• Alpert, J. C., Kanamitsu M. , Caplan P. M. , Sela J. G. , White G. H. , and Kalnay E. , 1988: Mountain induced gravity wave drag parameterization in the NMC Medium-Range Forecast Model. Preprints, Eighth Conf. on Numerical Weather Prediction, Baltimore, MD, Amer. Meteor. Soc., 726–733.

  • Alpert, J. C., Hong S.-Y. , and Kim Y.-J. , 1996: Sensitivity of cyclogenesis to lower tropospheric enhancement of gravity wave drag using the Environmental Modeling Center Medium Range Model. Preprints, 11th Conf. on Numerical Weather Prediction, Norfolk, VA, Amer. Meteor. Soc., 322–323.

  • Arakawa, A., and Schubert W. H. , 1974: Interaction of a cumulus cloud ensemble with the large-scale environment. Part I. J. Atmos. Sci., 31, 674701, doi:10.1175/1520-0469(1974)031<0674:IOACCE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Balling, R. C., 1985: Warm season nocturnal precipitation in the Great Plains of the United States. J. Appl. Meteor. Climatol., 24, 13831387, doi:10.1175/1520-0450(1985)024<1383:WSNPIT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Basu, B. K., 2007: Diurnal variation in precipitation over India during the summer monsoon season: Observed and model predicted. Mon. Wea. Rev., 135, 21552167, doi:10.1175/MWR3355.1.

    • Search Google Scholar
    • Export Citation
  • Betts, A., Ball J. H. , Beljaars A. C. M. , Miller M. J. , and Viterbo P. A. , 1996: The land surface–atmosphere interaction: A review based on observational and global modeling perspectives. J. Geophys. Res., 101, 72097225, doi:10.1029/95JD02135.

    • Search Google Scholar
    • Export Citation
  • Boone, A., and Coauthors, 2009: AMMA Land Surface Model Intercomparison Project Phase 2 (ALMIP2). GEWEX News, Vol. 19, No. 4, International GEWEX Project Office, Silver Spring, MD, 9–10. [Available online at http://www.gewex.org/images/Nov2009.pdf.]

  • Boyd, E., Cornforth R. J. , Lamb P. J. , Tarhule A. , Lele M. I. , and Brouder A. , 2013: Building resilience to face recurring environmental crisis in African Sahel. Nat. Climate Change, 3, 631637, doi:10.1038/nclimate1856.

    • Search Google Scholar
    • Export Citation
  • Boyle, J., and Klein S. A. , 2010: Impact of horizontal resolution on climate model forecasts of tropical precipitation and diabatic heating for the TWP-ICE period. J. Geophys. Res., 115, D23113, doi:10.1029/2010JD014262.

    • Search Google Scholar
    • Export Citation
  • Brier, G. W., 1965: Diurnal and semidiurnal atmospheric tides in relation to precipitation variations. Mon. Wea. Rev., 93, 93100, doi:10.1175/1520-0493(1965)093<0093:DASATI>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • Cappelaere, B., and Coauthors, 2009: The AMMA-CATCH experiment in the cultivated Sahelian area of south-west Niger—Investigating water cycle response to a fluctuating climate and changing environment. J. Hydrol., 375, 3451, doi:10.1016/j.jhydrol.2009.06.021.

    • Search Google Scholar
    • Export Citation
  • Chen, H., Yu R. , Li J. , Yuan W. , and Zhou T. , 2010: Why nocturnal long-duration rainfall presents an eastward delayed diurnal phase along the Yangtze River. J. Climate, 23, 905917, doi:10.1175/2009JCLI3187.1.

    • Search Google Scholar
    • Export Citation
  • Chen, T. C., Yen M. C. , Hsieh J. C. , and Arritt R. W. , 1999: Diurnal and seasonal variations of the rainfall measured by the automatic rainfall and meteorological telemetry system in Taiwan. Bull. Amer. Meteor. Soc., 80, 22992312, doi:10.1175/1520-0477(1999)080<2299:DASVOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Chou, M.-D., and Suarez M. J. , 1994: An efficient thermal infrared radiation parameterization for use in general circulation models, NASA Tech. Memo. TM-1994-104606, Series on Global Modeling and Data Assimilation, Vol. 3, 102 pp.

  • Clark, A. J., Gallus W. A. Jr., and Chen T. C. , 2007: Comparison of the diurnal precipitation cycle in convection-resolving and non-convection-resolving mesoscale models. Mon. Wea. Rev., 135, 34563473, doi:10.1175/MWR3467.1.

    • Search Google Scholar
    • Export Citation
  • Couvreux, F., Rio C. , Guichard F. , Lothon M. , Canut G. , Bouniol D. , and Gounou A. , 2012: Initiation of daytime local convection in a semi-arid region analysed with high-resolution simulations and AMMA observations. Quart. J. Roy. Meteor. Soc., 138, 5671, doi:10.1002/qj.903.

    • Search Google Scholar
    • Export Citation
  • Dai, A., 2001: Global precipitation and thunderstorm frequencies. Part II: Diurnal variations. J. Climate, 14, 11121128, doi:10.1175/1520-0442(2001)014<1112:GPATFP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dai, A., and Trenberth K. E. , 2004: The diurnal cycle and its depiction in the Community Climate System Model. J. Climate, 17, 930951, doi:10.1175/1520-0442(2004)017<0930:TDCAID>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dai, A., Giorgi F. , and Trenberth K. E. , 1999: Observed and model simulated diurnal cycles of precipitation over the contiguous United States. J. Geophys. Res., 104, 63776402, doi:10.1029/98JD02720.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., Manning K. W. , Carbone R. E. , Trier S. B. , and Tuttle J. D. , 2003: Coherence of warm-season continental rainfall in numerical weather prediction models. Mon. Wea. Rev., 131, 26672679, doi:10.1175/1520-0493(2003)131<2667:COWCRI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • 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 advancements 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
  • Fritsch, J. M., and Carbone R. E. , 2004: Improving quantitative precipitation forecasts in the warm season: A USWRP research and development strategy. Bull. Amer. Meteor. Soc., 85, 955965, doi:10.1175/BAMS-85-7-955.

    • Search Google Scholar
    • Export Citation
  • Gallus, W. A., Jr., 1999: Eta simulations of three extreme precipitation events: Sensitivity to resolution and convective parameterization. Wea. Forecasting, 14, 405426, doi:10.1175/1520-0434(1999)014<0405:ESOTEP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Garratt, J. R., Krummel P. , and Kowalczyk E. A. , 1993: The surface-energy balance at local and regional scales—A comparison of general circulation model results with observations. J. Climate, 6, 10901109, doi:10.1175/1520-0442(1993)006<1090:TSEBAL>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., and Shields C. , 1999: Tests of precipitation parameterizations available in latest version of NCAR Regional Climate Model (RegCM) over continental United States. J. Geophys. Res., 104, 63536376, doi:10.1029/98JD01164.

    • Search Google Scholar
    • Export Citation
  • Gounou, A., 2011: The driving processes behind the diurnal cycles of the West African monsoon. Ph.D. thesis, Universite Paul Sabatier, Toulouse III, Toulouse, France, 200 pp. [Available online at http://francoise.guichard.m.free.fr/fg2cd/these_publiee_AG.pdf.]

  • Gray, W. M., and Jacobson R. W. , 1977: Diurnal variation of deep cumulus convection. Mon. Wea. Rev., 105, 11711188, doi:10.1175/1520-0493(1977)105<1171:DVODCC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hannay, C., Neale R. , and Bacmeister J. , 2012: Timeslice experiments at high resolution. What does the resolution buy us? CESM Atmosphere Model Working Group Meeting, Boulder, NCAR. [Available online at http://www.cesm.ucar.edu/working_groups/Atmosphere/Presentations/2012/hannay2.pdf.]

  • Haurwitz, B., and Cowley A. D. , 1973: The diurnal and semidiurnal barometric oscillations, global distribution, and annual variation. Pure Appl. Geophys., 102, 193222, doi:10.1007/BF00876607.

    • Search Google Scholar
    • Export Citation
  • Hodges, K. I., and Thorncroft C. C. , 1997: Distribution and statistics of African mesoscale convective weather systems based on the ISCCP Meteosat imagery. Mon. Wea. Rev., 125, 28212837, doi:10.1175/1520-0493(1997)125<2821:DASOAM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., and Pan H. L. , 1996: Nonlocal boundary layer vertical diffusion in a Medium-Range Forecast Model. Mon. Wea. Rev., 124, 23222339, doi:10.1175/1520-0493(1996)124<2322:NBLVDI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., and Leetmaa A. , 1999: An evaluation of the NCEP RSM for regional climate modeling. J. Climate, 12, 592609, doi:10.1175/1520-0442(1999)012<0592:AEOTNR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., Yign N. , and Dudhia J. , 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 23182341, doi:10.1175/MWR3199.1.

    • Search Google Scholar
    • Export Citation
  • Huang, W.-R., and Chan J. C. L. , 2011: Maintenance mechanisms for the early-morning maximum summer rainfall over southeast China. Quart. J. Roy. Meteor. Soc., 137, 959968, doi:10.1002/qj.815.

    • Search Google Scholar
    • Export Citation
  • Huffman, G. J., and Coauthors, 2007: The TRMM Multisatellite Precipitation Analysis (TMPA): Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J. Hydrometeor., 8, 3855, doi:10.1175/JHM560.1.

    • Search Google Scholar
    • Export Citation
  • Juang, H.-M., and Kanamitsu M. , 1994: The NMC nested regional spectral model. Mon. Wea. Rev., 122, 326, doi:10.1175/1520-0493(1994)122<0003:TNNRSM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Juang, H.-M., Hong S.-Y. , and Kanamitsu M. , 1997: The NCEP Regional Spectral Model: An update. Bull. Amer. Meteor. Soc., 78, 21252143, doi:10.1175/1520-0477(1997)078<2125:TNRSMA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kain, J. S., 2004: The Kain–Fritsch convective parameterization: An update. J. Appl. Meteor. Climatol., 43, 170181, doi:10.1175/1520-0450(2004)043<0170:TKCPAU>2.0.CO;2.

    • 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
  • Kang, H.-S., and Hong S.-Y. , 2008: Sensitivity of the simulated East Asian summer monsoon climatology to four convective parameterization schemes. J. Geophys. Res., 113, D15119, doi:10.1029/2007JD009692.

    • Search Google Scholar
    • Export Citation
  • Kim, J., and Mahrt L. , 1992: Momentum transport by gravity waves. J. Atmos. Sci., 49, 735748, doi:10.1175/1520-0469(1992)049<0735:MTBGW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Koo, M.-S., and Hong S.-Y. , 2010: Diurnal variations of simulated precipitation over East Asia in two regional climate models. J. Geophys. Res., 115, D05105, doi:10.1029/2009JD012574.

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

    • Search Google Scholar
    • Export Citation
  • Laing, A. G., Carbone R. , Levizzani V. , and Tuttle J. , 2008: The propagation and diurnal cycles of deep convection in northern tropical Africa. Quart. J. Roy. Meteor. Soc., 134, 93109, doi:10.1002/qj.194.

    • Search Google Scholar
    • Export Citation
  • Lebel, T., and Coauthors, 2009: AMMA-CATCH studies in the Sahelian region of West-Africa: An overview. J. Hydrol., 375, 313, doi:10.1016/j.jhydrol.2009.03.020.

    • Search Google Scholar
    • Export Citation
  • Lee, M., and Coauthors, 2007: Sensitivity to horizontal resolution in the AGCM simulations of warm season diurnal cycle of precipitation over the United States and northern Mexico. J. Climate, 20, 18621881, doi:10.1175/JCLI4090.1.

    • Search Google Scholar
    • Export Citation
  • Li, Z., Yang D. , Hong Y. , Zhang J. , and Qi Y. , 2014: Characterizing spatiotemporal variations of hourly rainfall by gauge and radar in the mountainous Three Gorges region. J. Appl. Meteor. Climatol., 53, 873889, doi:10.1175/JAMC-D-13-0277.1.

    • Search Google Scholar
    • Export Citation
  • Lin, X., Randall D. A. , and Fowler L. D. , 2000: Diurnal variability of the hydrological cycle and radiative fluxes: Comparisons between observations and a GCM. J. Climate, 13, 41594179, doi:10.1175/1520-0442(2000)013<4159:DVOTHC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lin, Y., and Coauthors, 2012: TWP-ICE global atmospheric model intercomparison: Convection responsiveness and resolution impact. J. Geophys. Res.,117, D09111, doi:10.1029/2011JD017018.

  • McGarry, M. M., and Reed R. J. , 1978: Diurnal variations in convective activity and precipitation during phases II and III of GATE. Mon. Wea. Rev., 106, 101113, doi:10.1175/1520-0493(1978)106<0101:DVICAA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Mohr, K. I., 2004: Interannual, monthly, and regional variability in the wet season diurnal cycle of precipitation in sub-Saharan Africa. J. Climate, 17, 24412453, doi:10.1175/1520-0442(2004)017<2441:IMARVI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Moorthi, S., and Suarez M. J. , 1992: Relaxed Arakawa–Schubert: A parameterization of moist convection for general circulation models. Mon. Wea. Rev., 120, 9781002, doi:10.1175/1520-0493(1992)120<0978:RASAPO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Nikulin, G., and Coauthors, 2012: Precipitation climatology in an ensemble of CORDEX-Africa regional climate simulations. J. Climate, 25, 60576078, doi:10.1175/JCLI-D-11-00375.1.

    • Search Google Scholar
    • Export Citation
  • Oki, T., and Kanae S. , 2006: Global hydrological cycles and world water resources. Science, 313, 10681072, doi:10.1126/science.1128845.

    • Search Google Scholar
    • Export Citation
  • Pan, H.-L., and Wu W.-S. , 1995: Implementing a mass flux convective parameterization package for the NMC Medium-Range Forecast Model. NMC Office Note 409, Camp Springs, MD, 40 pp. [Available online at http://www.lib.ncep.noaa.gov/ncepofficenotes/files/01408A42.pdf.]

  • Parker, D. J., and Coauthors, 2005: The diurnal cycle of West African monsoon circulation. Quart. J. Roy. Meteor. Soc., 131, 28392860, doi:10.1256/qj.04.52.

    • Search Google Scholar
    • Export Citation
  • Pu, B., and Cook K. H. , 2010: Dynamics of the West African westerly jet. J. Climate, 23, 62636276, doi:10.1175/2010JCLI3648.1.

  • Ramage, C. S., 1952: Diurnal variation of summer rainfall over east China, Korea and Japan. J. Atmos. Sci., 9, 8386, doi:10.1175/1520-0469(1952)009<0083:DVOSRO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ramage, C. S., 1971: Monsoon Meteorology. Academic Press, 295 pp.

  • Randall, D. A., Harshvardhan, and Dazlich D. A. , 1991: Diurnal variability of the hydrological cycle in a general circulation model. J. Atmos. Sci., 48, 4062, doi:10.1175/1520-0469(1991)048<0040:DVOTHC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Randall, D. A., Khairoutdinov M. , and Grabowski W. , 2003: Breaking the cloud parameterization deadlock. Bull. Amer. Meteor. Soc., 84, 15471564, doi:10.1175/BAMS-84-11-1547.

    • Search Google Scholar
    • Export Citation
  • Ridout, J. A., Jin Y. , and Liou C.-S. , 2005: A cloud-base quasi-balance constraint for parameterized convection: Application to the Kain–Fritsch cumulus scheme. Mon. Wea. Rev., 133, 33153334, doi:10.1175/MWR3034.1.

    • Search Google Scholar
    • Export Citation
  • Rio, C., Hourdin F. , Grandpeix J.-Y. , and Lafore J.-P. , 2009: Shifting the diurnal cycle of parameterized deep convection over land. Geophys. Res. Lett., 36, L07809, doi:10.1029/2008GL036779.

    • Search Google Scholar
    • Export Citation
  • Roca, R., Chambon P. , Jobard I. , Kirstetter P. , Gosset M. , and Berges J. C. , 2010: Comparing satellite and surface rainfall products over West Africa at meteorologically relevant scales during the AMMA campaign using error estimates. J. Appl. Meteor. Climatol., 49, 715731, doi:10.1175/2009JAMC2318.1.

    • 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
  • Schwartz, B. E., and Bosart L. F. , 1979: The diurnal variability of Florida rainfall. Mon. Wea. Rev., 107, 15351545, doi:10.1175/1520-0493(1979)107<1535:TDVOFR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Shine, D. W., Cocke S. , and LaRow T. E. , 2007: Diurnal cycle of precipitation in a climate model. J. Geophys. Res., 112, D13109, doi:10.1029/2006JD008333.

    • Search Google Scholar
    • Export Citation
  • Slingo, A., Wilderspin R. , and Brentnall S. , 1987: Simulation of the diurnal cycle of outgoing longwave radiation with an atmospheric GCM. Mon. Wea. Rev., 115, 14511457, doi:10.1175/1520-0493(1987)115<1451:SOTDCO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Sperber, K. R., and Yasunari T. , 2006: Workshop on monsoon climate systems: Toward better prediction of the monsoon. Bull. Amer. Meteor. Soc., 87, 13391403, doi:10.1175/BAMS-87-10-1399.

    • Search Google Scholar
    • Export Citation
  • Taylor, C. M., and Clark D. B. , 2001: The diurnal cycle and African easterly waves: A land surface perspective. Quart. J. Roy. Meteor. Soc., 127, 845867, doi:10.1002/qj.49712757308.

    • Search Google Scholar
    • Export Citation
  • Taylor, C. M., Gounou A. , Guichard F. , Harris P. P. , Ellis R. J. , Couvreux F. , and Kauwe M. D. , 2011: Frequency of Sahelian storm initiation enhanced over mesoscale soil-moisture patterns. Nat. Geosci., 4, 430433, doi:10.1038/ngeo1173.

    • Search Google Scholar
    • Export Citation
  • Taylor, C. M., Birch C. E. , Parker D. J. , Dixon N. , Guichard F. , Nikulin G. , and Lister G. M. S. , 2013: Modeling soil moisture–precipitation feedback in the Sahel: Importance of spatial scale versus convective parameterization. Geophys. Res. Lett., 40, 62136218, doi:10.1002/2013GL058511.

    • Search Google Scholar
    • Export Citation
  • Taylor, K. E., 2001: Summarizing multiple aspects of model performance in a single diagram. J. Geophys. Res., 106, 71837192, doi:10.1029/2000JD900719.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., Dai A. , Rasmussen R. M. , and Parsons D. B. P. , 2003: The changing character of precipitation. Bull. Amer. Meteor. Soc., 84, 12051217, doi:10.1175/BAMS-84-9-1205.

    • Search Google Scholar
    • Export Citation
  • Vischel, T., Quantin G. , Lebel T. , Viarre J. , Gosset M. , Cazenave F. , and Panthou G. , 2011: Generation of high-resolution rain fields in West Africa: Evaluation of dynamic interpolation methods. J. Hydrometeor., 12, 14651482, doi:10.1175/JHM-D-10-05015.1.

    • Search Google Scholar
    • Export Citation
  • Wai, M. M.-K., Welsh P. T. , and Ma W.-M. , 1996: Interaction of secondary circulations with the summer monsoon and diurnal rainfall over Hong Kong. Bound.-Layer Meteor., 81, 123146, doi:10.1007/BF00119062.

    • Search Google Scholar
    • Export Citation
  • Wallace, J. M., 1975: Diurnal variations in precipitation and thunderstorm frequency over the conterminous United States. Mon. Wea. Rev., 103, 406419, doi:10.1175/1520-0493(1975)103<0406:DVIPAT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wang, W., and Seaman N. L. , 1997: A comparison study of convective parameterization schemes in a mesoscale model. Mon. Wea. Rev., 125, 252278, doi:10.1175/1520-0493(1997)125<0252:ACSOCP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wilks, D. S., 2005: Statistical Methods in the Atmospheric Sciences. 2nd ed. Academic Press, 627 pp.

  • Yang, G., and Slingo J. , 2001: The diurnal cycle in the tropics. Mon. Wea. Rev., 129, 784801, doi:10.1175/1520-0493(2001)129<0784:TDCITT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Yoshimura, K., and Kanamitsu M. , 2008: Dynamical global downscaling of global reanalysis. Mon. Wea. Rev., 136, 29832998, doi:10.1175/2008MWR2281.1.

    • Search Google Scholar
    • Export Citation
  • Yoshimura, K., Kanamitsu M. , and Dettinger M. , 2010: Regional downscaling for stable water isotopes: A case study of an atmospheric river event. J. Geophys. Res.,115, D18114, doi:10.1029/2010JD014032.

  • Yu, R., Zhou T. , Xiong A. , Zhu Y. , and Li J. , 2007: Diurnal variations of summer precipitation over contiguous China. Geophys. Res. Lett., 34, L01704, doi:10.1029/2006GL028129.

    • Search Google Scholar
    • Export Citation
  • Yuan, W., Wu R. , Zhang M. , Lin W. , Li J. , and Fu Y. , 2013: Diurnal cycle of summer precipitation over subtropical East Asia in CAM5. J. Climate, 26, 31593172, doi:10.1175/JCLI-D-12-00119.1.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Regional model domain (excluding buffer zone). The solid black box is for high-resolution (10 km) simulations. The Benin mesoscale site is outlined as a gray box.

  • View in gallery

    Normalized amplitudes of precipitation diurnal variations derived from (a) TMPA, (b) RAS, (c) SAS, and (d) KF2 during the monsoon season (April–September).

  • View in gallery

    Phase (peak timing) of precipitation diurnal variations derived from (a) TMPA, (b) RAS, (c) SAS, and (d) KF2 during the monsoon season (April–September). LST is indicated by the color bar with a 3-h interval.

  • View in gallery

    Hovmöller diagram (time–lon distribution) for the normalized (by daily mean) rainfall diurnal variation derived from (a) observations (TMPA) and (b)–(d) model simulations with different CPSs (RAS, SAS, and KF2). The y axis indicates LST.

  • View in gallery

    (a) Normalized diurnal variations derived from satellite and RSM output (dashed lines) and fitted models from the first two harmonics (continuous lines), and (b) fitted diurnal and (c) fitted semidiurnal cycles of precipitation amount from TMPA, RAS, SAS, and KF2 over land.

  • View in gallery

    As in Fig. 5, but for the ocean.

  • View in gallery

    Seasonal (April–September) evolution of the diurnal cycle for total precipitation averaged over the Benin mesoscale domain derived from ALMIP2, RAS, SAS, and KF2 at (a) coarse and (b) high resolution. Continuous lines are the first two harmonics fitted from dashed lines.

  • View in gallery

    Seasonal (April–September) evolution of the diurnal cycle for CAPE (above) and convective precipitation (below) averaged over the Benin mesoscale domain derived from CFSR, RAS, SAS, and KF2 at (a) coarse and (b) high resolution. For convective precipitation, continuous lines are the first two harmonics fitted from dashed lines. The color bar (center) indicates shortwave downward radiation (W m−2) from ALMIP2.

  • View in gallery

    Seasonal (April–September) evolution of the diurnal cycle for large-scale precipitation averaged over the Benin mesoscale domain derived from ALMIP2, RAS, SAS, and KF2 at (a) coarse and (b) high resolution. Continuous lines are the first two harmonics fitted from dashed lines.

  • View in gallery

    Taylor diagrams summarizing the statistics of precipitation amount from the observations and each sensitivity experiment.

  • View in gallery

    Vertical profiles of the anomalies from the ensemble mean in (top) specific humidity (kg kg−1) and (middle) temperature (K). (bottom) Vertical profiles of omega (Pa s−1). Negative (positive) values mean rising (subsiding) motion.

  • View in gallery

    Seasonal (April–September) evolution of the diurnal cycle for (a) SHF and (b) LHF averaged over the Benin mesoscale domain derived from CFSR, RAS, SAS, and KF2 at coarse and high resolution.

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

The Diurnal Cycle of Precipitation in Regional Spectral Model Simulations over West Africa: Sensitivities to Resolution and Cumulus Schemes

View More View Less
  • 1 * Institute of Industrial Science, University of Tokyo, Tokyo, Japan
  • | 2 Advanced Radar Research Center, University of Oklahoma, and NOAA/National Severe Storms Laboratory, and National Weather Center, Norman, Oklahoma
  • | 3 Atmosphere and Ocean Research Institute, University of Tokyo, Kashiwa, Japan
  • | 4 Department of Atmospheric Science, Kongju National University, Gongju, South Korea
  • | 5 Atmospheric Sciences Research Center, University at Albany, State University of New York, Albany, New York
  • | 6 ** School of Meteorology, University of Oklahoma, Norman, Oklahoma
  • | 7 Advanced Radar Research Center, and School of Civil Engineering and Environmental Sciences, University of Oklahoma, and National Weather Center, Norman, Oklahoma
Full access

Abstract

As a basic form of climate patterns, the diurnal cycle of precipitation (DCP) can provide a key test bed for model reliability and development. In this study, the DCP over West Africa was simulated by the National Centers for Environmental Prediction (NCEP) Regional Spectral Model (RSM) during the monsoon season (April–September) of 2005. Three convective parameterization schemes (CPSs), single-layer simplified Arakawa–Schubert (SAS), multilayer relaxed Arakawa–Schubert (RAS), and new Kain–Fritsch (KF2), were evaluated at two horizontal resolutions (20 and 10 km). The Benin mesoscale site was singled out for additional investigation of resolution effects. Harmonic analysis was used to characterize the phase and amplitude of the DCP. Compared to satellite observations, the overall spatial distributions of amplitude were well captured at regional scales. The RSM properly reproduced the observed late afternoon peak over land and the early morning peak over ocean. Nevertheless, the peak time was early. Sensitivity experiments of CPSs showed similar spatial patterns of rainfall totals among the schemes; CPSs mainly affected the amplitude of the diurnal cycle, while the phase was not significantly shifted. There is no clear optimal pairing of resolution and CPS. However, it is found that the sensitivity of DCP to CPSs and resolution varies with the partitioning between convective and stratiform, which implies that appropriate partitioning needs to be considered for future development of CPSs in global or regional climate models.

Corresponding author address: Yang Hong, National Weather Center, ARRC, Ste. 4610, 120 David L. Boren Blvd., Norman, OK 73072. E-mail: yanghong@ou.edu

Abstract

As a basic form of climate patterns, the diurnal cycle of precipitation (DCP) can provide a key test bed for model reliability and development. In this study, the DCP over West Africa was simulated by the National Centers for Environmental Prediction (NCEP) Regional Spectral Model (RSM) during the monsoon season (April–September) of 2005. Three convective parameterization schemes (CPSs), single-layer simplified Arakawa–Schubert (SAS), multilayer relaxed Arakawa–Schubert (RAS), and new Kain–Fritsch (KF2), were evaluated at two horizontal resolutions (20 and 10 km). The Benin mesoscale site was singled out for additional investigation of resolution effects. Harmonic analysis was used to characterize the phase and amplitude of the DCP. Compared to satellite observations, the overall spatial distributions of amplitude were well captured at regional scales. The RSM properly reproduced the observed late afternoon peak over land and the early morning peak over ocean. Nevertheless, the peak time was early. Sensitivity experiments of CPSs showed similar spatial patterns of rainfall totals among the schemes; CPSs mainly affected the amplitude of the diurnal cycle, while the phase was not significantly shifted. There is no clear optimal pairing of resolution and CPS. However, it is found that the sensitivity of DCP to CPSs and resolution varies with the partitioning between convective and stratiform, which implies that appropriate partitioning needs to be considered for future development of CPSs in global or regional climate models.

Corresponding author address: Yang Hong, National Weather Center, ARRC, Ste. 4610, 120 David L. Boren Blvd., Norman, OK 73072. E-mail: yanghong@ou.edu

1. Introduction

The intensification of the hydrological cycle and extreme hydrological events by natural factors and human activities call for great concern at the global scale (Oki and Kanae 2006). The Sahel has been receiving attention since it suffered from persistent and severe drought conditions from the late 1960s to the mid-1980s, which dramatically impacted resident communities that depend on ecosystem services for their livelihoods. In the mid-1990s, the rainfall returned to near- or above-normal amounts (relative to the 1941–2012 period). However, the recent period (2005–12) experienced three food crises triggered by severe drought linked to a recurrent lack of precipitation in this region (Boyd et al. 2013). More accurate prediction of West African monsoon (WAM) precipitation now, and under climate change, is required to better direct humanitarian aid responses as well as adaptive planning measures.

Current climate models are deficient in simulating the West Africa (WA) seasonal rainfall due to the complex monsoon system. These deficiencies at the seasonal scale may originate and accumulate from much smaller time scales, for example, at the diurnal cycle scale (Taylor and Clark 2001; Parker et al. 2005; Shine et al. 2007). Precipitation frequency and intensity in WA, particularly in extreme events, are likely to change within the context of the changing climate (Trenberth et al. 2003). To assess the evolving characteristics of precipitation, it is essential to examine rain events at a subdaily time scale (Roca et al. 2010). The diurnal cycle is a key test for model reliability and tool for model development, for example, to validate physical parameterizations (Slingo et al. 1987; Garratt et al. 1993; Dai et al. 1999) and to enhance the understanding of important mechanisms that drive the diurnal cycle (Randall et al. 1991), as well as to provide insights for improvements in the representation of subgrid-scale processes (Giorgi and Shields 1999; Betts et al. 1996; Lin et al. 2000). Increased accuracy in the representation of these processes at the subdaily time scale can significantly improve short-term precipitation forecasts, which is indispensable to mitigating the negative impacts of droughts in the WA region.

Precipitation in current climate models is generated by either grid-resolvable (large scale) forcing or subgrid (convective parameterization) processes. The former can be simulated explicitly, while the latter must be parameterized. This consequently increases the uncertainties in reproducing observed precipitation. Most current GCMs are known to produce a diurnal cycle of continental precipitation that is in phase with insolation, with maximum rainfall occurring around midday instead of during the late afternoon (Yang and Slingo 2001; Rio et al. 2009). The bias of the diurnal cycle of precipitation (DCP) indicates that there is disparity in how convective processes are represented in models compared to the real world. The primary cause of the errors has been demonstrated to be the weakness associated with convective parameterization schemes (CPSs; Fritsch and Carbone 2004; Shine et al. 2007), since it is still generally computationally prohibitive to run GCMs and RCMs at cloud-resolving resolutions (e.g., finer than 4 km). A recent study conducted by Koo and Hong (2010) demonstrated that the CPS is the most important trigger in simulating DCP over land. Even more so, complex dynamic and thermodynamic features of the WAM climatological system, such as African easterly waves (AEW), African easterly jet (AEJ), and mesoscale convective systems (MCSs), may add additional difficulty for CPSs’ performance in terms of DCP.

CPS-dependent simulation uncertainties are related to the models’ dependency on horizontal resolution, the impact of which is still controversial. Enhanced skill in resolving the DCP with increasing horizontal resolution is popular logic, since higher resolution better captures the topography, which can significantly affect both rainfall formation (Wallace 1975; Schwartz and Bosart 1979; Balling 1985) and large-scale circulation. This has been demonstrated with several climate models by Shine et al. (2007), Hannay et al. (2012), Davis et al. (2003), Clark et al. (2007), and Yuan et al. (2013). On the contrary, increasing resolution has little effect on the forecast of precipitation in mesoscale models using CPSs (Gallus 1999) and a mixed impact on the American warm season DCP (Lee et al. 2007). To clarify the effects of CPSs and resolution on the simulated diurnal variation of precipitation, sensitivity experiments were conducted in this study at both the regional and the mesoscale. This study will help test the performance of current CPSs over this specific region and help establish a basis for developing new CPSs.

The structure of the manuscript is as follows. In section 2 we introduce the precipitation verification data, model and experimental setup, and method of analysis. Section 3 includes a systematic description of the DCP and presentation of results, including the sensitivity of DCP to CPSs and horizontal resolution. Section 4 provides a summary and our conclusions.

2. Data and methods

a. Precipitation observations and reanalysis data

Observed precipitation at the regional scale was obtained from Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis (TMPA). TMPA combines precipitation estimates from multiple satellites and gauge observations to provide the best estimate of precipitation in each grid box (Huffman et al. 2007). This study uses the research product 3B42 V6 (ftp://disc2.nascom.nasa.gov/data/TRMM/Gridded/3B42\_V6) with a fine spatial (0.25° × 0.25°) and temporal (3 hourly; 0000, 0300, …, 2100 UTC) resolution within the global latitude belt from 50°S to 50°N. The 3-hourly average data are downscaled into hourly data by using a bilinear interpolation method. This study also benefits from the very high quality in situ observational data from rain gauge stations provided by African Monsoon Multidiscplinary Analysis (AMMA) Land Surface Model Intercomparison Project Phase 2 (ALMIP2; Boone et al. 2009), which is supported by the AMMA–Coupling the Tropical Atmosphere and the Hydrological Cycle (CATCH) observing system (Cappelaere et al. 2009; Lebel et al. 2009). The data are provided at a 0.05° resolution with a 30-min temporal resolution using the Lagrangian kriging interpolation technique (Vischel et al. 2011). This study only focuses on one heavily instrumented supersite square in Benin.

The National Centers for Environmental Prediction (NCEP) Climate Forecast System Reanalysis (CFSR) data (Saha et al. 2010) are utilized to examine the local surface-driven processes. The CFSR is a global reanalysis with a horizontal resolution of ~38 km (T382) with 64 vertical levels. The guess field of CFSR is taken from the 6 hourly forecast from a coupled atmosphere and ocean system with an interactive sea ice model. It also includes assimilation of satellite radiances.

b. Model and experimental setup

1) Model description

In this study, the NCEP/Environmental Modeling Center Regional Spectral Model (RSM) is used. A detailed model description of the RSM is given by Juang and Kanamitsu (1994), Juang et al. (1997), and Hong and Leetmaa (1999). The RSM has been widely used over East Asia because of its ability to reproduce that region’s summer monsoon. It has also been developed with incorporation of stable water isotopes to study atmospheric river events over the United States (Yoshimura et al. 2010). To date, very few studies have been conducted over the WAM region using the RSM. Therefore, this study provides an opportunity to evaluate the model performance with the influence of WAM dynamics. Investigating model deficiencies can give us insights for future model development.

In the RSM, perturbations of temperature, mixing ratio, pressure, and divergence are spectrally represented and nudged by a two-dimensional cosine series. A regional climate model with this spectral nudging technique has smaller large-scale biases and does not depend on the domain size. In addition, a global downscaling technique (Yoshimura and Kanamitsu 2008) can also be physically better expressed to constrain the large-scale background in a realistic way, compared to the conventional lateral boundary nudging technique.

Major physical processes adopted in the RSM include land surface effects [Noah land surface model (Ek et al. 2003)], long- and shortwave radiation [Chou scheme (Chou and Suarez 1994)], cloud–radiation interaction, planetary boundary scheme (Hong and Pan 1996), large-scale condensation, gravity wave drag (Alpert et al. 1988, 1996), and enhanced topography, as well as vertical and horizontal diffusion (Hong and Pan 1996; Kim and Mahrt 1992). Since deep and shallow convection are very important physical processes for triggering precipitation and are of special concern to this study, a more detailed description of the convective parameterization schemes are introduced below.

Deep convective parameterization schemes have proven to be one of the most challenging parts of numerical atmospheric modeling. Three CPSs are available in the current RSM: relaxed Arakawa–Schubert (RAS; Moorthi and Suarez 1992), simplified Arakawa–Schubert (SAS; Pan and Wu 1995), and the new Kain–Fritsch approach (KF2; Kain 2004). In this study, these three CPSs are used to study the sensitivity of the diurnal variation of simulated precipitation. Both RAS and SAS made simplifications of the standard Arakawa and Schubert scheme (AS; Arakawa and Schubert 1974), which postulated that the level of activity of convection is such that their stabilizing effect balances the destabilization by large-scale processes. RAS has been widely tested for seasonal forecasts at NCEP, while SAS has advantages for medium-range forecasts (Kalnay et al. 1996). The main difference between RAS and SAS lies in two aspects: 1) the cloud model and 2) the treatment of downdrafts (Kang and Hong 2008). Specifically, the clouds in RAS have different tops, while SAS has only one type of cloud. RAS does not include any downdraft mechanisms, but SAS considers saturated downdrafts. Different from the AS scheme, the closure assumption in KF2 is based on the convective available potential energy (CAPE) for an entraining parcel, and this can provide more reasonable precipitation rates. KF2 made modifications to the cloud model by permitting cloud development (variable cloud radius and cloud-depth threshold) for deep convection. KF2 has been widely tested for climate and mesoscale modeling studies (Wang and Seaman 1997; Ridout et al. 2005).

2) Experimental design

Figure 1 shows the model domain over West and North Africa based on a Mercator map projection with 20-km horizontal resolution. The area enclosed by the black box indicates the high-resolution simulation domain (10 km). The RSM uses 28 vertical sigma levels, consistent with the NCEP–U.S. Department of Energy (DOE) Atmospheric Model Intercomparison Project (AMIP)-II reanalysis (R2; Kanamitsu et al. 2002) data to avoid errors due to vertical interpolation of large-scale fields from the global model (Hong and Leetmaa 1999). Table 1 shows the details of the model location and the corresponding grid numbers.

Fig. 1.
Fig. 1.

Regional model domain (excluding buffer zone). The solid black box is for high-resolution (10 km) simulations. The Benin mesoscale site is outlined as a gray box.

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

Table 1.

Geographical characteristics of the regional model domain.

Table 1.

The lateral boundary and initial conditions are taken from the reanalysis-nudged isotope-incorporated AGCM simulation conducted by Yoshimura and Kanamitsu (2008). In their method, the surface boundary conditions, sea surface temperature, and sea ice distribution are taken from R2. The R2 data are used for the large-scale forcing to drive the AGCM to produce the large-scale analysis by using the global downscaling technique. The time steps for the model integration vary by season and are 40, 60, 40, and 30 s for spring, summer, autumn, and winter, respectively. Simulation results are generated at hourly intervals and interpolated onto 12 vertical pressure levels: 1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, and 100 hPa. All experiments (coarse and fine resolution with different convection schemes) start at 0000 UTC 1 January 2004 and terminate at 2300 UTC 31 December 2005. The first year of simulation (e.g., January–December 2004) is regarded as spinup and discarded. The WAM season (April–September) of 2005 is chosen for the main analysis. Six sensitivity experiments with three CPSs and two resolutions are indicated in Table 2.

Table 2.

Description of sensitivity experiments.

Table 2.

c. Analysis method

In this study, we applied harmonic analysis (Dai 2001; Basu 2007; Koo and Hong 2010) to characterize the diurnal cycle of precipitation, focusing on the relative amplitudes and phases of the diurnal (24-h period) and semidiurnal (12-h period) variation. The analysis procedures found in Haurwitz and Cowley (1973) were followed. The basic idea of harmonic analysis is to represent the fluctuations or variations in a time series by adding a series of sine and cosine functions (Wilks 2005). A more accurate fit is obtained as more sinusoids are added. However, we will only focus on the first two harmonics, which have physical meaning and dominate the total precipitation. Mathematical equations to fit the diurnal variation with the first and second harmonic function can be represented by the following form:
e1
where
e2
is the hourly precipitation (24 points, n = 24), is the daily mean value, and and indicate the first harmonic (diurnal cycle) and second harmonic (semidiurnal cycle), respectively. Here, k = 1 or 2, and t = (0000, 0100, 0200, … , 2300) is the time in hours (UTC). In addition, and are amplitude and phase, respectively. They can be calculated using the following equations:
e3
e4
Coefficients and , corresponding to , can be obtained according to (5) and (6) using multiple regression methods:
e5
e6
The diurnal cycle of precipitation is well known to be driven by solar heating or surface and atmospheric oscillation (referred to as atmospheric tides), while the mechanisms for explaining the semidiurnal cycle are complex and can be linked to many physical processes that vary by region. For example, Brier (1965) demonstrated that in the United States is related to solar and lunar tidal forces. Mohr (2004) found that a semidiurnal cycle of precipitation exists at 10°N in sub-Saharan Africa, which is related to the frequency and life cycle of the organized convective systems. Studies focusing on southeastern China indicated that can be linked to low-level atmospheric thermal advection (Ramage 1952; Chen et al. 2010), convergence of surface wind and water vapor flux (Wai et al. 1996; Chen et al. 1999), radiative cooling at cloud top (Chen et al. 2010), and land–sea differential radiative heating–cooling (Huang and Chan 2011).

3. Simulation results and discussion

a. The diurnal cycle of precipitation at regional scale

1) Amplitude

Normalized amplitude is more appropriate for evaluating diurnal variation since the amplitude depends on the daily mean values. The values normalized by the daily average precipitation will be more useful for long-term datasets since they are less affected by interannual variability. Figure 2 describes the spatial distributions of the normalized amplitude explained by the diurnal cycle of precipitation from TMPA and three simulations with different CPSs. From the observations, we can see a large part of the domain has amplitudes between approximately 0.4 and 0.8 in central and West Africa. High values of amplitude (more than 1.6) always occur in a vegetation gradient zone (transition from savanna to desert). Amplitudes are usually weakened over ocean compared to those over continental areas. In these regions, the mean precipitation is very low, but sudden rainfall is frequent (Koo and Hong 2010). The amplitudes from the simulations (RAS, SAS, and KF2) are comparable to those from the observations, not only in terms of absolute value but also regarding the spatial pattern. Obvious differences among the three CPSs can be found over the regions with higher amplitude. For example, KF2 produces the highest amplitudes among the three schemes over the North Atlantic Ocean and in particular during the pure monsoon season (July). RAS and SAS show very similar patterns after the monsoon onset, for example, during July and August. However, KF2 and SAS show similar patterns in premonsoon months, such as April and May. This may indicate that CPSs have different responses to the large-scale forcing and feature different characteristics of the monsoon dynamics for different months, although the large-scale monsoon circulations are similar in all the simulations.

Fig. 2.
Fig. 2.

Normalized amplitudes of precipitation diurnal variations derived from (a) TMPA, (b) RAS, (c) SAS, and (d) KF2 during the monsoon season (April–September).

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

2) Phase

The diurnal cycle of the precipitation phase is shown in Fig. 3 from both the observations and the simulations. The color bar indicates local solar time (LST) with a 3 hourly interval. Cyan represents early morning and orange indicates late afternoon. As shown in Fig. 3a from satellite observations, the hour of maximum precipitation has embedded features appearing from the late afternoon to evening (~1500–2100 LST) and at nighttime (~2100–0300 LST) over the interior land including the intertropical convergence zone (ITCZ). As for the ocean, maximum precipitation mainly occurs during daybreak or (early) morning (~0300–0900 LST). The late afternoon peak over land is related to deep convection, which is driven by the increased sensible heat flux and will lead to significant upward motion of water vapor. The early morning peak over the ocean can be explained by the so-called static radiation–convection and dynamic radiation–convection mechanisms (Ramage 1971; Gray and Jacobson 1977). The nighttime peak may be related to nocturnal low-level jets (NLLJs), which bring moisture from the ocean to the land during the monsoon and impact rainfall variations (Sperber and Yasunari 2006). Pu and Cook (2010) find that in the West African westerly jet region, winds exhibit a semidiurnal cycle, which peaks at 0500 and 1700 LST. Their findings help to explain why there is a mixed phase (~0300–0600 and ~1500–1800 LST) over the eastern Atlantic and the West African coast in TMPA observations as shown in Fig. 3a. The development of NLLJs would also help to sustain the strong diurnal cycle of organized deep convection, which initiates in the afternoon and further develops at night (Laing et al. 2008; Gounou 2011). The nighttime peak can also be related to the effect of mesoscale convective systems (Nikulin et al. 2012), which are usually initiated around ~1700–1800 LST, but precipitate at maximum intensity several hours later during their mature phase (McGarry and Reed 1978; Hodges and Thorncroft 1997). Regarding model simulations, RAS, SAS, and KF2 all capture the distinct contrast between land (late afternoon) and ocean (early morning) well. However, the phases are ~3 h earlier and there is no obvious nighttime peak over coastal areas.

Fig. 3.
Fig. 3.

Phase (peak timing) of precipitation diurnal variations derived from (a) TMPA, (b) RAS, (c) SAS, and (d) KF2 during the monsoon season (April–September). LST is indicated by the color bar with a 3-h interval.

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

An interesting feature is that RAS and SAS have a distinct early morning peak over the transition area (~15°–20°N) that is not shown in KF2 or the observations, which both have the mixed timing. This phenomenon (mixed pattern) is related to the propagation of large-scale precipitation while RAS and SAS cannot capture the propagation signal well. The propagation characteristics of the rainfall system can be depicted by using a Hovmöller diagram. To evaluate whether the model is able to replicate the propagating signal in the east–west direction, diurnally averaged Hovmöller diagrams (Fig. 4) were constructed using the hourly normalized (by daily mean) rainfall between 2.5° and 38°N from both observations and simulations. The Hovmöller diagram of TMPA (Fig. 4a) clearly shows a coherent propagating rainfall axis along 20°–5°W over ocean areas from July to September. From April to June, this signal appears to be fairly weak. Only simulations from KF2 capture this westward propagation in the corresponding region, although for only a few months (June, July, and September), and the signal is quite noisy. RAS, SAS, and KF2 even produce eastward propagation of rainfall systems in premonsoon months (April–May). These results indicate that the RSM with CPSs has difficulties simulating the propagating rainfall system, which can explain the missed mixed phase in the simulations. In addition, both observations and simulations exhibit a slow propagation signal in the eastern part of the analysis domain (from 5°W to 25°E). However, RAS and SAS simulate the afternoon rainfall ~2–3 h earlier, as seen in the shift of the propagating rainfall axis in simulations compared to that in the observations. This can further explain why the mixed phase is advanced in simulations (RAS and SAS) compared to that in the observations (TMPA), as shown in Fig. 3. The other interesting observation is that the mixed early morning and late afternoon peaks shown in the three CPSs are close to the North Atlantic Ocean region, for example, around Morocco. This may be related to the high altitude, as shown by Dai and Trenberth (2004) and Koo and Hong (2010).

Fig. 4.
Fig. 4.

Hovmöller diagram (time–lon distribution) for the normalized (by daily mean) rainfall diurnal variation derived from (a) observations (TMPA) and (b)–(d) model simulations with different CPSs (RAS, SAS, and KF2). The y axis indicates LST.

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

3) Land versus ocean

Figures 5 and 6 compare the normalized diurnal variation of precipitation amount from TMPA, RAS, SAS, and KF2 over land and ocean as well as their corresponding first and second harmonics. According to TMPA observations, the time of maximum precipitation varies by region, with an afternoon peak (~1500–1800 LST) over land and a morning peak (~0800–1200 LST) over ocean. Peak time also varies among the 6 months and there is a ~1–2-h delay in the later monsoon months (July–September) compared to the premonsoon months (April–June). Large seasonal variability is not seen in the amplitudes, but differs by region. Land exhibits slightly higher (~0.4–0.5) amplitudes compared to those of the ocean (~0.25). For the simulations, the general features (i.e., periodicity and monthly variations of amplitude) are fairly well reproduced compared to the observations. In Fig. 5, the simulated amplitude is higher than observed, and the phase is advanced ~2–3 h prematurely. Moreover, RAS shows the closest amplitude, but always advanced the phase too much compared to SAS and KF2. The only noticeable difference between SAS and KF2 is manifested by exaggerated amplitude; the peak time is almost the same. Sensitivity of diurnal variation to the CPSs is also different over land versus ocean. Large uncertainties can be found over land between the three CPSs; however, the disparity over ocean is quite small, as seen in Fig. 6.

Fig. 5.
Fig. 5.

(a) Normalized diurnal variations derived from satellite and RSM output (dashed lines) and fitted models from the first two harmonics (continuous lines), and (b) fitted diurnal and (c) fitted semidiurnal cycles of precipitation amount from TMPA, RAS, SAS, and KF2 over land.

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

Fig. 6.
Fig. 6.

As in Fig. 5, but for the ocean.

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

Diurnal and semidiurnal cycles can be described by the first and second harmonics. From Figs. 5 and 6, we can see the total variance is dominated by the first harmonic (diurnal cycle). The semidiurnal cycle is weaker than the diurnal cycle, and the amplitude is reduced to ⅓ over land and, in particular, almost ⅕ or less over ocean. The simulated variance explained by the diurnal and semidiurnal cycles is larger than the TMPA over land, but smaller (larger) for diurnal (semidiurnal) cycles over ocean. This differs from previous studies as shown by Koo and Hong (2010), whose simulation results reveal similar amplitudes of (over ocean) and (over land and ocean) compared to the observations. The afternoon peak shown in is shifted ~2 h ahead over land but shifted almost 5 h ahead over ocean. For the semidiurnal cycle, SAS and KF2 are similar compared to TMPA for almost all the months, not only in amplitude but also in phase. However, RAS sometimes has out-of-phase behavior, as we can see for May and September of Figs. 5 and 6 .

b. The diurnal cycle of precipitation at mesoscale

1) Total precipitation

In the following section, the diurnal cycle of precipitation over the Benin mesoscale domain simulated by different CPSs (RAS, SAS, and KF2) at different spatial resolutions (20 and 10 km) will be compared to observations and discussed in detail. Figure 7 shows the diurnal cycle of total precipitation at the Benin mesoscale domain in both observations (ALMIP2) and three simulations (RAS, SAS, and KF2) from April to September. The 20- and 10-km results are presented in Figs. 7a and 7b, respectively. In the observations, there are often two peaks, one in early morning–nighttime and the other in late afternoon, but the late afternoon peak is more pronounced. In the coarse-resolution trial, all of the simulations generate larger amplitudes during the first 3 months, but smaller amplitudes in July–September. In addition, the time of the simulated precipitation maximum always precedes the observations, which is a general problem in current climate models. Among the simulations, the RAS peak time is the earliest and is associated with the weakest amplitude. SAS and KF2 show similar trends and there are only slight differences between the two in the premonsoon months (April and May), not only regarding amplitude but also phase. After the monsoon onset, differences between SAS and KF2 increase, especially for the amplitude (July and August) and phase (September). The month showing the worst features is July, which has very weak precipitation in all three of the simulations. The reason for this is not clear yet, but it may be due to the parent GCM’s discrepancy in incorrectly representing the large-scale monsoon dynamics (not shown). As for the high-resolution simulations, there are no systematic advantages in simulating the phase and amplitude. For example, precipitation simulated using RAS is larger than that at coarse resolution for almost all the months. However, SAS and KF2 display much more complex behavior among different months and there is no uniform conclusion. In general, differences among the three CPSs are smaller during the premonsoon season at coarse resolution, but amplified at high resolution. The differences in the later 3 months are large at coarse resolution, but reduced at high resolution. This indicates that CPSs are sensitive to the resolution as well as monsoon dynamics.

Fig. 7.
Fig. 7.

Seasonal (April–September) evolution of the diurnal cycle for total precipitation averaged over the Benin mesoscale domain derived from ALMIP2, RAS, SAS, and KF2 at (a) coarse and (b) high resolution. Continuous lines are the first two harmonics fitted from dashed lines.

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

2) Convective versus large-scale precipitation

To explain why there is no significant improvement at high resolution for any of the three CPSs, convective precipitation and large-scale precipitation are further examined separately in Figs. 8 and 9. In the RSM, convective precipitation is generated by the deep and shallow convective schemes, while the large-scale precipitation is formulated based on the saturation condensation. Compared to the total precipitation (Fig. 7), convective precipitation only has a single diurnal peak, which always occurs in late afternoon. In 20-km simulations (Fig. 8a, below), SAS produces the largest amount of precipitation and the differences compared to RAS and KF2 are amplified in the later monsoon months from July to September. However, in 10-km runs (Fig. 8b, below), the results are quite different. The convective precipitation rate is substantially larger for RAS, but reduced for SAS and KF2. Differences among RAS, SAS, and KF2 are more apparent from April to June, while they converge from July to September in terms of phase and amplitude. Regarding large-scale precipitation at coarse resolution (Fig. 9a), RAS has obvious early morning and nighttime peaks, which are supposed to dominate the early morning phase of total precipitation (Fig. 7a). However, there is almost no large-scale precipitation in SAS except during June, but there is a relatively large amount in KF2, for example, during May, June, and September. When compared to coarse-resolution simulations, SAS and KF2 with high resolution produce much more large-scale precipitation, but it is reduced in most of the months for RAS. An obvious early morning peak can be found from April to July, while the peak time shifts to late afternoon in August, associated with the enlarged amplitude. Compared to Figs. 8a (below) and 8b (below), high-resolution simulations produce more large-scale precipitation than coarse-resolution cases because at higher resolution it is easier for a grid box to become saturated, which therefore dominates the total precipitation. The above analysis reveals that partitioning between convective and large-scale precipitation is significantly modulated by resolution and the role of different CPSs is weakened when resolution increases.

Fig. 8.
Fig. 8.

Seasonal (April–September) evolution of the diurnal cycle for CAPE (above) and convective precipitation (below) averaged over the Benin mesoscale domain derived from CFSR, RAS, SAS, and KF2 at (a) coarse and (b) high resolution. For convective precipitation, continuous lines are the first two harmonics fitted from dashed lines. The color bar (center) indicates shortwave downward radiation (W m−2) from ALMIP2.

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

Fig. 9.
Fig. 9.

Seasonal (April–September) evolution of the diurnal cycle for large-scale precipitation averaged over the Benin mesoscale domain derived from ALMIP2, RAS, SAS, and KF2 at (a) coarse and (b) high resolution. Continuous lines are the first two harmonics fitted from dashed lines.

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

3) Taylor diagram

To evaluate the model performance in a more quantitative way, the Taylor diagram (Taylor 2001) was adopted to graphically summarize the similarity [in terms of correlation R, centered root-mean-square error (RMSE), and standard deviation (std dev)] between the observations and simulations. The Taylor diagram is especially useful in summarizing model skills with a large ensemble. Figure 10 is a Taylor diagram showing R, RMSE (mm h−1), and std dev (mm h−1) for precipitation amount from April to September over the Benin mesoscale domain. Statistics are calculated based on the time series as shown in Fig. 7. A number is assigned to each model sensitivity experiment in terms of the DCP. The star (reference point) indicates statistics from the observations. The dashed line shows the contour of the reference std dev and gray lines are RMSE contours. Generally speaking, numbers nearest to the reference point will have the best agreement with the observations since they have high R and low RMSE. Numbers on the dashed arc indicate that the model has the correct std dev. Figure 10 indicates that simulations exhibit different diurnal variabilities over the months, as reflected from the std dev. Before monsoon onset (April and May), almost all of the experiments have larger diurnal variation since they lie far outside of the dashed arc. However, experiments show good agreement in June and less variation from July to September. There is no overall optimal pairing that shows better agreement in all six of the months. For example, SAS_10 lies closest to the reference point in May and August, but relatively far in other months. There is no dominant advantage for one specific convective parameterization scheme either since they show different statistics during different months. As we show, high-resolution experiments do not always show improvement compared to coarse resolution. Slight improvement can only be found in May and September for the SAS convection scheme.

Fig. 10.
Fig. 10.

Taylor diagrams summarizing the statistics of precipitation amount from the observations and each sensitivity experiment.

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

c. Comprehensive evaluation of model sensitivity experiments

To understand the performance in producing DCP among different sensitivity experiments, it is therefore necessary to explore key mechanisms that drive the diurnal nature of precipitation from a physical perspective. Specifically, physical processes linked to land surface conditions and atmospheric instability, which are important triggers of precipitation, will be discussed below in detail.

1) Convective available potential energy

CAPE is one of the useful convective indices for investigating the trigger of deep convection. It is defined as the amount of potential energy of an air parcel lifted to a certain level vertically through the atmosphere. The higher the CAPE, the more energy is available for convection. In this section, examination of CAPE will be discussed at the mesoscale to evaluate the atmospheric instability and the relationship with the simulated diurnal variation of precipitation. Figures 8a and 8b further examine the relationship between the diurnal cycle of CAPE and the diurnal cycle of convective precipitation for both coarse (20 km) and high resolution (10 km). Shortwave downward radiation (SDR) at the diurnal time scale from ALMIP2 observations is also shown as the color bar in Fig. 8 (center). Additional knowledge can be gained from Fig. 8 compared to the previous discussion. The onset of CAPE agrees well with sunrise (indicated from SDR). However, the peak time for CAPE lags behind the time of maximum SDR since the atmosphere will continue to be heated in the afternoon. Simulated CAPE of later monsoon months (July–September) is smaller than that in premonsoon months (April–June). This trend is also demonstrated in the CFSR data. The reduced CAPE in the later 3 months is due to the reduced downward shortwave radiation (Fig. 8, center) and the surface temperature (not shown). In addition, the RSM (both coarse and high resolution) does not produce enough dynamics and fluctuation for CAPE compared to the reanalysis, except during May and June. Although the peak time of simulated CAPE is well captured at coarse resolution (Fig. 8a), the maximum CAPE is delayed in premonsoon months and advanced ~4–5 h in the later 3 months at high resolution. For coarse resolution, SAS seems to perform better than RAS and KF2, in terms of both amplitude and phase. However, simulated CAPE from KF2 is the most sensitive to resolution and KF2 yields better representation at high resolution from June to August (Fig. 8b).

The diurnal variation of convective precipitation follows the trend of CAPE well except for RAS in the 20-km run, whose maximum convective precipitation lies ahead of CAPE (except during April). In addition, the higher CAPE is, the larger amount of convective precipitation can be obtained. Although RAS, SAS, and KF2 display high similarities of CAPE, the differences can be amplified in convective precipitation significantly, as shown during April and May for RAS as well as SAS from July to September. As for high-resolution results in the later 3 months, CAPE in RAS and SAS is not as sensitive as that for KF2, which does not have a remarkable decrease. In addition, the proportionality relationship between convective precipitation and CAPE found at coarse resolution disappears at high resolution. This indicates that the role of CAPE is weakened in high-resolution simulations.

2) Vertical profiles

Vertical profiles are useful tools for studying the stable–unstable structure of the atmospheric boundary layer since the change of vertical profiles is affected by surface heating–cooling, moisture transfer, cloud formation, advection, and subsidence. As different CPSs have different assumptions for the cloud layer, as well as the entrainment–detrainment effect, vertical profiles can provide insight into how heat fluxes are transported and therefore enhance the convective initiation and increase precipitation. To understand the possible reasons for the differences among the six sensitivity experiments, area-averaged vertical anomalies from the six-ensemble mean in specific humidity (kg kg−1) and temperature (K) are further examined in Fig. 11. In general, the overall anomaly patterns along the vertical profile are similar for all three CPSs; however, they are different in terms of the magnitude, and resolution has a large impact on the simulations. For the specific humidity simulated by coarse resolution (solid lines in Fig. 11, top), dry moisture anomalies are commonly apparent for all three CPSs in all tropospheric layers with a maximum value at ~700–800 hPa from July to September. However, during the premonsoon months (April and May), wet anomalies can be found in the lower troposphere (below 900 hPa). In contrast, high-resolution simulations show different behavior with a wet anomaly for all months in all tropospheric layers from July to September, but a dry anomaly below 900 hPa from April to June. Additionally, large diversity can be found among the three CPSs in the coarse-resolution simulations. The magnitude of the anomaly increases, generally, from KF2, RAS, and SAS, in this order. However, for high-resolution runs, differences among the three CPSs are very small, and KF2 has the wettest anomaly. These vertical differences can be explained according to the cloud models and treatment of downdrafts adopted for the CPS. RAS and KF2 have different cloud tops and allow convective moist air to detrain to the environment at different levels, but SAS has only one deep cloud, and detrainment is allowed only at the cloud tops. Also, SAS considers saturated downdrafts, while this is not true for RAS. It means that RAS and KF2 allow more flexibility in the interaction between the vertical profiles and the environment. A cloud model with one layer like SAS constrains bringing moisture from the environment and leads to a drier bias compared to the cloud model with ensemble layers. That can explain why SAS always shows the biggest anomaly and RAS and KF2 look quite similar. We can also find that differences in the last 3 months are not as large as those for the first 3 months, not only for coarse resolution, but also for high resolution.

Fig. 11.
Fig. 11.

Vertical profiles of the anomalies from the ensemble mean in (top) specific humidity (kg kg−1) and (middle) temperature (K). (bottom) Vertical profiles of omega (Pa s−1). Negative (positive) values mean rising (subsiding) motion.

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

For temperature anomalies (see Fig. 11, middle), differences among the six sensitivity experiments are not as significant as those for moisture anomalies. Generally, high and coarse resolutions show opposite anomalies, and there is no uniform hot–cool bias for all the vertical layers. High-resolution simulations exhibit cool anomalies below 900 hPa, while the hot anomalies occur at coarse resolution. For the troposphere above 600 hPa, high resolution tends to have hot anomalies but coarse resolution tends to have cool anomalies. These differences in vertical profiles of specific humidity and temperature are in line with those in surface fluxes. Both of them serve as the scaffolding needed to investigate the vertical heating and moistening, as described in detail in the following section.

The vertical velocity (omega) is the direct measurement of how a model creates condensation and precipitation. Therefore, analyzing this variable gives us a more comprehensive understanding of the differences among various CPSs. The domain-averaged omega from April to September is shown in Fig. 11 (bottom). Negative values of omega indicate updrafts, which are associated with convective activity. The overall distributions of omega along the vertical levels are similar for all six experiments. Diversities among the six experiments become apparent below 500 hPa with the largest diversity occurring near the surface. Omega among RAS, SAS, and KF2 are close to each other for the 10-km simulations. Differences among these three CPSs increase at coarse resolution. For example, in premonsoon months (April and May) and late monsoon months (August and September), large differences between the solid lines and dashed lines (Fig. 11, bottom) reveal that omega is more sensitive to the change of horizontal resolution than to the change of CPS. Looking into the details, RAS configured with high horizontal resolution produces the largest vertical velocity differences compared to coarse resolution over the whole study period. This can be directly linked to the total precipitation and could explain why there are obvious increases at high resolution (Fig. 7b) compared to coarse resolution (Fig. 7a) for RAS, but not for SAS and KF2.

3) Flux comparison

The diurnal cycle of convection is also closely related to the strong diurnal cycle of sensible heat flux (SHF) and latent heat flux (LHF) in response to the solar forcing. Comparison of SHF and LHF on the diurnal time scale from the six experiments is depicted separately in Figs. 12a and 12b. As observed from the reanalysis, the Benin mesoscale site is characterized by large SHF and small LHF in premonsoon months, as the soils are dry. When the monsoon rain comes and the soil becomes wetter, SHF decreases and LHF increases, resulting in a high source of moisture from the surface, which is favorable for the development of moist convection. In addition, SHF (LHF) derived from the reanalysis shows a clear decreasing (increasing) trend from April to September. However, these trends are not apparent in the model simulations. The largest difference between the simulations and reanalysis happens in premonsoon months, whereas the model does a decent job of reproducing the diurnal amplitude for the later 3 months. The underestimation of SHF and overestimation of LHF compensate each other and demonstrate a reasonable overall energy budget from the model perspective. The other thing we notice is the significantly reduced SHF and LHF at high resolution compared to those at coarse resolution. According to the energy budget, it is highly possible that the magnitude of the incoming solar flux at the surface is reduced because of the increase in cloud fraction. A more interesting observation from Fig. 12a is the different degree of sensitivity of SHF to resolution and CPS. In the premonsoon months, substantial differences can be found among different CPSs, while in later periods much larger differences are found between different resolutions. However, the effect of CPS on LHF (Fig. 12b) is not as apparent as that for SHF. It is the changing of the resolution that dominates the main difference in LHF among the six experiments. The reduction in LHF at high resolution can easily be caused by high values of specific humidity near the surface (Fig. 11, top), thus reducing the moisture gradient between the surface and the immediate atmosphere.

Fig. 12.
Fig. 12.

Seasonal (April–September) evolution of the diurnal cycle for (a) SHF and (b) LHF averaged over the Benin mesoscale domain derived from CFSR, RAS, SAS, and KF2 at coarse and high resolution.

Citation: Weather and Forecasting 30, 2; 10.1175/WAF-D-14-00013.1

Combined results from vertical profiles (Fig. 11) and a flux comparison (Fig. 12) further elucidate the simulated precipitation differences. Figure 11 shows that the lower troposphere (below 900 hPa) is warmer at coarse resolution than at high resolution. This warming can be explained by the larger SHF (as shown in Fig. 12a) from the surface, which potentially causes the enhanced mixing of the boundary layer. The enhanced mixing will further reduce the moisture near the surface (Fig. 11, top) but increase it above, which transports moisture from the land to the atmosphere and increases the moist static energy. This results in the enhancement of convective initiation and increase in the amount of precipitation as shown in Fig. 8. As mentioned by Couvreux et al. (2012), SHF has a dominant role for deep initiation compared to LHF over this semiarid condition. For lower SHF, increasing the LHF is not sufficient for deep convection to initiate. For example, at coarse resolution, even though the LHF in KF2 is much smaller than RAS and SAS (Fig. 12b), the amount of convective precipitation in KF2 is still comparable and even larger than RAS (Fig. 8a) since the magnitude of SHF does not vary too much. Conversely, when SHF is similar among the three CPS, the larger the LHF, the larger the amount of convective precipitation would be. This might explain why there is significantly reduced convective precipitation at high resolution compared to that at coarse resolution from another aspect, which is partly due to the reduction in the supply of moisture (LHF). This further substantiated the reduced proportion of convective precipitation, which is produced by subgrid-scale parameterization.

4. Conclusions

In this study, the diurnal cycle of precipitation (DCP) during the 2005 West African monsoon is simulated by the NCEP RSM and evaluated against a rain gauge network at the mesoscale and satellite observations at the regional scale. The effects of convective parameterization schemes and resolution on model realism are discussed. Conclusions from the analysis can be summarized as follows.

a. General behavior of the RSM

Comparison of model-simulated DCP with observations reveals that the RSM could capture precipitation’s diurnal variation, with a late afternoon peak over land and an early morning peak over ocean. This spatial pattern is well captured compared to satellite observations in most areas. In terms of the phase for the diurnal cycle, the RSM advances the maximum precipitation too much, which is more apparent in diurnal variations than semidiurnal variations , over land and ocean. This is the general deficiency of current climate models in consensus with previous research. Furthermore, component analysis was conducted by separating total precipitation into the convective and large-scale parts. Results indicated that convective precipitation takes a large proportion and dominates the phase of the total precipitation. Results from the mesoscale comparison further displayed the complexity of the DCP, which has a large seasonal variability. In addition to the prevailing peak that occurred during the late afternoon, there is also an observed morning peak that does not appear at the regional scale. Because the mechanisms for the occurrence of morning peak precipitation are complex, in general, the RSM cannot capture it very well. This study further found that the morning peak is more closely related to the large-scale precipitation, which has an obvious morning maximum precipitation.

b. Effects of resolution

Contrary to the expectation that increasing resolution can improve the simulated diurnal variation of precipitation, results shown in this study do not show any clear advantage in high-resolution runs. Specifically, high resolution tends to generate higher precipitation amounts compared to those of coarse-resolution simulations. This is mainly due to the increase in large-scale precipitation according to the component analysis. This relatively robust result is in agreement with previous studies (Boyle and Klein 2010; Lin et al. 2012). Our study highlighted the controversial issue related to the resolution effects. Although the topography can be better resolved at higher resolution, this cannot ensure the overall improved performance. Moreover, Benin is a very flat domain, and the role of topography in the formation of precipitation is not very noteworthy. It is also noted that some dominant inherent biases coming from other important physical processes, which are not well represented within the current framework, are expected to play a more important role in the simulation. This explanation is still in its infancy, but some hints can be gleaned from the sensitivity tests of CPSs due to changes in the resolution.

c. Effects of convective parameterization schemes

Effects of convective parameterization schemes on the simulated DCP are systematically examined both at the regional and mesoscale by changing the CPSs adopted in the RSM, which are RAS, SAS, and KF2. These three CPSs show very similar patterns in terms of the spatial distributions of the phase and amplitude at the regional scale, although a significant difference tends to occur within the ITCZ region. Regarding the seasonal evolution of the diurnal cycle, there is no obvious shift of the peak time at the regional scale, and the main difference is found for amplitude. It is noted that the degree of sensitivity of the diurnal cycle to CPS is different in the premonsoon season and pure monsoon season as well as at different resolutions, as indicated by the comparison for total precipitation, convective precipitation, and large-scale precipitation, as well as the vertical profiles and surface fluxes. Generally speaking, simulations among the three CPSs at coarse resolution show smaller diversity in the premonsoon season than that in the pure monsoon season. However, this pattern is reversed in high-resolution simulations. That is, the diversity among the three CPSs is larger in the first 3 months than that in the later 3 months. This may be related to the complex monsoon dynamics, which needs further investigation. These findings are of significance because they provide a foundation for future development of CPSs with the consideration of resolution effects and large-scale climate dynamics (monsoon systems). Statistics summarized in Taylor diagrams (Fig. 10) do not show any overwhelming advantage for one specific CPS–resolution pairing for all months. One possible explanation is that atmospheric conditions that can satisfy the assumption of CPSs have large seasonal variability.

Acknowledgments

This study was supported by the Japan Society for the Promotion of Science “KAKENHI” (23226012); the Japanese Ministry of Education, Culture, Sports, Science, and Technology “SOUSEI” Program; the Japanese Ministry of Environment Environmental Research and Technology Development Fund (S-10: comprehensive study to develop a global climate change risks management strategy); and the seventh research announcement for Precipitation Measuring Mission (PMM) managed by Japan Aerospace Exploration Agency. We would like to acknowledge the AMMA Land Surface Model Intercomparison Project (ALMIP) and the AMMA-CATCH network for use of the mesoscale precipitation fields. The authors are also grateful to the editor and three anonymous reviewers for their reviews and valuable suggestions during the review process.

APPENDIX

Future Work

There still remain some limitations in this research that merit further consideration. Key feedback mechanisms involved in the diurnal cycle, including solar radiation heating, cloud formation, and land–atmosphere interactions, as well as the model biases, deserve further progress (Trenberth et al. 2003). Improvements and efforts related to this study can be extended by focusing on the following aspects.

First of all, a large discrepancy was found between simulations and observations in terms of the phase. This may be related to the planetary boundary layer (PBL) scheme. The current PBL parameterization adopted in the RSM is the Medium-Range Forecast Model (MRF) PBL scheme (Hong and Pan 1996). This scheme has an early development of PBL and therefore simulations in this study advanced the time of maximum precipitation. It would be better to apply a revised vertical diffusion package, for example, the Yonsei University PBL (YSU PBL; Hong et al. 2006), which has a nonlocal turbulent mixing coefficient in the PBL. The YSU PBL has been successfully tested in Weather Research and Forecast (WRF) Model simulations and it can rectify the well-known problems in MRF, especially the overly rapid growth of the PBL. Therefore, it would be expected that the phase can be simulated somewhat more reasonably by using the YSU PBL scheme.

The other aspect that is important in West Africa but not involved in this study is the issue of land–atmosphere coupling (Koster et al. 2004). Recent studies (Taylor et al. 2011, 2013) highlight the short-time-scale feedbacks between soil moisture and rainfall with different spatial resolutions, as well as the mesoscale gradients in surface fluxes, which are important for the generation of convection in West Africa. However, the coupling between the land surface and the atmosphere is only computed on the GCM scale, which misses the finescale gradients of soil moisture and temperature. Therefore, including this coupling may influence the diurnal variability and better represent interactions between surface heterogeneity, clouds, and precipitation.

One of the reasons why convective parameterization schemes behave poorly is because CPSs are designed to represent specific mechanisms using assumptions that apply correctly over particular regions or models. However, when applied on other specific weather conditions (e.g., the monsoon systems of West Africa), the parameterization assumptions might not be accurate enough. To avoid this problem, well-resolved explicit convection is needed (Randall et al. 2003). In addition, the inconspicuous improvement of the diurnal cycle due to the increasing horizontal resolution indicates that an explicit simulation of small-scale cloud and boundary layer processes at finer horizontal scale has the potential to improve the model’s results. This can be realized by embedding the cloud-resolving models (CRMs) within each GCM grid cell, so-called superparameterization. However, it is strongly limited by the current computational capabilities. Currently, the compromise solution is in how the physical processes associated with convection should be reasonably partitioned between the parameterized part and the explicitly resolved part, which needs deep understanding of the basic issues.

Finally, motivated by the fact that the transition zone shows special characteristics of the diurnal cycle (Yu et al. 2007; Li et al. 2014) and to make our analysis more robust, the transition region in West Africa (~15°–20°N) needs to be divided into several subregions in the future to better identify how spatial patterns of the DCP will respond to changes in CPS and resolution in these subregions.

REFERENCES

  • Alpert, J. C., Kanamitsu M. , Caplan P. M. , Sela J. G. , White G. H. , and Kalnay E. , 1988: Mountain induced gravity wave drag parameterization in the NMC Medium-Range Forecast Model. Preprints, Eighth Conf. on Numerical Weather Prediction, Baltimore, MD, Amer. Meteor. Soc., 726–733.

  • Alpert, J. C., Hong S.-Y. , and Kim Y.-J. , 1996: Sensitivity of cyclogenesis to lower tropospheric enhancement of gravity wave drag using the Environmental Modeling Center Medium Range Model. Preprints, 11th Conf. on Numerical Weather Prediction, Norfolk, VA, Amer. Meteor. Soc., 322–323.

  • Arakawa, A., and Schubert W. H. , 1974: Interaction of a cumulus cloud ensemble with the large-scale environment. Part I. J. Atmos. Sci., 31, 674701, doi:10.1175/1520-0469(1974)031<0674:IOACCE>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Balling, R. C., 1985: Warm season nocturnal precipitation in the Great Plains of the United States. J. Appl. Meteor. Climatol., 24, 13831387, doi:10.1175/1520-0450(1985)024<1383:WSNPIT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Basu, B. K., 2007: Diurnal variation in precipitation over India during the summer monsoon season: Observed and model predicted. Mon. Wea. Rev., 135, 21552167, doi:10.1175/MWR3355.1.

    • Search Google Scholar
    • Export Citation
  • Betts, A., Ball J. H. , Beljaars A. C. M. , Miller M. J. , and Viterbo P. A. , 1996: The land surface–atmosphere interaction: A review based on observational and global modeling perspectives. J. Geophys. Res., 101, 72097225, doi:10.1029/95JD02135.

    • Search Google Scholar
    • Export Citation
  • Boone, A., and Coauthors, 2009: AMMA Land Surface Model Intercomparison Project Phase 2 (ALMIP2). GEWEX News, Vol. 19, No. 4, International GEWEX Project Office, Silver Spring, MD, 9–10. [Available online at http://www.gewex.org/images/Nov2009.pdf.]

  • Boyd, E., Cornforth R. J. , Lamb P. J. , Tarhule A. , Lele M. I. , and Brouder A. , 2013: Building resilience to face recurring environmental crisis in African Sahel. Nat. Climate Change, 3, 631637, doi:10.1038/nclimate1856.

    • Search Google Scholar
    • Export Citation
  • Boyle, J., and Klein S. A. , 2010: Impact of horizontal resolution on climate model forecasts of tropical precipitation and diabatic heating for the TWP-ICE period. J. Geophys. Res., 115, D23113, doi:10.1029/2010JD014262.

    • Search Google Scholar
    • Export Citation
  • Brier, G. W., 1965: Diurnal and semidiurnal atmospheric tides in relation to precipitation variations. Mon. Wea. Rev., 93, 93100, doi:10.1175/1520-0493(1965)093<0093:DASATI>2.3.CO;2.

    • Search Google Scholar
    • Export Citation
  • Cappelaere, B., and Coauthors, 2009: The AMMA-CATCH experiment in the cultivated Sahelian area of south-west Niger—Investigating water cycle response to a fluctuating climate and changing environment. J. Hydrol., 375, 3451, doi:10.1016/j.jhydrol.2009.06.021.

    • Search Google Scholar
    • Export Citation
  • Chen, H., Yu R. , Li J. , Yuan W. , and Zhou T. , 2010: Why nocturnal long-duration rainfall presents an eastward delayed diurnal phase along the Yangtze River. J. Climate, 23, 905917, doi:10.1175/2009JCLI3187.1.

    • Search Google Scholar
    • Export Citation
  • Chen, T. C., Yen M. C. , Hsieh J. C. , and Arritt R. W. , 1999: Diurnal and seasonal variations of the rainfall measured by the automatic rainfall and meteorological telemetry system in Taiwan. Bull. Amer. Meteor. Soc., 80, 22992312, doi:10.1175/1520-0477(1999)080<2299:DASVOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Chou, M.-D., and Suarez M. J. , 1994: An efficient thermal infrared radiation parameterization for use in general circulation models, NASA Tech. Memo. TM-1994-104606, Series on Global Modeling and Data Assimilation, Vol. 3, 102 pp.

  • Clark, A. J., Gallus W. A. Jr., and Chen T. C. , 2007: Comparison of the diurnal precipitation cycle in convection-resolving and non-convection-resolving mesoscale models. Mon. Wea. Rev., 135, 34563473, doi:10.1175/MWR3467.1.

    • Search Google Scholar
    • Export Citation
  • Couvreux, F., Rio C. , Guichard F. , Lothon M. , Canut G. , Bouniol D. , and Gounou A. , 2012: Initiation of daytime local convection in a semi-arid region analysed with high-resolution simulations and AMMA observations. Quart. J. Roy. Meteor. Soc., 138, 5671, doi:10.1002/qj.903.

    • Search Google Scholar
    • Export Citation
  • Dai, A., 2001: Global precipitation and thunderstorm frequencies. Part II: Diurnal variations. J. Climate, 14, 11121128, doi:10.1175/1520-0442(2001)014<1112:GPATFP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dai, A., and Trenberth K. E. , 2004: The diurnal cycle and its depiction in the Community Climate System Model. J. Climate, 17, 930951, doi:10.1175/1520-0442(2004)017<0930:TDCAID>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dai, A., Giorgi F. , and Trenberth K. E. , 1999: Observed and model simulated diurnal cycles of precipitation over the contiguous United States. J. Geophys. Res., 104, 63776402, doi:10.1029/98JD02720.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., Manning K. W. , Carbone R. E. , Trier S. B. , and Tuttle J. D. , 2003: Coherence of warm-season continental rainfall in numerical weather prediction models. Mon. Wea. Rev., 131, 26672679, doi:10.1175/1520-0493(2003)131<2667:COWCRI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • 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 advancements 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
  • Fritsch, J. M., and Carbone R. E. , 2004: Improving quantitative precipitation forecasts in the warm season: A USWRP research and development strategy. Bull. Amer. Meteor. Soc., 85, 955965, doi:10.1175/BAMS-85-7-955.

    • Search Google Scholar
    • Export Citation
  • Gallus, W. A., Jr., 1999: Eta simulations of three extreme precipitation events: Sensitivity to resolution and convective parameterization. Wea. Forecasting, 14, 405426, doi:10.1175/1520-0434(1999)014<0405:ESOTEP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Garratt, J. R., Krummel P. , and Kowalczyk E. A. , 1993: The surface-energy balance at local and regional scales—A comparison of general circulation model results with observations. J. Climate, 6, 10901109, doi:10.1175/1520-0442(1993)006<1090:TSEBAL>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., and Shields C. , 1999: Tests of precipitation parameterizations available in latest version of NCAR Regional Climate Model (RegCM) over continental United States. J. Geophys. Res., 104, 63536376, doi:10.1029/98JD01164.

    • Search Google Scholar
    • Export Citation
  • Gounou, A., 2011: The driving processes behind the diurnal cycles of the West African monsoon. Ph.D. thesis, Universite Paul Sabatier, Toulouse III, Toulouse, France, 200 pp. [Available online at http://francoise.guichard.m.free.fr/fg2cd/these_publiee_AG.pdf.]

  • Gray, W. M., and Jacobson R. W. , 1977: Diurnal variation of deep cumulus convection. Mon. Wea. Rev., 105, 11711188, doi:10.1175/1520-0493(1977)105<1171:DVODCC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hannay, C., Neale R. , and Bacmeister J. , 2012: Timeslice experiments at high resolution. What does the resolution buy us? CESM Atmosphere Model Working Group Meeting, Boulder, NCAR. [Available online at http://www.cesm.ucar.edu/working_groups/Atmosphere/Presentations/2012/hannay2.pdf.]

  • Haurwitz, B., and Cowley A. D. , 1973: The diurnal and semidiurnal barometric oscillations, global distribution, and annual variation. Pure Appl. Geophys., 102, 193222, doi:10.1007/BF00876607.

    • Search Google Scholar
    • Export Citation
  • Hodges, K. I., and Thorncroft C. C. , 1997: Distribution and statistics of African mesoscale convective weather systems based on the ISCCP Meteosat imagery. Mon. Wea. Rev., 125, 28212837, doi:10.1175/1520-0493(1997)125<2821:DASOAM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., and Pan H. L. , 1996: Nonlocal boundary layer vertical diffusion in a Medium-Range Forecast Model. Mon. Wea. Rev., 124, 23222339, doi:10.1175/1520-0493(1996)124<2322:NBLVDI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., and Leetmaa A. , 1999: An evaluation of the NCEP RSM for regional climate modeling. J. Climate, 12, 592609, doi:10.1175/1520-0442(1999)012<0592:AEOTNR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., Yign N. , and Dudhia J. , 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 23182341, doi:10.1175/MWR3199.1.

    • Search Google Scholar
    • Export Citation
  • Huang, W.-R., and Chan J. C. L. , 2011: Maintenance mechanisms for the early-morning maximum summer rainfall over southeast China. Quart. J. Roy. Meteor. Soc., 137, 959968, doi:10.1002/qj.815.

    • Search Google Scholar
    • Export Citation
  • Huffman, G. J., and Coauthors, 2007: The TRMM Multisatellite Precipitation Analysis (TMPA): Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J. Hydrometeor., 8, 3855, doi:10.1175/JHM560.1.

    • Search Google Scholar
    • Export Citation
  • Juang, H.-M., and Kanamitsu M. , 1994: The NMC nested regional spectral model. Mon. Wea. Rev., 122, 326, doi:10.1175/1520-0493(1994)122<0003:TNNRSM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Juang, H.-M., Hong S.-Y. , and Kanamitsu M. , 1997: The NCEP Regional Spectral Model: An update. Bull. Amer. Meteor. Soc., 78, 21252143, doi:10.1175/1520-0477(1997)078<2125:TNRSMA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kain, J. S., 2004: The Kain–Fritsch convective parameterization: An update. J. Appl. Meteor. Climatol., 43, 170181, doi:10.1175/1520-0450(2004)043<0170:TKCPAU>2.0.CO;2.

    • 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
  • Kang, H.-S., and Hong S.-Y. , 2008: Sensitivity of the simulated East Asian summer monsoon climatology to four convective parameterization schemes. J. Geophys. Res., 113, D15119, doi:10.1029/2007JD009692.

    • Search Google Scholar
    • Export Citation
  • Kim, J., and Mahrt L. , 1992: Momentum transport by gravity waves. J. Atmos. Sci., 49, 735748, doi:10.1175/1520-0469(1992)049<0735:MTBGW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Koo, M.-S., and Hong S.-Y. , 2010: Diurnal variations of simulated precipitation over East Asia in two regional climate models. J. Geophys. Res., 115, D05105, doi:10.1029/2009JD012574.

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

    • Search Google Scholar
    • Export Citation
  • Laing, A. G., Carbone R. , Levizzani V. , and Tuttle J. , 2008: The propagation and diurnal cycles of deep convection in northern tropical Africa. Quart. J. Roy. Meteor. Soc., 134, 93109, doi:10.1002/qj.194.

    • Search Google Scholar
    • Export Citation
  • Lebel, T., and Coauthors, 2009: AMMA-CATCH studies in the Sahelian region of West-Africa: An overview. J. Hydrol., 375, 313, doi:10.1016/j.jhydrol.2009.03.020.

    • Search Google Scholar
    • Export Citation
  • Lee, M., and Coauthors, 2007: Sensitivity to horizontal resolution in the AGCM simulations of warm season diurnal cycle of precipitation over the United States and northern Mexico. J. Climate, 20, 18621881, doi:10.1175/JCLI4090.1.

    • Search Google Scholar
    • Export Citation
  • Li, Z., Yang D. , Hong Y. , Zhang J. , and Qi Y. , 2014: Characterizing spatiotemporal variations of hourly rainfall by gauge and radar in the mountainous Three Gorges region. J. Appl. Meteor. Climatol., 53, 873889, doi:10.1175/JAMC-D-13-0277.1.

    • Search Google Scholar
    • Export Citation
  • Lin, X., Randall D. A. , and Fowler L. D. , 2000: Diurnal variability of the hydrological cycle and radiative fluxes: Comparisons between observations and a GCM. J. Climate, 13, 41594179, doi:10.1175/1520-0442(2000)013<4159:DVOTHC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lin, Y., and Coauthors, 2012: TWP-ICE global atmospheric model intercomparison: Convection responsiveness and resolution impact. J. Geophys. Res.,117, D09111, doi:10.1029/2011JD017018.

  • McGarry, M. M., and Reed R. J. , 1978: Diurnal variations in convective activity and precipitation during phases II and III of GATE. Mon. Wea. Rev., 106, 101113, doi:10.1175/1520-0493(1978)106<0101:DVICAA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Mohr, K. I., 2004: Interannual, monthly, and regional variability in the wet season diurnal cycle of precipitation in sub-Saharan Africa. J. Climate, 17, 24412453, doi:10.1175/1520-0442(2004)017<2441:IMARVI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Moorthi, S., and Suarez M. J. , 1992: Relaxed Arakawa–Schubert: A parameterization of moist convection for general circulation models. Mon. Wea. Rev., 120, 9781002, doi:10.1175/1520-0493(1992)120<0978:RASAPO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Nikulin, G., and Coauthors, 2012: Precipitation climatology in an ensemble of CORDEX-Africa regional climate simulations. J. Climate, 25, 60576078, doi:10.1175/JCLI-D-11-00375.1.

    • Search Google Scholar
    • Export Citation
  • Oki, T., and Kanae S. , 2006: Global hydrological cycles and world water resources. Science, 313, 10681072, doi:10.1126/science.1128845.

    • Search Google Scholar
    • Export Citation
  • Pan, H.-L., and Wu W.-S. , 1995: Implementing a mass flux convective parameterization package for the NMC Medium-Range Forecast Model. NMC Office Note 409, Camp Springs, MD, 40 pp. [Available online at http://www.lib.ncep.noaa.gov/ncepofficenotes/files/01408A42.pdf.]

  • Parker, D. J., and Coauthors, 2005: The diurnal cycle of West African monsoon circulation. Quart. J. Roy. Meteor. Soc., 131, 28392860, doi:10.1256/qj.04.52.

    • Search Google Scholar
    • Export Citation
  • Pu, B., and Cook K. H. , 2010: Dynamics of the West African westerly jet. J. Climate, 23, 62636276, doi:10.1175/2010JCLI3648.1.

  • Ramage, C. S., 1952: Diurnal variation of summer rainfall over east China, Korea and Japan. J. Atmos. Sci., 9, 8386, doi:10.1175/1520-0469(1952)009<0083:DVOSRO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ramage, C. S., 1971: Monsoon Meteorology. Academic Press, 295 pp.

  • Randall, D. A., Harshvardhan, and Dazlich D. A. , 1991: Diurnal variability of the hydrological cycle in a general circulation model. J. Atmos. Sci., 48, 4062, doi:10.1175/1520-0469(1991)048<0040:DVOTHC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Randall, D. A., Khairoutdinov M. , and Grabowski W. , 2003: Breaking the cloud parameterization deadlock. Bull. Amer. Meteor. Soc., 84, 15471564, doi:10.1175/BAMS-84-11-1547.

    • Search Google Scholar
    • Export Citation
  • Ridout, J. A., Jin Y. , and Liou C.-S. , 2005: A cloud-base quasi-balance constraint for parameterized convection: Application to the Kain–Fritsch cumulus scheme. Mon. Wea. Rev., 133, 33153334, doi:10.1175/MWR3034.1.

    • Search Google Scholar
    • Export Citation
  • Rio, C., Hourdin F. , Grandpeix J.-Y. , and Lafore J.-P. , 2009: Shifting the diurnal cycle of parameterized deep convection over land. Geophys. Res. Lett., 36, L07809, doi:10.1029/2008GL036779.

    • Search Google Scholar
    • Export Citation
  • Roca, R., Chambon P. , Jobard I. , Kirstetter P. , Gosset M. , and Berges J. C. , 2010: Comparing satellite and surface rainfall products over West Africa at meteorologically relevant scales during the AMMA campaign using error estimates. J. Appl. Meteor. Climatol., 49, 715731, doi:10.1175/2009JAMC2318.1.

    • 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
  • Schwartz, B. E., and Bosart L. F. , 1979: The diurnal variability of Florida rainfall. Mon. Wea. Rev., 107, 15351545, doi:10.1175/1520-0493(1979)107<1535:TDVOFR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Shine, D. W., Cocke S. , and LaRow T. E. , 2007: Diurnal cycle of precipitation in a climate model. J. Geophys. Res., 112, D13109, doi:10.1029/2006JD008333.

    • Search Google Scholar
    • Export Citation
  • Slingo, A., Wilderspin R. , and Brentnall S. , 1987: Simulation of the diurnal cycle of outgoing longwave radiation with an atmospheric GCM. Mon. Wea. Rev., 115, 14511457, doi:10.1175/1520-0493(1987)115<1451:SOTDCO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Sperber, K. R., and Yasunari T. , 2006: Workshop on monsoon climate systems: Toward better prediction of the monsoon. Bull. Amer. Meteor. Soc., 87, 13391403, doi:10.1175/BAMS-87-10-1399.

    • Search Google Scholar
    • Export Citation
  • Taylor, C. M., and Clark D. B. , 2001: The diurnal cycle and African easterly waves: A land surface perspective. Quart. J. Roy. Meteor. Soc., 127, 845867, doi:10.1002/qj.49712757308.

    • Search Google Scholar
    • Export Citation
  • Taylor, C. M., Gounou A. , Guichard F. , Harris P. P. , Ellis R. J. , Couvreux F. , and Kauwe M. D. , 2011: Frequency of Sahelian storm initiation enhanced over mesoscale soil-moisture patterns. Nat. Geosci., 4, 430433, doi:10.1038/ngeo1173.

    • Search Google Scholar
    • Export Citation
  • Taylor, C. M., Birch C. E. , Parker D. J. , Dixon N. , Guichard F. , Nikulin G. , and Lister G. M. S. , 2013: Modeling soil moisture–precipitation feedback in the Sahel: Importance of spatial scale versus convective parameterization. Geophys. Res. Lett., 40, 62136218, doi:10.1002/2013GL058511.

    • Search Google Scholar
    • Export Citation
  • Taylor, K. E., 2001: Summarizing multiple aspects of model performance in a single diagram. J. Geophys. Res., 106, 71837192, doi:10.1029/2000JD900719.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., Dai A. , Rasmussen R. M. , and Parsons D. B. P. , 2003: The changing character of precipitation. Bull. Amer. Meteor. Soc., 84, 12051217, doi:10.1175/BAMS-84-9-1205.

    • Search Google Scholar
    • Export Citation
  • Vischel, T., Quantin G. , Lebel T. , Viarre J. , Gosset M. , Cazenave F. , and Panthou G. , 2011: Generation of high-resolution rain fields in West Africa: Evaluation of dynamic interpolation methods. J. Hydrometeor., 12, 14651482, doi:10.1175/JHM-D-10-05015.1.

    • Search Google Scholar
    • Export Citation
  • Wai, M. M.-K., Welsh P. T. , and Ma W.-M. , 1996: Interaction of secondary circulations with the summer monsoon and diurnal rainfall over Hong Kong. Bound.-Layer Meteor., 81, 123146, doi:10.1007/BF00119062.

    • Search Google Scholar
    • Export Citation
  • Wallace, J. M., 1975: Diurnal variations in precipitation and thunderstorm frequency over the conterminous United States. Mon. Wea. Rev., 103, 406419, doi:10.1175/1520-0493(1975)103<0406:DVIPAT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wang, W., and Seaman N. L. , 1997: A comparison study of convective parameterization schemes in a mesoscale model. Mon. Wea. Rev., 125, 252278, doi:10.1175/1520-0493(1997)125<0252:ACSOCP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wilks, D. S., 2005: Statistical Methods in the Atmospheric Sciences. 2nd ed. Academic Press, 627 pp.

  • Yang, G., and Slingo J. , 2001: The diurnal cycle in the tropics. Mon. Wea. Rev., 129, 784801, doi:10.1175/1520-0493(2001)129<0784:TDCITT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Yoshimura, K., and Kanamitsu M. , 2008: Dynamical global downscaling of global reanalysis. Mon. Wea. Rev., 136, 29832998, doi:10.1175/2008MWR2281.1.

    • Search Google Scholar
    • Export Citation
  • Yoshimura, K., Kanamitsu M. , and Dettinger M. , 2010: Regional downscaling for stable water isotopes: A case study of an atmospheric river event. J. Geophys. Res.,115, D18114, doi:10.1029/2010JD014032.

  • Yu, R., Zhou T. , Xiong A. , Zhu Y. , and Li J. , 2007: Diurnal variations of summer precipitation over contiguous China. Geophys. Res. Lett., 34, L01704, doi:10.1029/2006GL028129.

    • Search Google Scholar
    • Export Citation
  • Yuan, W., Wu R. , Zhang M. , Lin W. , Li J. , and Fu Y. , 2013: Diurnal cycle of summer precipitation over subtropical East Asia in CAM5. J. Climate, 26, 31593172, doi:10.1175/JCLI-D-12-00119.1.

    • Search Google Scholar
    • Export Citation
Save