the Colorado Rocky Mountains. J. Geophys. Res., 115, D22101, doi:10.1029/2009JD013470.

    • Search Google Scholar
    • Export Citation
  • Barlage, M., , Tewari M. , , Chen F. , , Manning K. , , and Miguez-Macho G. , 2013: North American regional climate simulations with WRF/Noah-MP validation and the effect of ground water interaction. Proc. 14th WRF User's Workshop, Boulder, CO, WRF. [Available online at www2.mmm.ucar.edu/wrf/users/workshops/WS2013/ppts/5B.2.pdf.]

    • Search Google Scholar
    • Export Citation
  • Berbery, E. H., , and Rasmusson E. M. , 1999: Mississippi moisture budgets on regional scales. Mon. Wea. Rev.,127, 2654–2673, doi:10.1175/1520-0493(1999)127<2654:MMBORS>2.0.CO;2.

  • Berbery, E. H., , Luo Y. , , Mitchell K. , , and Betts A. K. , 2003: Eta model–estimated land surface processes and the hydrologic cycle of the Mississippi basin. J. Geophys. Res., 108, 8852, doi:10.1029/2002JD003192.

    • Search Google Scholar
    • Export Citation
  • Betts, A., , Hong S.-Y. , , and Pan H.-L. , 1996: Comparison of NCEP–NCAR reanalysis with 1987 FIFE data. Mon. Wea. Rev., 124, 14801498, doi:10.1175/1520-0493(1996)124<1480:CONNRW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Betts, A., , Viterbo P. , , and Wood E. , 1998: Surface energy and water balance for the Arkansas–Red River basin from the ECMWF reanalysis. J. Climate, 11, 28812897, doi:10.1175/1520-0442(1998)011<2881:SEAWBF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Betts, A., , Ball J. H. , , and Viterbo P. , 1999: Basin-scale surface water and energy budgets for the Mississippi from the ECMWF reanalysis. J. Geophys. Res., 104, 19 29319 306, doi:10.1029/1999JD900056.

    • Search Google Scholar
    • Export Citation
  • Black, T., 1994: The new NMC mesoscale Eta model: Description and forecast examples. Wea. Forecasting, 9, 265278, doi:10.1175/1520-0434(1994)009<0265:TNNMEM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Brunke, M. A., , Wang Z. , , and Zeng X. , 2011: An assessment of the uncertainties in ocean surface turbulent fluxes in 11 reanalysis, satellite-derived, and combined global datasets. J. Climate,24, 5469–5493, doi:10.1175/2011JCLI4223.1.

  • Campana, K., , and Caplan P. , 2005: Technical procedures bulletin for the T382 Global Forecast System. NOAA/NCEP/Environmental Modeling Center. [Available online at www.emc.ncep.noaa.gov/gc_wmb/Documentation/TPBoct05/T382.TPB.FINAL.htm.]

  • Chen, F., , Mitchell K. , , Schaake J. , , Xue Y. , , Pan H.-L. , , Koren V. , , Duan Q. , , and Betts A. , 1996: Modeling of land surface evaporation by four schemes and comparison with FIFE observations. J. Geophys. Res., 101, 7251–7268, doi:10.1029/95JD02165.

    • Search Google Scholar
    • Export Citation
  • Chen, M., , Shi W. , , Xie P. , , Silva V. B. S. , , Kousky V. E. , , Wayne Higgins R. , , and Janowiak J. E. , 2008: Assessing objective techniques for gauge-based analyses of global daily precipitation. J. Geophys. Res., 113, D04110, doi:10.1029/2007JD009132.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and et al. , 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597, doi:10.1002/qj.828.

    • Search Google Scholar
    • Export Citation
  • Dole, R., , Hoerling M. , , and Schubert S. , Eds., 2008: Reanalysis of historical climate data for key atmospheric features: Implications for attribution of causes of observed change. U.S. Climate Change Science Program Synthesis and Assessment Product 1.3, NOAA, Asheville, NC, 156 pp.

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

    • Search Google Scholar
    • Export Citation
  • Gochis, D. J., , and Chen F. , 2003: Implementation and testing of a grid-based routing scheme in the Noah land surface model. EGS–AGU–EUG Joint Assembly, Nice, France, EGU, Abstract 799.

  • Helfrich, S. R., , McNamara D. , , Ramsay B. H. , , Baldwin T. , , and Kasheta T. , 2007: Enhancements to and forthcoming developments in the Interactive Multisensor Snow and Ice Mapping System (IMS). Hydrol. Process., 21, 1576–1586, doi:10.1002/hyp.6720.

  • Hollinger, S. E., , and Isard S. A. , 1994: A soil moisture climatology of Illinois. J. Climate, 7, 822827, 833, doi:10.1175/1520-0442(1994)007<0822:ASMCOI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Holtslag, A. A. M., , and Ek M. , 1996: Simulation of surface fluxes and boundary layer development over the pine forest in HAPEX-MOBILHY. J. Appl. Meteor., 35, 202213, doi:10.1175/1520-0450(1996)035<0202:SOSFAB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Jacquemin, B., , and Noilhan J. , 1990: Sensitivity study and validation of a land surface parameterization using the HAPEX-MOBILHY dataset. Bound.-Layer Meteor., 52, 93134, doi:10.1007/BF00123180.

    • Search Google Scholar
    • Export Citation
  • Jarvis, P. G., 1976: The interpretation of the variations in leaf water potential and stomatal conductance found in canopies in the field. Philos. Trans. Roy. Soc. London.,B273, 593–610, doi:10.1098/rstb.1976.0035.

    • Search Google Scholar
    • Export Citation
  • Ji, M., , Leetmaa A. , , and Derber J. , 1995: An ocean analysis system for seasonal to interannual climate studies. Mon. Wea. Rev., 123, 460481, doi:10.1175/1520-0493(1995)123<0460:AOASFS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Josey, S. A., 2001: A Comparison of ECMWF, NCEP–NCAR, and SOC surface heat fluxes with moored buoy measurements in the subduction region of the northeast Atlantic. J. Climate, 14, 17801789, doi:10.1175/1520-0442(2001)014<1780:ACOENN>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., 2000: The US reanalysis program: Climatology for the new millennium. Rep. of the First Planning Workshop for the Next US Global Reanalysis, University of Maryland, College Park, College Park, MD, 19 pp. [Available online at http://old.usclivar.org/Pubs/ReanalysisWorkshop_Rep.pdf.]

  • Kalnay, E., and et al. , 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437470, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kanamitsu, M., and et al. , 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 16311643, doi:10.1175/BAMS-83-11-1631.

    • Search Google Scholar
    • Export Citation
  • Kistler, R., and et al. , 2001: The NCEP–NCAR 50-Year Reanalysis: Monthly means CD-ROM and documentation. Bull. Amer. Meteor. Soc., 82, 247267, doi:10.1175/1520-0477(2001)082<0247:TNNYRM>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kopp, T. J., , and Kiess R. B. , 1996: The Air Force Global Weather Central snow analysis model. Preprints, 15th Conf. on Weather Analysis and Forecasting, Norfolk, Virginia, Amer. Meteor. Soc., 220222.

  • Koren, V., , Schaake J. , , Mitchell K. , , Duan Q.-Y. , , Chen F. , , and Baker J. M. , 1999: A parameterization of snowpack and frozen ground intended for NCEP weather and climate models. J. Geophys. Res., 104, 19 56919 585, doi:10.1029/1999JD900232.

    • Search Google Scholar
    • Export Citation
  • Koster, R., and et al. , 2004: Regions of strong coupling between soil moisture and precipitation. Science,305, 1138–1140, doi:10.1126/science.1100217.

  • Koster, R., , Guo Z. , , Yang R. , , Dirmeyer P. A. , , Mitchell K. E. , , and Puma M. J. , 2009: On the nature of soil moisture in land surface models. J. Climate, 22, 43224335, doi:10.1175/2009JCLI2832.1.

    • Search Google Scholar
    • Export Citation
  • Liang, X., , Lettenmaier D. P. , , Wood E. F. , , and Burges S. J. , 1994: A simple hydrologically based model of land surface water and energy fluxes for GSMs. J. Geophys. Res., 99, 14 41514 428, doi:10.1029/94JD00483.

    • Search Google Scholar
    • Export Citation
  • Lin, Y., , Mitchell K. E. , , Rogers E. , , and DiMego G. J. , 2005: Using hourly and daily precipitation analyses to improve model water budget. Preprints, Ninth Symp. on Integrated Observing and Assimilation Systems for the Atmosphere, Oceans, and Land Surface, San Diego, CA, Amer. Meteor. Soc., 3.3. [Available online at http://ams.confex.com/ams/pdfpapers/84484.pdf.]

  • Livneh, B., , Xia Y. , , Mitchell K. E. , , Ek M. B. , , and Lettenmaier D. P. , 2010: Noah LSM snow model diagnostics and enhancements. J. Hydrometeor., 11, 721738, doi:10.1175/2009JHM1174.1.

    • Search Google Scholar
    • Export Citation
  • Lohmann, D., and et al. , 2004: Streamflow and water balance intercomparison of four land surface models in the North American Land Data Assimilation Systems project. J. Geophys. Res.,109, D07S91, doi:10.1029/2003JD003517.

  • Luo, Y., , Berbery H. , , Mitchell K. , , and Betts A. , 2007: Relationships between land surface and near-surface atmospheric variables in the NCEP North American Regional Reanalysis. J. Hydrometeor., 8, 11841203, doi:10.1175/2007JHM844.1.

    • Search Google Scholar
    • Export Citation
  • Mahrt, L., , and Ek M. , 1984: The influence of atmospheric stability on potential evaporation. J. Climate Appl. Meteor., 23, 222234, doi:10.1175/1520-0450(1984)023<0222:TIOASO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Mahrt, L., , and Pan H.-L. , 1984: A two-layer model of soil hydrology. Bound.-Layer Meteor., 29, 120, doi:10.1007/BF00119116.

  • Matson, M., , Ropelewski C. F. , , and Varnadore M. S. , 1986: An Atlas of Satellite-Derived Northern Hemisphere Snow Cover Frequency. National Weather Service, 75 pp.

  • Maurer, E. P., , O’Donnell G. , , Lettenmaier D. , , and Roads J. , 2001: Evaluation of the land surface water budget in NCEP/NCAR and NCEP/DOE reanalyses using an off-line hydrologic model. J. Geophys. Res., 106, 17 84117 862, doi:10.1029/2000JD900828.

    • Search Google Scholar
    • Export Citation
  • McNally, A. P., , Derber J. C. , , Wu W.-S. , , and Katz B. B. , 2006: The use of TOVS level-1 radiances in the NCEP SSI analysis system. Quart. J. Roy. Metor. Soc, 126, 689724, doi:10.1002/qj.49712656315.

    • Search Google Scholar
    • Export Citation
  • Meng, J., , Yang R. , , Wei H. , , Ek M. , , Gayno G. , , Xie P. , , and Mitchell K. , 2012: The land surface analysis in the NCEP Climate Forecast System Reanalysis. J. Hydrometeor.,13, 1621–1630, doi:10.1175/JHM-D-11-090.1.

  • Mesinger, F., , Janjić Z. I. , , Nickovic S. , , Gavrilov D. , , and Deaven D. G. , 1988: The step-mountain coordinate: Model description and performance for cases of alpine lee cyclogenesis and for a case of an Appalachian redevelopment. Mon. Wea. Rev., 116, 1493–1518, doi:10.1175/1520-0493(1988)116<1493:TSMCMD>2.0.CO;2.

  • Mesinger, F., and et al. , 2006: North American Regional Reanalysis. Bull. Amer. Meteor. Soc., 87, 343360, doi:10.1175/BAMS-87-3-343.

    • Search Google Scholar
    • Export Citation
  • Milly, P. C. D., , and Dunne K. A. , 2001: Trends in evaporation and surface cooling is the Mississippi River basin. Geophys. Res. Lett.,28, 1219–1222, doi:10.1029/2000GL012321.

  • Mitchell, K., and et al. , 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.

  • Music, B., , and Caya D. , 2007: Evaluation of the hydrological cycle over the Mississippi River basin as simulated by the Canadian Regional Climate Model (CRCM). J. Hydrometeor., 8, 969988, doi:10.1175/JHM627.1.

    • Search Google Scholar
    • Export Citation
  • Niu, G.-Y., , Yang Z.-L. , , Dickinson R. E. , , Gulden L. E. , , and Su H. , 2007: Development of a simple groundwater model for use in climate models and evaluation with Gravity Recovery and Climate Experiment data. J. Geophys. Res., 112, D07103, doi:10.1029/2006JD007522.

    • Search Google Scholar
    • Export Citation
  • Niu, G.-Y., and et al. , 2011: The community Noah land surface model with multi-parameterization options (Noah-MP): 1. Model description and evaluation with local-scale measurements. J. Geophys. Res., 116, D12109, doi:10.1029/2010JD015139.

    • Search Google Scholar
    • Export Citation
  • Noilhan, J., , and Planton S. , 1989: A simple parameterization of land surface processes for meteorological models. Mon. Wea. Rev., 117, 536549, doi:10.1175/1520-0493(1989)117<0536:ASPOLS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Onogi, K., and et al. , 2007: The JRA-25 Reanalysis. J. Meteor. Soc. Japan, 85, 369432, doi:10.2151/jmsj.85.369.

  • Pan, H.-L., , and Mahrt L. , 1987: Interaction between soil hydrology and boundary-layer development. Bound.-Layer Meteor., 38, 185202, doi:10.1007/BF00121563.

    • Search Google Scholar
    • Export Citation
  • Peixóto, J. P., , and Oort A. H. , 1983: The atmospheric branch of the hydrological cycle and climate. Variations of the Global Water Budget, D. Reidel, 5–65, doi:10.1007/978-94-009-6954-4_2.

  • Rienecker, M., and et al. , 2011: MERRA: NASA’s Modern-Era Retrospective Analysis for Research and Applications. J. Climate, 24, 36243648, doi:10.1175/JCLI-D-11-00015.1.

    • Search Google Scholar
    • Export Citation
  • Roads, J., , and Betts A. , 2000: NCEP–NCAR and ECMWF reanalysis surface water and energy budgets for the Mississippi River basin. J. Hydrometeor., 1, 8894, doi:10.1175/1525-7541(2000)001<0088:NNAERS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Roads, J., , Chen S. , , Kanamitsu M. , , and Juang H. , 1999: Surface water characteristics in the NCEP global spectral model and reanalysis. J. Geophys. Res., 104, 19 30719 327, doi:10.1029/98JD01166.

    • Search Google Scholar
    • Export Citation
  • Roads, J., , Kanamitsu M. , , and Stewart R. , 2002: CSE water and energy budgets in the NCEP–DOE Reanalysis II. J. Hydrometeor., 3, 227248, doi:10.1175/1525-7541(2002)003<0227:CWAEBI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Roads, J., and et al. , 2003: GCIP water and energy synthesis (WEBS). J. Geophys. Res.,108, 8609, doi:10.1029/2002JD002583.

  • Robock, A., and et al. , 2000: The Global Soil Moisture Data Bank. Bull. Amer. Meteor. Soc.,81, 1281–1299, doi:10.1175/1520-0477(2000)081<1281:TGSMDB>2.3.CO;2.

  • Rodell, M., and et al. , 2004: The Global Land Data Assimilation System. Bull. Amer. Meteor. Soc., 85, 381394, doi:10.1175/BAMS-85-3-381.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and et al. , 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc.,91, 1015–1057, doi:10.1175/2010BAMS3001.1.

  • Saha, S., and et al. , 2014: The NCEP Climate Forecast System version 2. J. Climate, 27, 21852208, doi:10.1175/JCLI-D-12-00823.1.

  • Schaake, J. C., , Koren V. I. , , Duan Q.-Y. , , Mitchell K. , , and Chen F. , 1996: Simple water balance model for estimating runoff at different spatial and temporal scales. J. Geophys. Res., 101, 74617475, doi:10.1029/95JD02892.

    • Search Google Scholar
    • Export Citation
  • Seager, R., , and Henderson N. , 2013: Diagnostic computation of moisture budgets in the ERA-Interim Reanalysis with reference to analysis of CMIP-archived atmospheric model data. J. Climate, 26, 78767901, doi:10.1175/JCLI-D-13-00018.1.

    • Search Google Scholar
    • Export Citation
  • Sheffield, J., , Livneh B. , , and Wood E. F. , 2012: Representation of terrestrial hydrology and large-scale drought of the continental United States from the North American Regional Reanalysis. J. Hydrometeor., 13, 856876, doi:10.1175/JHM-D-11-065.1.

    • Search Google Scholar
    • Export Citation
  • Srinivasan, G., and et al. , 2000: Soil moisture simulations in revised AMIP models. J. Geophys. Res., 105, 26 63526 644, doi:10.1029/2000JD900443.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , Fasullo J. T. , , and Mackaro J. , 2011: Atmospheric moisture transports from ocean to land and global energy flows in reanalyses. J. Climate,24, 4907–4924, doi:10.1175/2011JCLI4171.1.

  • Uppala, S., and et al. , 2005: The ERA-40 re-analysis. Quart. J. Roy. Meteor., 131, 29613012, doi:10.1256/qj.04.176.

  • Wang, F., , Wang L. , , Koike T. , , Zhou H. , , Yang K. , , Wang A. , , and Li W. , 2011: Evaluation and application of a fine-resolution global data set in a semiarid mesoscale river basin with a distributed biosphere hydrological model. J. Geophys. Res., 116, D21108, doi:10.1029/2011JD015990.

    • Search Google Scholar
    • Export Citation
  • Wang, L., and et al. , 2009: Development of a distributed biosphere hydrological model and its evaluation with the Southern Great Plains Experiments (SGP97 and SGP99). J. Geophys. Res., 114, D08107, doi:10.1029/2008JD010800.

    • Search Google Scholar
    • Export Citation
  • Xie, P., , and Arkin P. A. , 1997: Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor. Soc., 78, 2539–2558, doi:10.1175/1520-0477(1997)078<2539:GPAYMA>2.0.CO;2.

  • Xie, P., , Chen M. , , Yatagai A. , , Hayasaka T. , , Fukushima Y. , , and Yang S. , 2007: A gauge based analysis of daily precipitation over East Asia. J. Hydrometeor., 8, 607626, doi:10.1175/JHM583.1.

    • Search Google Scholar
    • Export Citation
  • Yang, R., , Mitchell K. E. , , Meng J. , , and Ek M. , 2011: Summer-season forecast experiments with the NCEP Climate Forecast System using different land models and different initial land states. J. Climate,24, 2319–2334, doi:10.1175/2010JCLI3797.1.

Save