• Atlas, R., , Wolfson N. , , and Terry J. , 1993: The effect of SST and soil moisture anomalies on GLA model simulations of the 1988 U.S. summer drought. J. Climate, 6, 20342048, doi:10.1175/1520-0442(1993)006<2034:TEOSAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bosilovich, M., , and Sun W. Y. , 1999: Numerical simulation of the 1993 Midwestern flood: Land–atmosphere interactions. J. Climate, 12, 14901505, doi:10.1175/1520-0442(1999)012<1490:NSOTMF>2.0.CO;2.

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

    • Search Google Scholar
    • Export Citation
  • Delworth, T., , and Manabe S. , 1989: The influence of soil wetness on near-surface atmospheric variability. J. Climate, 2, 14471462, doi:10.1175/1520-0442(1989)002<1447:TIOSWO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dirmeyer, P. A., 2011: The terrestrial segment of soil moisture-climate coupling. Geophys. Res. Lett., 38, L16702, doi:10.1029/2011GL048268.

    • Search Google Scholar
    • Export Citation
  • Dirmeyer, P. A., , Koster R. D. , , and Guo Z. , 2006: Do global models properly represent the feedback between land and atmosphere? J. Hydrometeor., 7, 11771198, doi:10.1175/JHM532.1.

    • Search Google Scholar
    • Export Citation
  • D’Odorico, P., , and Porporato A. , 2004: Preferential states in soil moisture and climate dynamics. Proc. Natl. Acad. Sci. USA, 101, 88488851, doi:10.1073/pnas.0401428101.

    • Search Google Scholar
    • Export Citation
  • Douville, H., , Chauvin F. , , and Broqua H. , 2001: Influence of soil moisture on the Asian and African monsoons. Part I: Mean monsoon and daily precipitation. J. Climate, 14, 23812403, doi:10.1175/1520-0442(2001)014<2381:IOSMOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ek, M. B., , and Holtslag A. A. M. , 2004: Influence of soil moisture on boundary layer cloud development. J. Hydrometeor., 5, 8699, doi:10.1175/1525-7541(2004)005<0086:IOSMOB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Elguindi, N., and Coauthors, 2011: Regional climatic model RegCM user manual version 4.1. The Abdus Salam International Centre for Theoretical Physics, Trieste, Italy, 32 pp. [Available online at http://gforge.ictp.it/gf/download/docmanfileversion/21/544/ReferenceMan4.1.pdf.]

  • Findell, K. L., , and Eltahir E. A. B. , 1997: An analysis of the soil moisture-precipitation feedback, based on direct observations from Illinois. Water Resour. Res., 33, 725735, doi:10.1029/96WR03756.

    • Search Google Scholar
    • Export Citation
  • Findell, K. L., , and Eltahir E. A. B. , 2003: Atmospheric controls on soil moisture–boundary layer interactions. Part II: Feedbacks within the continental United States. J. Hydrometeor., 4, 570583, doi:10.1175/1525-7541(2003)004<0570:ACOSML>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Findell, K. L., , Gentine P. , , Linter B. R. , , and Kerr C. , 2011: Probability of afternoon precipitation in eastern United States and Mexico enhanced by high evaporation. Nat. Geosci., 4, 434–439, doi:10.1038/ngeo1174.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., , Mearns L. O. , , Shields C. , , and Mayer L. , 1996: A regional model study of the importance of local versus remote controls of the 1988 drought and the 1993 flood over the central United States. J. Climate, 9, 11501162, doi:10.1175/1520-0442(1996)009<1150:ARMSOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., and Coauthors, 2012: RegCM4: Model description and illustrative basic performance over selected CORDEX domains. Climate Res., 52, 729, doi:10.3354/cr01018.

    • Search Google Scholar
    • Export Citation
  • Gu, H. H., , Wang G. L. , , Yu Z. B. , , and Mei R. , 2012: Assessment on impact of climate change on East and South Asia using the RegCM4 regional climate model. Climatic Change, 114, 301317, doi:10.1007/s10584-012-0411-y.

    • Search Google Scholar
    • Export Citation
  • Guo, Z., and Coauthors, 2006: GLACE: The Global Land–Atmosphere Coupling Experiment. Part II: Analysis. J. Hydrometeor., 7, 611625, doi:10.1175/JHM511.1.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., , and Pan H.-L. , 2000: Impact of soil moisture anomalies on seasonal, summertime circulation over North America in a regional climate model. J. Geophys. Res., 105, 29 62529 634, doi:10.1029/2000JD900276.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kanamitsu, M., , Ebisuzaki W. , , Woollen J. , , Yang S.-K. , , Hnilo J. J. , , Fiorino M. , , and Potter G. L. , 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
  • Kim, J. E., , and Hong S. Y. , 2007: Impact of soil moisture anomalies on summer rainfall over East Asia: A regional climate model study. J. Climate, 20, 57325743, doi:10.1175/2006JCLI1358.1.

    • Search Google Scholar
    • Export Citation
  • Kim, Y. J., , and Wang G. L. , 2007: Impact of initial soil moisture anomalies on subsequent precipitation over North America in the Coupled Land–Atmosphere Model CAM3–CLM3. J. Hydrometeor., 8, 513533, doi:10.1175/JHM611.1.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., , and Suarez M. J. , 1995: Relative contributions of land and ocean processes to precipitation variability. J. Geophys. Res.,100 (D7), 13 775–13 790, doi:10.1029/95JD00176.

  • Koster, R. D., , Dirmeyer P. A. , , Hahmann A. N. , , Ijpelaar R. , , Tyahla L. , , Cox P. , , and Suarez M. J. , 2002: Comparing the degree of land–atmosphere interaction in four atmospheric general circulation models. J. Hydrometeor., 3, 363375, doi:10.1175/1525-7541(2002)003<0363:CTDOLA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Coauthors, 2004a: Realistic initialization of land surface states: Impacts on subseasonal forecast skill. J. Hydrometeor., 5, 10491063, doi:10.1175/JHM-387.1.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Coauthors, 2004b: Regions of strong coupling between soil moisture and precipitation. Science, 305, 11381140, doi:10.1126/science.1100217.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Coauthors, 2006: GLACE: The Global Land–Atmosphere Coupling Experiment. Part I: Overview. J. Hydrometeor., 7, 590610, doi:10.1175/JHM510.1.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., , Schubert S. D. , , and Suarez M. J. , 2009: Analyzing the concurrence of meteorological droughts and warm periods, with implications for the determination of evaporative regime. J. Climate, 22, 33313341, doi:10.1175/2008JCLI2718.1.

    • Search Google Scholar
    • Export Citation
  • Kumar, S. V., and Coauthors, 2006: Land information system: An interoperable framework for high resolution land surface modeling. Environ. Modell. Software, 21, 14021415, doi:10.1016/j.envsoft.2005.07.004.

    • Search Google Scholar
    • Export Citation
  • Leung, L. R., , Huang M. , , Qian Y. , , and Liang X. , 2011: Climate-soil-vegetation control on groundwater table dynamics and its feedbacks in a climate model. Climate Dyn., 36, 5781, doi:10.1007/s00382-010-0746-x.

    • Search Google Scholar
    • Export Citation
  • Liu, Z. Y., , Notaro M. , , and Kutzbach J. , 2006: Assessing global vegetation–climate feedbacks from observations. J. Climate, 19, 787814, doi:10.1175/JCLI3658.1.

    • Search Google Scholar
    • Export Citation
  • Mei, R., , and Wang G. , 2011: Impact of sea surface temperature and soil moisture on summer precipitation in the United States based on observational data. J. Hydrometeor., 12, 10861099, doi:10.1175/2011JHM1312.1.

    • Search Google Scholar
    • Export Citation
  • Mei, R., , and Wang G. , 2012: Summer land–atmosphere coupling strength in the United States: Comparison among observations, reanalysis data and numerical models. J. Hydrometeor., 13, 10101022, doi:10.1175/JHM-D-11-075.1.

    • Search Google Scholar
    • Export Citation
  • Namias, J., 1952: The annual course of month-to-month persistence in climatic anomalies. Bull. Amer. Meteor. Soc., 33, 279285.

  • Namias, J., 1959: Persistence of mid-tropospheric circulations between adjacent months and seasons. The Atmosphere and the Sea in Motion: Scientific Contributions to the Rossby Memorial Volume, Rockefeller Institute Press, 240–248.

  • Rayner, N. A., , Parker D. E. , , Horton E. B. , , Folland C. K. , , Alexander L. V. , , Rowell D. P. , , Kent E. C. , , and Kaplan A. , cited 2006: Met Office’s Global Ice Coverage and Sea Surface Temperatures (GISST), and Monthly Night Marine Air Temperature/SST Anomalies (MOHMATN4)/(MOHSST6) data (1856–2006). NCAS British Atmospheric Data Centre. [Available online at http://badc.nerc.ac.uk/view/badc.nerc.ac.uk__ATOM__dataent_GISST.]

  • Rodell, M., and Coauthors, 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
  • Ruiz-Barradas, A., , and Nigam S. , 2005: Warm season precipitation variability over the U.S. Great Plains in observations, NCEP and ERA-40 reanalyses, and NCAR and NASA atmospheric model simulations. J. Climate, 18, 18081830, doi:10.1175/JCLI3343.1.

    • Search Google Scholar
    • Export Citation
  • Ruiz-Barradas, A., , and Nigam S. , 2006: Great plains hydroclimate variability: The view from North American Regional Reanalysis. J. Climate, 19, 30043010, doi:10.1175/JCLI3768.1.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 10151057, doi:10.1175/2010BAMS3001.1.

    • Search Google Scholar
    • Export Citation
  • Salvucci, G. D., , Saleem J. A. , , and Kaufmann R. , 2002: Investigating soil moisture feedbacks on precipitation with tests of Granger causality. Adv. Water Resour., 25, 13051312, doi:10.1016/S0309-1708(02)00057-X.

    • Search Google Scholar
    • Export Citation
  • Schär, C., , Lüthi D. , , Beyerle U. , , and Heise E. , 1999: The soil–precipitation feedback: A process study with a regional climate model. J. Climate, 12, 722741, doi:10.1175/1520-0442(1999)012<0722:TSPFAP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Schlosser, C. A., , and Milly P. C. D. , 2002: A model-based investigation of soil moisture predictability and associated climate predictability. J. Hydrometeor., 3, 483501, doi:10.1175/1525-7541(2002)003<0483:AMBIOS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Seneviratne, S. I., , Lüthi D. , , Litschi M. , , and Schär C. , 2006: Land–atmosphere coupling and climate change in Europe. Nature, 443, 205209, doi:10.1038/nature05095.

    • Search Google Scholar
    • Export Citation
  • Seth, A., , and Giorgi F. , 1998: The effects of domain choice on summer precipitation simulation and sensitivity in a regional climate model. J. Climate, 11, 26982712, doi:10.1175/1520-0442(1998)011<2698:TEODCO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Sheffield, J., , Goteti G. , , and Wood E. F. , 2006: Development of a 50-yr high-resolution global dataset of meteorological forcings for land surface modeling. J. Climate, 19, 30883111, doi:10.1175/JCLI3790.1.

    • Search Google Scholar
    • Export Citation
  • Shukla, J., , and Mintz Y. , 1982: Influence of land–surface evapotranspiration on the earth’s climate. Science, 215, 14981501, doi:10.1126/science.215.4539.1498.

    • Search Google Scholar
    • Export Citation
  • Siqueira, M., , Katul G. , , and Porporato A. , 2009: Soil moisture feedbacks on convection triggers: The role of soil–plant hydrodynamics. J. Hydrometeor., 10, 96112, doi:10.1175/2008JHM1027.1.

    • Search Google Scholar
    • Export Citation
  • Steiner, A. L., , Pal J. S. , , Rauscher S. A. , , Belel J. L. , , Diffenbaugh N. S. , , Boone A. , , Sloan L. C. , , and Giorgi F. , 2009: Land surface coupling in regional climate simulations of the West African monsoon. Climate Dyn., 33, 869892, doi:10.1007/s00382-009-0543-6.

    • Search Google Scholar
    • Export Citation
  • Sud, Y. C., , Mocko D. M. , , Lau K.-M. , , and Atlas R. , 2003: Simulating the midwestern U.S. drought of 1988 with a GCM. J. Climate, 16, 39463965, doi:10.1175/1520-0442(2003)016<3946:STMUDO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Sun, S. S., , and Wang G. L. , 2012: The complexity of using a feedback parameter to quantify the soil moisture-precipitation relationship. J. Geophys. Res., 117, D11113, doi:10.1029/2011JD017173.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , and Guillemot C. J. , 1996: Physical processes involved in the 1988 drought and 1993 floods in North America. J. Climate, 9, 12881298, doi:10.1175/1520-0442(1996)009<1288:PPIITD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wei, J., , and Dirmeyer P. A. , 2010: Toward understanding the large-scale land-atmosphere coupling in the models: Roles of different processes. Geophys. Res. Lett., 37, L19707, doi:10.1029/2010GL044769.

    • Search Google Scholar
    • Export Citation
  • Wei, J., , and Dirmeyer P. A. , 2012: Dissecting soil moisture-precipitation coupling. Geophys. Res. Lett., 39, L19711, doi:10.1029/2012GL053038.

    • Search Google Scholar
    • Export Citation
  • Wei, J., , Dickinson R. E. , , and Chen H. , 2008: A negative soil moisture–precipitation relationship and its causes. J. Hydrometeor., 9, 13641376, doi:10.1175/2008JHM955.1.

    • Search Google Scholar
    • Export Citation
  • Yeh, T. C., , Wetherald R. T. , , and Manabe S. , 1984: The effect of soil moisture on the short-term climate and hydrology change—A numerical experiment. Mon. Wea. Rev., 112, 474490, doi:10.1175/1520-0493(1984)112<0474:TEOSMO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Zeng, X. B., , Barlage M. , , Castro C. , , and Fling K. , 2010: Comparison of land–precipitation coupling strength using observations and models. J. Hydrometeor., 11, 979994, doi:10.1175/2010JHM1226.1.

    • Search Google Scholar
    • Export Citation
  • Zhang, J., , Wang W. C. , , and Leung L. R. , 2008a: Contribution of land–atmosphere coupling to summer climate variability over the contiguous United States. J. Geophys. Res., 113, D22109, doi:10.1029/2008JD010136.

    • Search Google Scholar
    • Export Citation
  • Zhang, J., , Wang W. C. , , and Wei J. , 2008b: Assessing land–atmosphere coupling using soil moisture from the Global Land Data Assimilation System and observational precipitation. J. Geophys. Res., 113, D17119, doi:10.1029/2008JD009807.

    • Search Google Scholar
    • Export Citation
  • Zhang, J., , Wu L. , , and Dong W. , 2011: Land–atmosphere coupling and summer climate variability over East Asia. J. Geophys. Res., 116, D05117, doi:10.1029/2010JD014714.

    • Search Google Scholar
    • Export Citation
  • Zheng, X. Y., , and Eltahir E. A. B. , 1998: A soil moisture–rainfall feedback mechanism: 2. Numerical experiments. Water Resour. Res., 34, 777785, doi:10.1029/97WR03497.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    The RegCM4 domain (shaded) and study area (within blue frame) over 0°–60°N, 40°–150°E.

  • View in gallery

    (a) Mean annual and seasonal precipitation (mm month−1) from CRU data during 1982–2001. As in (a), but for the difference in precipitation between (b) GLDAS and CRU (GLDAS minus CRU), (c) RegCM4 and CRU (RegCM4 minus CRU), and (d) CFSR and CRU (CFSR minus CRU). Here the CRU, RegCM4, and CFSR data are upscaled to 1° × 1°.

  • View in gallery

    As in Fig. 2, but for the standard deviation of precipitation.

  • View in gallery

    (a) Annual and seasonal averages of volumetric soil moisture in the upper 1-m depth from GLDAS data. As in (a), but for the difference between (b) RegCM4 and GLDAS (RegCM4 minus GLDAS) and (c) CFSR and GLDAS (CFSR minus GLDAS). Here the RegCM4 and CFSR data are upscaled to 1° × 1° to match the GLDAS data.

  • View in gallery

    As in Fig. 4, but for the standard deviation of volumetric soil moisture.

  • View in gallery

    As in Fig. 4, but for ET (mm day−1).

  • View in gallery

    As in Fig. 6, but for the standard deviation.

  • View in gallery

    Conditional-correlation related indicators between 1-day soil moisture and 21-day (left to right) precipitation [PC(P)], temperature [−PC(T)], evaporation [PC(ET) × σ(ET)], and EF [PC(EF) × σ(EF)] for GLDAS data during 1982–2001 in (top to bottom) spring [March–May (MAM)], summer [June–August (JJA)], and autumn [September–November (SON)]. Only the PC values that passed the 5% SI significance test are shaded. Note that as most PC(T) values are negative, plotted here are −PC(T)s for a color scale consistence with other panels in the figure.

  • View in gallery

    As in Fig. 8, but for CFSR reanalysis data during 1991–2010.

  • View in gallery

    As in Fig. 8, but for RegCM4 outputs during 1982–2001.

  • View in gallery

    (left to right) GLACE coupling strength indicators , , , and for RegCM4 in (top to bottom) spring (MAM), summer (JJA), and autumn (SON).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 89 89 13
PDF Downloads 75 75 5

Diagnosing the Strength of Land–Atmosphere Coupling at Subseasonal to Seasonal Time Scales in Asia

View More View Less
  • 1 State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing, China, and Department of Civil and Environmental Engineering, and Center for Environmental Sciences and Engineering, University of Connecticut, Storrs, Connecticut
  • | 2 Department of Civil and Environmental Engineering, and Center for Environmental Sciences and Engineering, University of Connecticut, Storrs, Connecticut
  • | 3 Department of Civil and Environmental Engineering, and Center for Environmental Sciences and Engineering, University of Connecticut, Storrs, Connecticut, and Computer Science and Mathematics Division, Climate Change Science Institute, Oak Ridge National Laboratory, Oak Ridge, Tennessee
  • | 4 State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing, China, and Department of Geoscience, University of Nevada, Las Vegas, Las Vegas, Nevada
  • | 5 State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing, China
© Get Permissions
Full access

Abstract

This paper focuses on diagnosing the strength of soil moisture–atmosphere coupling at subseasonal to seasonal time scales over Asia using two different approaches: the conditional correlation approach [applied to the Global Land Data Assimilation System (GLDAS) data, the Climate Forecast System Reanalysis (CFSR) data, and output from the regional climate model, version 4 (RegCM4)] and the Global Land–Atmosphere Coupling Experiment (GLACE) approach applied to the RegCM4. The conditional correlation indicators derived from the model output and the two observational/reanalysis datasets agree fairly well with each other in the spatial pattern of the land–atmosphere coupling signal, although the signal in CFSR data is stronger and spatially more extensive than the GLDAS data and the RegCM4 output. Based on the impact of soil moisture on 2-m air temperature, the land–atmosphere coupling hotspots common to all three data sources include the Indochina region in spring and summer, the India region in summer and fall, and north-northeastern China and southwestern Siberia in summer. For precipitation, all data sources produce a weak and spatially scattered signal, indicating the lack of any strong coupling between soil moisture and precipitation, for both precipitation amount and frequency. Both the GLACE approach and the conditional correlation approach (applied to all three data sources) identify evaporation and evaporative fraction as important links for the coupling between soil moisture and precipitation/temperature. Results on soil moisture–temperature coupling strength from the GLACE-type experiment using RegCM4 are in good agreement with those from the conditional correlation analysis applied to output from the same model, despite substantial differences between the two approaches in the terrestrial segment of the land–atmosphere coupling.

Corresponding author address: Dr. Guiling Wang, Department of Civil and Environmental Engineering, University of Connecticut, 261 Glenbrook St., Storrs, CT 06269. E-mail: gwang@engr.uconn.edu

Abstract

This paper focuses on diagnosing the strength of soil moisture–atmosphere coupling at subseasonal to seasonal time scales over Asia using two different approaches: the conditional correlation approach [applied to the Global Land Data Assimilation System (GLDAS) data, the Climate Forecast System Reanalysis (CFSR) data, and output from the regional climate model, version 4 (RegCM4)] and the Global Land–Atmosphere Coupling Experiment (GLACE) approach applied to the RegCM4. The conditional correlation indicators derived from the model output and the two observational/reanalysis datasets agree fairly well with each other in the spatial pattern of the land–atmosphere coupling signal, although the signal in CFSR data is stronger and spatially more extensive than the GLDAS data and the RegCM4 output. Based on the impact of soil moisture on 2-m air temperature, the land–atmosphere coupling hotspots common to all three data sources include the Indochina region in spring and summer, the India region in summer and fall, and north-northeastern China and southwestern Siberia in summer. For precipitation, all data sources produce a weak and spatially scattered signal, indicating the lack of any strong coupling between soil moisture and precipitation, for both precipitation amount and frequency. Both the GLACE approach and the conditional correlation approach (applied to all three data sources) identify evaporation and evaporative fraction as important links for the coupling between soil moisture and precipitation/temperature. Results on soil moisture–temperature coupling strength from the GLACE-type experiment using RegCM4 are in good agreement with those from the conditional correlation analysis applied to output from the same model, despite substantial differences between the two approaches in the terrestrial segment of the land–atmosphere coupling.

Corresponding author address: Dr. Guiling Wang, Department of Civil and Environmental Engineering, University of Connecticut, 261 Glenbrook St., Storrs, CT 06269. E-mail: gwang@engr.uconn.edu

1. Introduction

As a main land surface state variable, soil moisture affects the subseasonal to seasonal variability of the atmosphere by influencing the surface water and energy budget through its effects on evapotranspiration (ET). In several different regions, it was found that the anomalies in surface fluxes induced by soil moisture anomalies can lead to precipitation anomalies (Delworth and Manabe 1989; Koster and Suarez 1995; Zheng and Eltahir 1998; Schlosser and Milly 2002). While anomalies in the atmosphere can dissipate within a matter of days, the process of soil moisture depletion through ET is slow, with a characteristic time scale ranging from weeks to months. Atmospheric response to soil moisture anomalies (originally induced by precipitation anomalies) can therefore provide a mechanism for precipitation persistence, contributing to the memory of the hydrometeorological system.

The topic of land–atmosphere coupling has been studied for decades based on both numerical modeling experiments and observational data analysis. The list of publications on land–atmosphere interactions is long and extensive. In the modeling field, earlier studies (Namias 1952, 1959; Shukla and Mintz 1982; Yeh et al. 1984) found strong sensitivity of the climate response to ET and soil moisture anomalies in North America using atmospheric general circulation models (AGCMs). Later on, the role of soil moisture in seasonal climate extremes, especially in drought and flood years, were examined (e.g., Atlas et al. 1993; Trenberth and Guillemot 1996; Sud et al. 2003), and the focus has also been shifted from the impact of perpetuating soil moisture anomalies to the impact of initial soil moisture anomalies (e.g., Seth and Giorgi 1998; Hong and Pan 2000; Kim and Wang 2007). More recently, the Global Land–Atmosphere Coupling Experiment (GLACE) project proposed the concept of land–atmosphere coupling strength (Koster et al. 2004b, 2006; Guo et al. 2006) and quantified the strength based on the average among 12 participating GCMs. Several hotspots for soil moisture–precipitation and soil moisture–temperature coupling were found in the transition zones between dry and wet climate regimes, such as the central United States, West Africa, and India in boreal summer.

Regional climate models were also widely used in studies of land–atmosphere interactions (e.g., Giorgi et al. 1996; Bosilovich and Sun 1999; Schär et al. 1999; Kim and Hong 2007; Seneviratne et al. 2006; Leung et al. 2011; Zhang et al. 2008a, 2011). In most of the studies, soil moisture feedback has been demonstrated to enhance precipitation and temperature variability and other climate extremes such as heat waves. For example, Seneviratne et al. (2006) found that the warming-induced increase of summer temperature variability in Europe is mainly attributed to the feedback between land surface and the atmosphere. Zhang et al. (2008a) identified areas of strong soil moisture–precipitation feedback in the northern United States in the summer season using the Weather Research and Forecasting Model (WRF); later, Zhang et al. (2011) used the same model to study the soil moisture impact on precipitation and temperature over East Asia and found that land–atmosphere interaction plays an important role in summer climate variability, especially over the climatic and ecological transition zones (consistent with findings of Koster et al. 2004b).

In the observational field, most of the studies (e.g., Findell and Eltahir 1997; Salvucci et al. 2002; D’Odorico and Porporato 2004; Ruiz-Barradas and Nigam 2005, 2006; Dirmeyer et al. 2006; Wei and Dirmeyer 2010, 2012; Zhang et al. 2008b; Mei and Wang 2011, 2012) supported a positive soil moisture–precipitation feedback, and some (Findell and Eltahir 2003; Ek and Holtslag 2004; Siqueira et al. 2009) demonstrated a negative feedback. For instance, Findell and Eltahir (1997) found a positive correlation between soil moisture and subsequent precipitation based on observational data in Illinois, and D’Odorico and Porporato (2004), using the same data, found that soil moisture appears to influence the frequency of subsequent precipitation instead of the amount. Zhang et al. (2008b) assessed the land–atmosphere coupling strength globally using the correlation between soil moisture from the Global Land Data Assimilation System (GLDAS) and observational precipitation and found that regions of strong correlation include the northern continental United States, the Sahel, southern Europe, central Eurasia, the region from Mongolia to northern China, and southwestern China. Mei and Wang (2011, 2012) proposed a conditional correlation analysis to quantify soil moisture–precipitation coupling strength and found that soil moisture feedback tends to enhance subsequent precipitation anomalies during years when the precipitation anomalies are large. Findell and Eltahir (2003) investigated the impact of soil moisture on triggering and development of convection in regions of the United States and found that different early morning atmospheric conditions can lead to a “wet soil advantage regime” (positive feedback) or a “dry soil advantage regime” (negative feedback).

Compared to the regions of North America and Europe, the impact of soil moisture on seasonal climate in Asia remains poorly understood. While it is included in several global analyses, focused regional studies are relatively few, and findings on the role of soil moisture in Asia are inconclusive. For example, the GLACE study demonstrated a strong signal of soil moisture feedback on warm season climate variability over northern India, central China, and part of western Asia (Koster et al. 2004b, 2006), which was later confirmed in a regional climate modeling study by Zhang et al. (2011). In contrast, Douville et al. (2001) found that soil moisture has a relatively weak impact on precipitation in Asia; Kim and Hong (2007) also found that the initial soil moisture anomalies on the summer rainfall in East Asia are not significant based on one regional climate model, though their impact on dynamic circulation is obvious.

In this paper, we present a focused regional study on the role of soil moisture in the subseasonal to seasonal climate of Asia based on the regional climate model, version 4 (RegCM4), and two different datasets: GLDAS and the National Centers for Environmental Prediction (NCEP) Climate Forecast System Reanalysis (CFSR) data. Two methods are applied here to quantify the strength of land–atmosphere coupling: the conditional correlation analysis method proposed by Mei and Wang (2011) and the GLACE numerical experimental approach (Koster et al. 2004b). The data and model description (with basic evaluation of datasets) are introduced in section 2. The conditional correlation method and its results are presented in section 3. Design of the GLACE-1-type experiment and its results are documented in section 4. Summary and discussions are given in section 5.

2. Data and model

In the absence of an observational dataset for soil moisture over the domain of interest, the land–atmosphere coupling in the GLDAS data and the CFSR reanalysis data are assessed and used for the model–data comparison. For each dataset, the soil moisture content in the upper 1-m depth for GLDAS and CFSR data will be used, and for the RegCM4, soil moisture content in the first seven layers (summing up to about 1.25 m) prorated to 1-m depth will be used. Also used in the conditional correlation analysis are daily precipitation (including both snow and rainfall), 2-m air temperature, latent heat flux (evaporation), and sensible heat flux [used to calculate the evaporative fraction (EF)]. The EF is defined as the ratio of daily latent heat flux to daily net surface radiation, and here the ratio of latent heat flux to the sum of sensible and latent heat fluxes is used as an approximation. The descriptions on the GLDAS data, CFSR reanalysis data, and the RegCM4 output and their comparisons are presented as follows.

a. GLDAS data

GLDAS data provide optimal fields of land surface states and fluxes with advanced land surface modeling and data assimilation techniques (Rodell et al. 2004). Currently, four land surface models (LSMs) are used in GLDAS: catchment, Noah, the Community Land Model (CLM), and the variable infiltration capacity (VIC) model. Each model is integrated with a large quantity of observation-based data and executed globally at high resolution (2.5° to 1 km) by the Land Information System (LIS; Kumar et al. 2006).

In this study, the newly published (October 2012) 1° × 1° 3-hourly GLDAS-2 dataset from the Noah model (note data from other models are not available yet) during the 1982–2001 time period is used (http://disc.sci.gsfc.nasa.gov/services/grads-gds/gldas). Compared with the previous version, GLDAS-1, spanning 1979 to present, GLDAS-2 data are extended from 1948 and generated with upgraded land surface models driven by the global meteorological forcing dataset from Princeton University (Sheffield et al. 2006), which fixed some previous problems in the forcing data. The derivation of GLDAS-2 meteorological forcing data involves downscaling from the NCEP–National Center for Atmospheric Research (NCAR) reanalysis data and bias correction against the University of East Anglia Climate Research Unit (CRU) data (for temperature) and the Global Precipitation Climatology Project data (for precipitation; Sheffield et al. 2006). Other improvements include the switch to Moderate Resolution Imaging Spectroradiometer (MODIS)-based land surface parameter datasets and incorporation of soil moisture initialization over desert. In addition, the bottom layer temperature dataset in the Noah model was also updated. The GLDAS dataset has been used in subseasonal climate forecast studies (e.g., Koster et al. 2004a; de Goncalves et al. 2006) and land–atmosphere coupling studies (e.g., Zhang et al. 2008b).

b. CFSR data

The NCEP CFSR dataset surpasses previous NCEP–NCAR reanalysis datasets owing to the benefits of using a higher resolution and a coupled atmosphere–ocean–land surface–sea ice technique. It covers the period from 1979 to present, with a 6-h temporal resolution and 0.5° spatial resolution (Saha et al. 2010). The primary improvements in this latest reanalysis product derive from 1) coupling to the ocean during the generation of the 6-h guess field, 2) an interactive sea ice model, and 3) the assimilation of satellite radiances for the entire period. In addition, the much higher horizontal and vertical resolution (T382L64) of the atmospheric model and assimilation improvements (including real-time monitoring) over the last 10–15 yr and the use of prescribed temporally varying CO2 concentrations also led to substantial improvements. The land surface model used in CFSR is the Noah land model. Note that this is the same land model from which the GLDAS-2 soil moisture data are derived.

Precipitation in CFSR data over a region in North Asia during the first decade of its data coverage differs significantly from the rest, which could lead to spuriously high soil moisture–precipitation correlation (not shown here). For example, the pre-1988 precipitation climatology in Inner Mongolia is more than twice of that of the post-1988 period, a feature not present in CRU or GLDAS data. To avoid such uncertainties, only data during 1991–2010 are used in this study.

c. RegCM4 output

The Abdus Salam International Center for Theoretical Physics (ICTP) RegCM4 (Steiner et al. 2009; Elguindi et al. 2011; Giorgi et al. 2012) is used. A previous study (Gu et al. 2012) has demonstrated that this model performs well in simulating the observed climate mean, variability, and extremes in most of Asia. The present study uses the exact same model configuration as Gu et al. (2012), including a 50-km horizontal resolution, 18 vertical levels, the CLM3.5 land surface scheme, and the Massachusetts Institute of Technology–Emanuel (MIT–EMAN) convection scheme. The domain is centered at 30°N and 100°E and consists of 264 (west–east) × 150 (south to north) grid points, covering most of Asia and the adjacent oceans (Fig. 1). The analysis will focus on the area over 0°–60°N, 40°–150°E (within the blue frame in Fig. 1).

Fig. 1.
Fig. 1.

The RegCM4 domain (shaded) and study area (within blue frame) over 0°–60°N, 40°–150°E.

Citation: Journal of Hydrometeorology 15, 1; 10.1175/JHM-D-13-0104.1

A continuous 22-yr simulation (referred to as the CTL run) is conducted for the period 1980–2001, driven with lateral boundary forcing from the NCEP–NCAR Reanalysis Project (NNRP) data (Kalnay et al. 1996; Kanamistu et al. 2002) and observed oceanic forcing from the Met Office Global Sea Surface Temperature (GISST; Rayner et al. 2006) of the same period. This simulation serves two purposes: to produce a single realization of the past land–atmosphere system for comparison with the GLDAS and CFSR data and to generate atmospheric initial conditions for the designed GLACE-1-type experiments (introduced in section 4a). The first 2 yr are discarded as model spinup, and output from the remaining 20 yr will be used for data evaluation and conditional correlation analysis.

d. Data evaluation

To verify the RegCM4 and the GLDAS and CFSR products against observational data, we compare the model and reanalysis data climatology for precipitation and temperature with the University of East Anglia CRU data version TS3.1. The RegCM4 output, GLDAS, and CFSR data are therefore aggregated into monthly data for this comparison. Note that the periods of the CFSR data do not completely overlap with the RegCM4 output and the GLDAS data. However, we assume that within this relatively short duration the characteristics of mean and standard deviation of precipitation and 2-m air temperature and the land–atmosphere coupling strength remain more or less the same and are therefore comparable.

The mean and standard deviation of the annual and seasonal precipitation from RegCM4, GLDAS, and CFSR data all show a very similar overall spatial pattern that agrees well with the CRU data (results not shown). Here the difference fields are presented (Figs. 2, 3) to allow for a more detailed comparison with the CRU data in the magnitudes of mean and standard deviation. The GLDAS precipitation data statistics closely resemble the CRU station-based data, which is not surprising as derivation of the GLDAS data involved bias correction against observation-based data (Sheffield et al. 2006). The only difference is found along the Himalaya region. The RegCM4 suffers from strong overestimation of precipitation over the Himalaya and Tibetan Plateau region in all seasons. Overestimation is also seen over Indochina in all seasons. Precipitation over India is underestimated, especially during the summer monsoon season. CFSR data suffer from similar overestimation of precipitation. Along the maximum precipitation belt over the Himalayas, RegCM4 and CFSR (which is largely model derived) are very similar in their precipitation spatial patterns and magnitude, and they both overestimate precipitation compared with observations (CRU and GLDAS). The good agreement between RegCM4 and CFSR is rather interesting, and the discrepancy with observations may to some extent also reflect uncertainties in observations (rather than just biases in the models). Note that the complex topography in the Himalaya region leads to strong spatial heterogeneity of precipitation that may not be captured by the sparse meteorological stations in that region. Both the GLDAS data and RegCM4 perform fairly well in capturing the standard deviation of precipitation, except over India during the monsoon when RegCM4 underestimates the precipitation standard deviation (Fig. 3). CFSR data show strong overestimation of precipitation standard deviation throughout the whole domain.

Fig. 2.
Fig. 2.

(a) Mean annual and seasonal precipitation (mm month−1) from CRU data during 1982–2001. As in (a), but for the difference in precipitation between (b) GLDAS and CRU (GLDAS minus CRU), (c) RegCM4 and CRU (RegCM4 minus CRU), and (d) CFSR and CRU (CFSR minus CRU). Here the CRU, RegCM4, and CFSR data are upscaled to 1° × 1°.

Citation: Journal of Hydrometeorology 15, 1; 10.1175/JHM-D-13-0104.1

Fig. 3.
Fig. 3.

As in Fig. 2, but for the standard deviation of precipitation.

Citation: Journal of Hydrometeorology 15, 1; 10.1175/JHM-D-13-0104.1

Relative to precipitation, the agreement among model, observation, and reanalysis data is much better for mean air temperature (results not shown). The only noticeable differences are in the Tibetan Plateau region, where both RegCM4 and CFSR show some cold bias, and in northeastern India, where RegCM4 has a warm bias. Standard deviation of temperature shows a larger degree of variation (results not shown), with CFSR overestimating the standard deviation over the Tibetan Plateau and RegCM4 underestimating over most of the domain.

From the above analysis, the 2-m temperature and precipitation from the GLDAS data closely resemble the CRU data, while strong biases exist over some regions in the RegCM4 and more so in the CFSR data. For soil moisture and ET (for which observational data are scarce), the GLDAS data are considered the best approximation to observation since they were generated by offline land surface models driven with realistic meteorological forcing and can thus avoid the biases related to model-produced atmospheric forcing (such as those in CFSR). Therefore, in the rest of the analysis, GLDAS data will be used as the reference for both soil moisture and ET data, and the land–atmosphere coupling strength based on GLDAS data will be considered the benchmark for RegCM4 and CFSR to compare against. However, it is important to keep in mind that even the coupling strength based on GLDAS is still subject to large uncertainties related to two factors: one is the potential deficiency of the Noah land model used to derive the soil moisture data, and the other is the lack of independency between the source of the soil moisture data and the source of data for other climate variables since the latter was used as the driving forcing for the former in GLDAS.

Soil moisture averaged in the top 1 m is used for all the conditional correlation analysis. Other depths (e.g., 0.4 and 2 m) were experimented on, and qualitatively, results from conditional correlation analysis are not sensitive to the depth of soil chosen. The mean and standard deviation for the top 1-m depth soil moisture are compared in Figs. 4 and 5, respectively. The main difference between RegCM4 output and GLDAS data exists over the Asian monsoon regions (including southeastern China, Indochina, and part of India), where the RegCM4 underestimates the mean soil moisture content. The CFSR reanalysis data tend to overestimate soil moisture, especially over central Asia (northwestern China). Compared with the GLDAS data, RegCM4 underestimates the standard deviation of soil moisture, especially in northeastern Asia and India, while CFSR overestimates the standard deviation of soil moisture over eastern China and part of the Tibetan Plateau (Fig. 5). However, different datasets agree well in the spatial pattern of soil moisture standard deviation (results not shown).

Fig. 4.
Fig. 4.

(a) Annual and seasonal averages of volumetric soil moisture in the upper 1-m depth from GLDAS data. As in (a), but for the difference between (b) RegCM4 and GLDAS (RegCM4 minus GLDAS) and (c) CFSR and GLDAS (CFSR minus GLDAS). Here the RegCM4 and CFSR data are upscaled to 1° × 1° to match the GLDAS data.

Citation: Journal of Hydrometeorology 15, 1; 10.1175/JHM-D-13-0104.1

Fig. 5.
Fig. 5.

As in Fig. 4, but for the standard deviation of volumetric soil moisture.

Citation: Journal of Hydrometeorology 15, 1; 10.1175/JHM-D-13-0104.1

ET in both RegCM4 and CFSR closely resembles the GLDAS data in a spatial pattern over the whole domain, and the agreement in magnitude is also very good for a major portion of the domain. The main difference is found over South and East Asia during the monsoon season when both RegCM and CFSR overestimate ET by more than 1 mm day−1 (Fig. 6). The difference in the standard deviation of ET is heavily influenced by mean ET, with overestimation of ET standard deviation where mean ET is overestimated (Fig. 7). The RegCM4 and CFSR data are very similar in their ET statistics (in terms of both their spatial patterns and magnitude), despite rather major differences between the two in soil moisture.

Fig. 6.
Fig. 6.

As in Fig. 4, but for ET (mm day−1).

Citation: Journal of Hydrometeorology 15, 1; 10.1175/JHM-D-13-0104.1

Fig. 7.
Fig. 7.

As in Fig. 6, but for the standard deviation.

Citation: Journal of Hydrometeorology 15, 1; 10.1175/JHM-D-13-0104.1

3. Conditional correlation analysis

According to previous studies (Koster et al. 2004b, 2009; Guo et al. 2006; Wei and Dirmeyer 2012; Mei and Wang 2012), evaporation or EF acts as a potential link for the coupling between soil moisture (SM) and precipitation (P)–temperature (T). For example, in the GLACE study, Koster et al. (2004b) found that strong SM–P coupling can only occur in climate transition zones where evaporation is relatively high but still sensitive to soil moisture variation. Later, Guo et al. (2006) decomposed the impact of soil moisture on climate variables (precipitation and temperature) into two segments (in the context of the GLACE study): the terrestrial segment with soil moisture impact on evaporation and the atmospheric segment with evaporation impact on precipitation and temperature. They found that the terrestrial segment plays an important role in explaining SM–PT coupling. Recently, Wei and Dirmeyer (2012) also dissected the SM–P coupling analytically into soil moisture–evaporation (SM–ET) coupling and evaporation–precipitation coupling and demonstrated that the SM–ET coupling is the principal factor in determining the spatial pattern and variation of SM–P coupling. On the other hand, Koster et al. (2009) found that evaporative control (through SM–EF relationship) explains well the drought-induced warming. Mei and Wang (2012) found strong lagged conditional correlation for SM–EF in areas of strong conditional correlation for SM–PT. Therefore, in this study, in addition to quantifying the SM–PT coupling, the coupling between soil moisture and evaporation–EF will be also examined based on the conditional correlation analysis.

a. Methodology

Here the conditional correlation method of Mei and Wang (2011, 2012) is used to quantify the coupling between soil moisture and various surface climate variables based on data from different sources, including the GLDAS data, the CFSR reanalysis data, and RegCM4 output. The 1-day soil moisture and the subsequent 21-day average of climate variables will be paired up in the conditional correlation analysis to accommodate potential comparison of results with previous studies (Findell and Eltahir 1997; D’Odorico and Porporato 2004; Mei and Wang 2011, 2012).

Prior to the conditional correlation analysis, raw gridded data in each dataset spanning 20 yr are preprocessed. First, daily precipitation, evaporation, EF, and 2-m temperature data are used to derive the 21-day running means for each variable. Then, both the 1-day soil moisture and 21-day climate variables are normalized using the mean and standard deviation corresponding to the time of the year in order to remove the seasonal cycle. After that, the conditional correlation between 1-day soil moisture and the subsequent 21-day climate variables (precipitation, 2-m temperature, evaporation, and EF) will be estimated based on each dataset using a two-step approach (Mei and Wang 2011) as follows.

  1. At each grid cell, the 20-yr data are divided into two categories, with 10 yr in each category according to the amount of grid-based seasonal precipitation. Specifically, the grid-based seasonal precipitation is used as the criterion to rank the years; the first and fourth quartiles are grouped together as the outer-quartiles category, and the second and third quartiles are grouped as the inner-quartiles category. Thus, the total number of data pairs (e.g., 1-day soil moisture accompanied with subsequent 21-day precipitation) in each category pool is 92 × 10 for the spring and summer seasons and 91 × 10 for the autumn season.
  2. The probability density function of the conditional correlation coefficient between 1-day soil moisture and subsequent 21-day climate variables is estimated in each category for each grid cell. Specifically, the PDF is derived from 10 000 repeatedly calculated correlations, each based on 16 data pairs randomly drawn without replacement from the data pool in each category and with a constraint to exclude temporal overlapping of climate variables between any two data pairs. Two parameters will be used to evaluate the strength of coupling between soil moisture and climate variables: the correlation coefficient at the peak probability density [referred to as the peak correlation (PC)] and the probability of the correlation coefficient being negative when the PC value is positive or being positive when the PC value is negative [referred to as the significance index (SI)]. In this study, PC values with SI lower than 0.1 will be considered significant.

Similar tothe product index proposed in Guo et al. (2006) and Dirmeyer (2011), a new index PC(υ) × σ(υ) (where υ stands for ET or EF) is proposed here to evaluate the strength of the terrestrial segment of the land–atmosphere coupling. Here the standard deviation σ(υ) is derived from the same conditional quartiles as PC(υ). A major pathway for soil moisture to influence climate is through its impact on ET (the terrestrial segment), which then influences the atmospheric boundary layer and convective processes (the atmospheric segment). For soil moisture to influence climate, in the atmospheric segment, the atmospheric processes must be sensitive to ET variations and the range of ET variation must be large enough; in the terrestrial segment, ET must be sensitive to soil moisture variations and the range of soil moisture variation must be large enough to cause a large range of ET variation. With PC(υ) measuring the impact of soil moisture on ET (or EF) and σ(υ) measuring the range of ET variation (or EF variation), the product index proposed here is conceptually similar to the sensitivity index of Dirmeyer (2011) (except that the slope of linear regression between soil moisture and ET is replaced by the peak correlation in our conditional correlation analysis here while the standard deviation of soil moisture is replaced by that of ET). Obviously, both PC(υ) and σ(υ). have to be suitably large in order to have a large value for PC(υ) × σ(υ). In this paper, only the PC(υ) × σ(υ) results in areas where PC(υ) passes the 10% significance test are presented.

b. Results on conditional correlation analysis

For consistency with the RegCM4 output, soil moisture within 1-m depth from the GLDAS and CFSR dataset is used for analysis. Extending the soil moisture depth to 2 m does not cause any qualitative difference in results. Similar to previous studies (Mei and Wang 2011, 2012), the signals in the inner quartiles are quite weak (although spatially consistent with the outer quartiles). Thus, only results from the outer quartiles are documented here. The PC values (for precipitation and temperature) and product index PC(υ) × σ(υ) (for evaporation and EF) derived from the GLDAS data, CFSR reanalysis data, and the RegCM4 outputs in the outer quartiles are presented in Figs. 810, respectively. Only the PC(υ) values passing the SI significance test (i.e., with SI < 0.1) are shaded.

Fig. 8.
Fig. 8.

Conditional-correlation related indicators between 1-day soil moisture and 21-day (left to right) precipitation [PC(P)], temperature [−PC(T)], evaporation [PC(ET) × σ(ET)], and EF [PC(EF) × σ(EF)] for GLDAS data during 1982–2001 in (top to bottom) spring [March–May (MAM)], summer [June–August (JJA)], and autumn [September–November (SON)]. Only the PC values that passed the 5% SI significance test are shaded. Note that as most PC(T) values are negative, plotted here are −PC(T)s for a color scale consistence with other panels in the figure.

Citation: Journal of Hydrometeorology 15, 1; 10.1175/JHM-D-13-0104.1

Fig. 9.
Fig. 9.

As in Fig. 8, but for CFSR reanalysis data during 1991–2010.

Citation: Journal of Hydrometeorology 15, 1; 10.1175/JHM-D-13-0104.1

Fig. 10.
Fig. 10.

As in Fig. 8, but for RegCM4 outputs during 1982–2001.

Citation: Journal of Hydrometeorology 15, 1; 10.1175/JHM-D-13-0104.1

For all variables examined, signals from the CFSR data (shown in Fig. 9) are more spatially continuous than that from GLDAS data (Fig. 8) and RegCM4 modeling (Fig. 10). One of the regions of large difference is the Tibetan Plateau, where data from different sources differ significantly in climatology of temperature and precipitation (Figs. 2, 3). As mentioned in section 2, a large degree of uncertainty exists in this region because of its complex terrain and the sparseness of observational stations that influences both the GLDAS data (through bias correction) and the CFSR data (through data assimilation). It therefore will be excluded from the rest of our analysis. With the Tibetan region excluded, the spatial pattern of strong signal for land–atmosphere coupling diagnosed in CFSR data, the GLDAS data, and the RegCM4 output are fairly consistent, although local differences (some of which are rather substantial) do exist.

For precipitation, the conditional correlation method shows a weak SM–P feedback, with scattered PC(P) signals from all three data sources mainly in the mid- and high latitudes. Where signal does exist, it tends to indicate a positive SM–P feedback. The SM–T feedback is stronger in all datasets, and the signal for significant PC(T) values (Figs. 810) shows some consistent pattern among different seasons. Areas of strong signal common to all three datasets (CFSR data, RegCM4, and GLDAS) include 1) the Indochina region in spring and summer, 2) India in summer and autumn, and 3) northern and northeastern China and southwestern Siberia in summer (with some spatial shift of signal in the GLDAS data). The CFSR data show a stronger signal for the SM–T feedback that is spatially much more extensive than the other two datasets in all three seasons. Most of the hotspots identified by GLDAS and RegCM4 are located within the hotspots identified by the CFSR data, with an exception in spring over India, where only RegCM4 supports a strong SM–T feedback and the CFSR and GLDAS data show little signal. Additional areas of strong SM–T feedback are found for the CFSR data in the Indochina region in autumn (Fig. 9).

Consistent with the notion that ET plays a major role in linking soil moisture to precipitation and temperature, all the hotspots for SM–P and SM–T coupling are within the hotspots of SM–ET coupling and SM–EF coupling, as shown by the PC(ET) × σ(ET) and PC(EF) × σ(EF) metrics in Figs. 810. The signals for the SM–ET coupling and SM–EF coupling are remarkably similar in their spatial pattern, and both collocated with strong SM–T coupling. The only exception is found over the Indochina region in the June–August (JJA) season, where the SM–EF signal better resembles that of the SM–T coupling while the SM–ET signal is missing.

Comparison of the SM–ET and SM–EF coupling signals among the three data sources suggests that the RegCM4 closely resembles the GLDAS data in both magnitude and spatial pattern, except over India, where RegCM4 overestimates the coupling strength compared with GLDAS. In contrast, CFSR data show a much stronger signal than both RegCM4 and GLDAS, especially over eastern China. Note that for all three data sources in Figs. 810, the “color saturating” large values for the EF product index are caused by very low net radiation in mid- and high latitudes giving rise to spuriously high EF and its standard deviation.

Based on the overall spatial pattern of the coupling strength indices, especially for temperature and ET as well as EF, it seems that the range of soil moisture variation (as reflected by the standard deviation of soil moisture in Fig. 5) plays an important role influencing the land–atmosphere coupling strength. As suggested in previous studies (e.g., Koster et al. 2006), the land–atmosphere coupling is strong over the transition zones between dry and wet climates. It is over such transition zones that the range of soil moisture variation (and therefore the range of ET variation) is large enough to influence atmospheric processes. Comparing the coupling strength indices with the soil moisture coefficient of variation (results not shown) indicates that strong signals for the two tend to be collocated. On a similar note, the overly strong signal in the CFSR data compared with the RegCM4 and GLDAS data might be also related to the fact that soil moisture standard deviation and coefficient of variation are much larger in the CFSR data.

In addition to precipitation amount, the soil moisture’s impact on precipitation frequency is also examined. Contrary to the finding that soil moisture may impose more significant influence on precipitation frequency than on precipitation amount in North America (e.g., D’Odorico and Porporato 2004; Mei and Wang 2011; Findell et al. 2011), results from all three datasets indicate that in Asia the impact of soil moisture on precipitation frequency (not shown) is much weaker than its impact on precipitation amount. Here frequency is defined based on precipitation at the daily time scale and thus may not completely account for the impact of processes or mechanisms at subdaily time scales.

4. GLACE experiments using RegCM4

The conditional correlation analysis provides one metric for the land–atmosphere coupling strength that can be easily applied to any existing dataset, as shown in section 3. In the context of numerical models, the GLACE approach of Koster et al. (2006) offers a unique, widely used framework along with an index to quantify the land–atmosphere coupling strength. Here we examine the strength of land–atmosphere coupling in the RegCM4 following the GLACE approach and compare the GLACE coupling strength index with the conditional correlation results based on output from the same model (in section 3).

a. Experimental design and diagnostic index description

Following the GLACE experimental design (Koster et al. 2004b), two ensembles are defined here: a W ensemble and an S ensemble, each containing 16 ensemble members. The atmospheric initial conditions generated from the CTL run (as described in section 2) at the beginning of each season (spring, summer, and autumn) during the 16-yr period 1986–2001 are used to initialize the corresponding experiments in the W ensemble (W1, W2, …, W16) and S ensemble (S1, S2, …, S16) for each season. For lateral boundary conditions, both ensembles are driven with the NNRP forcing and SST forcing from the year 1994. The year 1994 is chosen because it is neither an El Niño nor a La Niña year. The S ensemble experiments are otherwise the same as the W ensemble, except that soil moisture in all S ensemble experiments is prescribed based on soil moisture output from one single W ensemble experiment, W1. Here the experiment for 1986 is chosen as the W1 experiment. If a different member of the W ensemble is chosen as the W1 experiment, the results will differ, but only slightly (Koster et al. 2006; Guo et al. 2006). Unlike the GLACE experiment of Koster et al. (2004b) that focused on deeper layers where the soil moisture variation is slow, both soil ice and liquid soil moisture in each of the 10 soil layers at each time step are saved from the W1 experiment to overwrite those in the S ensemble in this study. So the impact of all soil moisture dynamics (as opposed to just the slow components) is retained in our GLACE-type experiment.

Following Koster et al. (2002), a diagnostic is defined to measure the intra-ensemble similarity of variables of interest (e.g., precipitation, temperature, and evaporation), and the difference between the two ensembles is used to quantify the soil moisture impact on atmospheric variable υ. Similar to Koster et al. (2006), the soil moisture’s control on “synoptic scale” climate variability is examined, that is, the variability of precipitation or 2-m temperature on time scales of about a week. In order to avoid the problems associated with initial “shocks” to the modeled atmosphere, data in the first week of each ensemble member are discarded and the remaining data are aggregated into time series of 6-day totals; thus, each member simulation has a time series of 14 six-day totals, and each ensemble has 16 such time series (with 224 six-day totals). The index, which quantifies the similarity of the 16 time series within each ensemble, is estimated according to the following equation:
e1
where is the variance of the ensemble-averaged time series for the variable υ (which is estimated based on a time series of 14 six-day totals, each representing the average across the 16 members), and is the variance of all the 224 six-day totals. As such, the difference in the Ω index between the W and S ensembles measures the degree to which soil moisture controls the temporal variation of the climate variable.

According to previous studies (e.g., Guo et al. 2006; Koster et al. 2009), the SM–ET (or EF) link is an important mechanism of terrestrial segment for land–atmosphere coupling, and the product index can be used to measure the strength of the SM–ET link. Here the is the GLACE index for evaporation and is the standard deviation generated from the W ensembles. In this study, such product index for both evaporation and evaporative fraction will be assessed.

b. Results on GLACE-type experiments

The GLACE index for both precipitation and surface temperature and the product index for both ET and EF over three seasons (spring, summer, and autumn) are shown in Fig. 11. Areas where the magnitude passes the 5% significance test are shaded. The distribution of is very scattered across the whole domain in all seasons, indicating that overall the coupling signal for precipitation is weak. Northern India in summer is the only region/season that stands out with a strong signal. Different from the scattered signals, the signals for are spatially continuous and evident in all three seasons. For instance, both the India region and part of northeastern China stand out as the hotspots of coupling in summer. In spring and autumn, the southern India region is the only hotspot. Except for the mid- and high-latitude regions where low net radiation (and therefore high EF and σEF) in cold seasons leads to spuriously high product index for EF, the signal elsewhere is very similar for and , and they both resemble the spatial pattern for SM–T coupling strength ().

Fig. 11.
Fig. 11.

(left to right) GLACE coupling strength indicators , , , and for RegCM4 in (top to bottom) spring (MAM), summer (JJA), and autumn (SON).

Citation: Journal of Hydrometeorology 15, 1; 10.1175/JHM-D-13-0104.1

The GLACE-type experimental results (Fig. 11) are consistent with the conditional correlation results based on the RegCM4 output (Fig. 10) in showing a weak SM–P feedback (i.e., the scattered in Fig. 10 and PC(P) in Fig. 11). No spatially coherent signal of SM–P coupling was found by either method. For the SM–T coupling, although the hotspots identified by the conditional correlation method are slightly more extensive than the GLACE experimental approach, both methods identify the following regions as hotspots for SM–T coupling: the Indian subcontinent and southwestern Indochina in all seasons and northeastern China in summer. The signal is spatially more extensive in summer than in other seasons. In addition, both the GLACE coupling strength index and the conditional correlation values [PC(ET) × σ(ET) and PC(EF) × σ(EF)] suggest strong impact of soil moisture on ET and EF over the hot spots for SM–T coupling, and both the SM–EF and SM–ET product indices are effective reflections of the SM–T coupling strength.

Substantial differences between the GLACE coupling strength and the conditional correlation results also exist in several regions, especially for the SM–ET and SM–EF coupling. The most notable are the mid- and high-latitude regions, where the conditional correlation suggests a strong signal while the GLACE index is not significant. Over India, where both methods suggest a strong coupling, the seasonality of the signal differs between the two, with not much difference in the conditional correlation and a strong contrast between summer and other seasons in the GLACE index.

Despite the substantial difference between the two methods in identifying areas of strong SM–ET or SM–EF coupling (i.e., the terrestrial segment of the land–atmosphere coupling), the areas of strong SM–T coupling identified by the two methods agree reasonably well. This decent agreement between the conditional correlation analysis and the GLACE-1 experimental analysis further indicates that the PV(υ)value combined with the SI significance test can provide an effective metric to quantify the soil moisture–atmosphere coupling.

5. Summary and discussion

This study diagnoses the strength of land–atmosphere coupling in Asia during spring, summer, and autumn seasons based on two different methods: the conditional correlation analysis method proposed by Mei and Wang (2011) and the classical GLACE approach (Koster et al. 2004b). One regional climate model (RegCM4) and two datasets (GLDAS and CFSR) are used here, and the coupling strength in GLDAS data is considered the benchmark in this study. The conditional correlation analysis approach can be applied to both model output and observational/reanalysis data, while the GLACE approach can only be applied to numerical models. Major findings are summarized as follows.

  1. Based on the conditional correlation method, the land–atmosphere coupling in RegCM4 is slightly stronger than in GLDAS data, but weaker than in CFSR data. Common to all three datasets, the signals for soil moisture–precipitation coupling are weak and are spatially scattered, with the only obvious hotspot located in part of the mid- and high-latitude (above 40°N) region in the summer (JJA) season. For 2-m temperature, the coupling strength signals are consistent and continuous in spatial distribution. The Indochina region (in spring and summer), India (in summer and fall), Indochina, and north-northeastern China and southwestern Siberia (in summer) are diagnosed as the hotspots. In addition, the soil moisture’s impact on precipitation frequency over Asia is even weaker than its impact on precipitation amount.
  2. The conditional correlation analysis for RegCM4 output and the GLACE-type experiment using the same model produce similar spatial patterns of land–atmosphere coupling strength, with the signal identified by the conditional correlation method being slightly more extensive spatially. These hotspots include the Indian subcontinent in almost all seasons examined, southwestern Indochina region in spring, and part of northern China in summer. This statement is based on the soil moisture’s impact on temperature, evaporation, and EF. The signal for soil moisture impact on precipitation is weak and spatially scattered based on both methods.
  3. Both evaporation and evaporative fraction have a vital role in linking soil moisture with precipitation and temperature. The product indices [ and PV(υ) × σ(υ)] for both ET and EF can effectively identify regions of strong soil moisture impact on temperature. The only exception is found over the Indochina region, where EF appears to be more robust as the terrestrial segment of land–atmosphere coupling, and the EF product index better resembles the SM–T coupling strength signal relative to the ET product index.

The strong land–atmosphere coupling signal in the CFSR data represents an overestimation compared with the GLDAS data. While it is difficult to pinpoint what might have caused this overestimation without manipulative experiments using the CFSR modeling system, the larger magnitude of CFSR soil moisture standard deviation (compared with GLDAS) might play a role, as larger standard deviation reflects a larger range of soil moisture variation that is a necessary condition for strong land–atmosphere coupling.

While many statistical methods exist for quantifying the land–atmosphere coupling, the Mei and Wang (2011) conditional correlation method is chosen here for two reasons. First, the method is applicable to soil moisture–atmosphere coupling for which the intrinsic time scale is from a few days to weeks. In fact, the methodology of Mei and Wang does not make any assumption regarding the time scale at which the land–atmosphere coupling occurs, making it applicable to coupling processes that take place at any time scale. On the other hand, the feedback parameter approach (Liu et al. 2006; Sun and Wang 2012), for example, explicitly assumes a time scale of coupling much longer than 1 month, and the Zeng et al. (2010) Gamma index also applies at monthly or seasonal time scales. Second, by dividing the data into outer quartiles and inner quartiles, the Mei and Wang method can focus specifically on more extreme conditions. This allows for consideration for potential threshold behaviors of the soil moisture impact on precipitation (e.g., Kim and Wang 2007).

Conceptually, as is the case for any correlation analysis, the conditional correlation analysis method used here can potentially suffer from the presence of low-frequency variability in the climate record and autocorrelation of precipitation (e.g., Wei et al. 2008; Wei and Dirmeyer 2010). However, our method is less influenced by such factors because of the strict restriction imposed to the sampling procedure to exclude overlapping of precipitation time periods in the calculation of any individual correlation coefficient used to derive the probability distribution function. Such restriction also leads to different behaviors when compared with other published correlation analysis methods (e.g., Wei et al. 2008; Dirmeyer 2011). For example, applying Wei et al.’s (2008) correlation method to the GLDAS, CFSR, and RegCM4 output produced a result similar to Wei et al.’s, with negative autocorrelation of 21-day precipitation throughout the domain (results not shown); applying the Mei and Wang (2011) approach to the autocorrelation of 21-day precipitation produces no negative PC value, with a few scattered grid cells in the entirety of Asia seeing a significant and positive PC value (results not shown). Another factor that makes the Mei and Wang (2011) approach less susceptible to the impact of precipitation persistence might be the use of normalized anomalies (as opposed to raw data) in the data analysis. For example, Wei and Dirmeyer (2010) found that precipitation persistence (defined as the lag-2-pentad precipitation autocorrelation) can have some impact on the land–atmosphere coupling strength in the GLACE context. The so-defined persistence is significant in several regions when estimated based on raw precipitation and soil moisture data in RegCM4 output, GLDAS, and CFSR data. However, when calculated based on normalized anomalies using the Mei and Wang approach, this autocorrelation is negligible for all datasets examined in this study.

Based both on the conditional correlation method and the GLACE approach and on the RegCM4 and the two available datasets (GLDAS and CFSR), the soil moisture–precipitation coupling in the Asian domain is weak, despite rather strong impact of soil moisture on temperature in a large portion of Asia. While the conditional correlation between soil moisture and subsequent precipitation may be influenced by precipitation autocorrelation, this does not confirm a cause-and-effect relationship; the GLACE experimental design ensures an unambiguous cause-and-effect relationship between soil moisture and precipitation variability in regions of high GLACE index. The high degree of similarity between the conditional correlation analysis results and the GLACE-type experiment results using the same model suggests that the lack of strong soil moisture–precipitation coupling in Asia as found in RegCM4, GLDAS data, and CFSR data might be a true characteristic of this regional climate system. However, the approaches used in this study focus more on quantifying the coupling strength. More targeted or detailed numerical experiments are necessary in order to develop physical process–based understanding on how the results come to be (e.g., why and how the soil moisture’s coupling to temperature is much stronger than its coupling to precipitation).

Given the lack of true observational data for soil moisture at the spatial scale of interest, findings from this study are subject to major uncertainties related to data quality. In both GLDAS and CFSR, soil moisture data were produced by the Noah land model. In GLDAS, atmospheric variables driving the land model were downscaled from reanalysis data, and precipitation and temperature were bias corrected against observation-based data before they were used as inputs for the land surface model. Uncertainties in the land–atmosphere coupling strength based on the GLDAS data therefore mainly come from uncertainties related to the Noah land model’s capability in simulating land surface processes. In CFSR, atmospheric variables were derived from the data assimilation system and thus had input from both a model and observational data. Therefore, any model-related errors in atmospheric forcing will be propagated into the land surface processes to influence soil moisture, and the resulting error in soil moisture then further influences the model atmospheric processes. The land–atmosphere coupling strength estimated based on CFSR data is therefore subject to uncertainties related to both the atmospheric model and the land model and is also influenced by the coupling strength inherit to the CFSR model. Soil moisture observational data are critically needed to better quantify the true coupling strength between land surface and the atmosphere.

Acknowledgments

This study was made possible by a scholarship from the China Scholarship Council that supported Di Liu’s visit to the University of Connecticut. Contribution from the University of Connecticut was supported by the NOAA CPPA (NA08OAR4310871) and the NSF (AGS-1063986), and contribution from Hohai University was supported by the National Basic Research Program of China (2010CB951101, 2013CBA01806) and the National Natural Science Foundation of China (Grant 41101015). The GLDAS data used in this study were acquired as part of the mission of NASA’s Earth Science Division and archived and distributed by the Goddard Earth Sciences (GES) Data and Information Services Center (DISC). The authors thank the two anonymous reviewers for very thorough and constructive reviews on an earlier version of this paper.

REFERENCES

  • Atlas, R., , Wolfson N. , , and Terry J. , 1993: The effect of SST and soil moisture anomalies on GLA model simulations of the 1988 U.S. summer drought. J. Climate, 6, 20342048, doi:10.1175/1520-0442(1993)006<2034:TEOSAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bosilovich, M., , and Sun W. Y. , 1999: Numerical simulation of the 1993 Midwestern flood: Land–atmosphere interactions. J. Climate, 12, 14901505, doi:10.1175/1520-0442(1999)012<1490:NSOTMF>2.0.CO;2.

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

    • Search Google Scholar
    • Export Citation
  • Delworth, T., , and Manabe S. , 1989: The influence of soil wetness on near-surface atmospheric variability. J. Climate, 2, 14471462, doi:10.1175/1520-0442(1989)002<1447:TIOSWO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dirmeyer, P. A., 2011: The terrestrial segment of soil moisture-climate coupling. Geophys. Res. Lett., 38, L16702, doi:10.1029/2011GL048268.

    • Search Google Scholar
    • Export Citation
  • Dirmeyer, P. A., , Koster R. D. , , and Guo Z. , 2006: Do global models properly represent the feedback between land and atmosphere? J. Hydrometeor., 7, 11771198, doi:10.1175/JHM532.1.

    • Search Google Scholar
    • Export Citation
  • D’Odorico, P., , and Porporato A. , 2004: Preferential states in soil moisture and climate dynamics. Proc. Natl. Acad. Sci. USA, 101, 88488851, doi:10.1073/pnas.0401428101.

    • Search Google Scholar
    • Export Citation
  • Douville, H., , Chauvin F. , , and Broqua H. , 2001: Influence of soil moisture on the Asian and African monsoons. Part I: Mean monsoon and daily precipitation. J. Climate, 14, 23812403, doi:10.1175/1520-0442(2001)014<2381:IOSMOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ek, M. B., , and Holtslag A. A. M. , 2004: Influence of soil moisture on boundary layer cloud development. J. Hydrometeor., 5, 8699, doi:10.1175/1525-7541(2004)005<0086:IOSMOB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Elguindi, N., and Coauthors, 2011: Regional climatic model RegCM user manual version 4.1. The Abdus Salam International Centre for Theoretical Physics, Trieste, Italy, 32 pp. [Available online at http://gforge.ictp.it/gf/download/docmanfileversion/21/544/ReferenceMan4.1.pdf.]

  • Findell, K. L., , and Eltahir E. A. B. , 1997: An analysis of the soil moisture-precipitation feedback, based on direct observations from Illinois. Water Resour. Res., 33, 725735, doi:10.1029/96WR03756.

    • Search Google Scholar
    • Export Citation
  • Findell, K. L., , and Eltahir E. A. B. , 2003: Atmospheric controls on soil moisture–boundary layer interactions. Part II: Feedbacks within the continental United States. J. Hydrometeor., 4, 570583, doi:10.1175/1525-7541(2003)004<0570:ACOSML>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Findell, K. L., , Gentine P. , , Linter B. R. , , and Kerr C. , 2011: Probability of afternoon precipitation in eastern United States and Mexico enhanced by high evaporation. Nat. Geosci., 4, 434–439, doi:10.1038/ngeo1174.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., , Mearns L. O. , , Shields C. , , and Mayer L. , 1996: A regional model study of the importance of local versus remote controls of the 1988 drought and the 1993 flood over the central United States. J. Climate, 9, 11501162, doi:10.1175/1520-0442(1996)009<1150:ARMSOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., and Coauthors, 2012: RegCM4: Model description and illustrative basic performance over selected CORDEX domains. Climate Res., 52, 729, doi:10.3354/cr01018.

    • Search Google Scholar
    • Export Citation
  • Gu, H. H., , Wang G. L. , , Yu Z. B. , , and Mei R. , 2012: Assessment on impact of climate change on East and South Asia using the RegCM4 regional climate model. Climatic Change, 114, 301317, doi:10.1007/s10584-012-0411-y.

    • Search Google Scholar
    • Export Citation
  • Guo, Z., and Coauthors, 2006: GLACE: The Global Land–Atmosphere Coupling Experiment. Part II: Analysis. J. Hydrometeor., 7, 611625, doi:10.1175/JHM511.1.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., , and Pan H.-L. , 2000: Impact of soil moisture anomalies on seasonal, summertime circulation over North America in a regional climate model. J. Geophys. Res., 105, 29 62529 634, doi:10.1029/2000JD900276.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kanamitsu, M., , Ebisuzaki W. , , Woollen J. , , Yang S.-K. , , Hnilo J. J. , , Fiorino M. , , and Potter G. L. , 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
  • Kim, J. E., , and Hong S. Y. , 2007: Impact of soil moisture anomalies on summer rainfall over East Asia: A regional climate model study. J. Climate, 20, 57325743, doi:10.1175/2006JCLI1358.1.

    • Search Google Scholar
    • Export Citation
  • Kim, Y. J., , and Wang G. L. , 2007: Impact of initial soil moisture anomalies on subsequent precipitation over North America in the Coupled Land–Atmosphere Model CAM3–CLM3. J. Hydrometeor., 8, 513533, doi:10.1175/JHM611.1.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., , and Suarez M. J. , 1995: Relative contributions of land and ocean processes to precipitation variability. J. Geophys. Res.,100 (D7), 13 775–13 790, doi:10.1029/95JD00176.

  • Koster, R. D., , Dirmeyer P. A. , , Hahmann A. N. , , Ijpelaar R. , , Tyahla L. , , Cox P. , , and Suarez M. J. , 2002: Comparing the degree of land–atmosphere interaction in four atmospheric general circulation models. J. Hydrometeor., 3, 363375, doi:10.1175/1525-7541(2002)003<0363:CTDOLA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Coauthors, 2004a: Realistic initialization of land surface states: Impacts on subseasonal forecast skill. J. Hydrometeor., 5, 10491063, doi:10.1175/JHM-387.1.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Coauthors, 2004b: Regions of strong coupling between soil moisture and precipitation. Science, 305, 11381140, doi:10.1126/science.1100217.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., and Coauthors, 2006: GLACE: The Global Land–Atmosphere Coupling Experiment. Part I: Overview. J. Hydrometeor., 7, 590610, doi:10.1175/JHM510.1.

    • Search Google Scholar
    • Export Citation
  • Koster, R. D., , Schubert S. D. , , and Suarez M. J. , 2009: Analyzing the concurrence of meteorological droughts and warm periods, with implications for the determination of evaporative regime. J. Climate, 22, 33313341, doi:10.1175/2008JCLI2718.1.

    • Search Google Scholar
    • Export Citation
  • Kumar, S. V., and Coauthors, 2006: Land information system: An interoperable framework for high resolution land surface modeling. Environ. Modell. Software, 21, 14021415, doi:10.1016/j.envsoft.2005.07.004.

    • Search Google Scholar
    • Export Citation
  • Leung, L. R., , Huang M. , , Qian Y. , , and Liang X. , 2011: Climate-soil-vegetation control on groundwater table dynamics and its feedbacks in a climate model. Climate Dyn., 36, 5781, doi:10.1007/s00382-010-0746-x.

    • Search Google Scholar
    • Export Citation
  • Liu, Z. Y., , Notaro M. , , and Kutzbach J. , 2006: Assessing global vegetation–climate feedbacks from observations. J. Climate, 19, 787814, doi:10.1175/JCLI3658.1.

    • Search Google Scholar
    • Export Citation
  • Mei, R., , and Wang G. , 2011: Impact of sea surface temperature and soil moisture on summer precipitation in the United States based on observational data. J. Hydrometeor., 12, 10861099, doi:10.1175/2011JHM1312.1.

    • Search Google Scholar
    • Export Citation
  • Mei, R., , and Wang G. , 2012: Summer land–atmosphere coupling strength in the United States: Comparison among observations, reanalysis data and numerical models. J. Hydrometeor., 13, 10101022, doi:10.1175/JHM-D-11-075.1.

    • Search Google Scholar
    • Export Citation
  • Namias, J., 1952: The annual course of month-to-month persistence in climatic anomalies. Bull. Amer. Meteor. Soc., 33, 279285.

  • Namias, J., 1959: Persistence of mid-tropospheric circulations between adjacent months and seasons. The Atmosphere and the Sea in Motion: Scientific Contributions to the Rossby Memorial Volume, Rockefeller Institute Press, 240–248.

  • Rayner, N. A., , Parker D. E. , , Horton E. B. , , Folland C. K. , , Alexander L. V. , , Rowell D. P. , , Kent E. C. , , and Kaplan A. , cited 2006: Met Office’s Global Ice Coverage and Sea Surface Temperatures (GISST), and Monthly Night Marine Air Temperature/SST Anomalies (MOHMATN4)/(MOHSST6) data (1856–2006). NCAS British Atmospheric Data Centre. [Available online at http://badc.nerc.ac.uk/view/badc.nerc.ac.uk__ATOM__dataent_GISST.]

  • Rodell, M., and Coauthors, 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
  • Ruiz-Barradas, A., , and Nigam S. , 2005: Warm season precipitation variability over the U.S. Great Plains in observations, NCEP and ERA-40 reanalyses, and NCAR and NASA atmospheric model simulations. J. Climate, 18, 18081830, doi:10.1175/JCLI3343.1.

    • Search Google Scholar
    • Export Citation
  • Ruiz-Barradas, A., , and Nigam S. , 2006: Great plains hydroclimate variability: The view from North American Regional Reanalysis. J. Climate, 19, 30043010, doi:10.1175/JCLI3768.1.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 10151057, doi:10.1175/2010BAMS3001.1.

    • Search Google Scholar
    • Export Citation
  • Salvucci, G. D., , Saleem J. A. , , and Kaufmann R. , 2002: Investigating soil moisture feedbacks on precipitation with tests of Granger causality. Adv. Water Resour., 25, 13051312, doi:10.1016/S0309-1708(02)00057-X.

    • Search Google Scholar
    • Export Citation
  • Schär, C., , Lüthi D. , , Beyerle U. , , and Heise E. , 1999: The soil–precipitation feedback: A process study with a regional climate model. J. Climate, 12, 722741, doi:10.1175/1520-0442(1999)012<0722:TSPFAP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Schlosser, C. A., , and Milly P. C. D. , 2002: A model-based investigation of soil moisture predictability and associated climate predictability. J. Hydrometeor., 3, 483501, doi:10.1175/1525-7541(2002)003<0483:AMBIOS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Seneviratne, S. I., , Lüthi D. , , Litschi M. , , and Schär C. , 2006: Land–atmosphere coupling and climate change in Europe. Nature, 443, 205209, doi:10.1038/nature05095.

    • Search Google Scholar
    • Export Citation
  • Seth, A., , and Giorgi F. , 1998: The effects of domain choice on summer precipitation simulation and sensitivity in a regional climate model. J. Climate, 11, 26982712, doi:10.1175/1520-0442(1998)011<2698:TEODCO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Sheffield, J., , Goteti G. , , and Wood E. F. , 2006: Development of a 50-yr high-resolution global dataset of meteorological forcings for land surface modeling. J. Climate, 19, 30883111, doi:10.1175/JCLI3790.1.

    • Search Google Scholar
    • Export Citation
  • Shukla, J., , and Mintz Y. , 1982: Influence of land–surface evapotranspiration on the earth’s climate. Science, 215, 14981501, doi:10.1126/science.215.4539.1498.

    • Search Google Scholar
    • Export Citation
  • Siqueira, M., , Katul G. , , and Porporato A. , 2009: Soil moisture feedbacks on convection triggers: The role of soil–plant hydrodynamics. J. Hydrometeor., 10, 96112, doi:10.1175/2008JHM1027.1.

    • Search Google Scholar
    • Export Citation
  • Steiner, A. L., , Pal J. S. , , Rauscher S. A. , , Belel J. L. , , Diffenbaugh N. S. , , Boone A. , , Sloan L. C. , , and Giorgi F. , 2009: Land surface coupling in regional climate simulations of the West African monsoon. Climate Dyn., 33, 869892, doi:10.1007/s00382-009-0543-6.

    • Search Google Scholar
    • Export Citation
  • Sud, Y. C., , Mocko D. M. , , Lau K.-M. , , and Atlas R. , 2003: Simulating the midwestern U.S. drought of 1988 with a GCM. J. Climate, 16, 39463965, doi:10.1175/1520-0442(2003)016<3946:STMUDO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Sun, S. S., , and Wang G. L. , 2012: The complexity of using a feedback parameter to quantify the soil moisture-precipitation relationship. J. Geophys. Res., 117, D11113, doi:10.1029/2011JD017173.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , and Guillemot C. J. , 1996: Physical processes involved in the 1988 drought and 1993 floods in North America. J. Climate, 9, 12881298, doi:10.1175/1520-0442(1996)009<1288:PPIITD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wei, J., , and Dirmeyer P. A. , 2010: Toward understanding the large-scale land-atmosphere coupling in the models: Roles of different processes. Geophys. Res. Lett., 37, L19707, doi:10.1029/2010GL044769.

    • Search Google Scholar
    • Export Citation
  • Wei, J., , and Dirmeyer P. A. , 2012: Dissecting soil moisture-precipitation coupling. Geophys. Res. Lett., 39, L19711, doi:10.1029/2012GL053038.

    • Search Google Scholar
    • Export Citation
  • Wei, J., , Dickinson R. E. , , and Chen H. , 2008: A negative soil moisture–precipitation relationship and its causes. J. Hydrometeor., 9, 13641376, doi:10.1175/2008JHM955.1.

    • Search Google Scholar
    • Export Citation
  • Yeh, T. C., , Wetherald R. T. , , and Manabe S. , 1984: The effect of soil moisture on the short-term climate and hydrology change—A numerical experiment. Mon. Wea. Rev., 112, 474490, doi:10.1175/1520-0493(1984)112<0474:TEOSMO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Zeng, X. B., , Barlage M. , , Castro C. , , and Fling K. , 2010: Comparison of land–precipitation coupling strength using observations and models. J. Hydrometeor., 11, 979994, doi:10.1175/2010JHM1226.1.

    • Search Google Scholar
    • Export Citation
  • Zhang, J., , Wang W. C. , , and Leung L. R. , 2008a: Contribution of land–atmosphere coupling to summer climate variability over the contiguous United States. J. Geophys. Res., 113, D22109, doi:10.1029/2008JD010136.

    • Search Google Scholar
    • Export Citation
  • Zhang, J., , Wang W. C. , , and Wei J. , 2008b: Assessing land–atmosphere coupling using soil moisture from the Global Land Data Assimilation System and observational precipitation. J. Geophys. Res., 113, D17119, doi:10.1029/2008JD009807.

    • Search Google Scholar
    • Export Citation
  • Zhang, J., , Wu L. , , and Dong W. , 2011: Land–atmosphere coupling and summer climate variability over East Asia. J. Geophys. Res., 116, D05117, doi:10.1029/2010JD014714.

    • Search Google Scholar
    • Export Citation
  • Zheng, X. Y., , and Eltahir E. A. B. , 1998: A soil moisture–rainfall feedback mechanism: 2. Numerical experiments. Water Resour. Res., 34, 777785, doi:10.1029/97WR03497.

    • Search Google Scholar
    • Export Citation
Save