• Adams, T., and Ostrowski J. , 2010: Short lead-time hydrologic ensemble forecasts from numerical weather prediction model ensembles. Proc. World Environmental and Water Resources Congress, Providence, RI, American Society of Civil Engineers, 2294–2304.

    • Search Google Scholar
    • Export Citation
  • Anctil, F., Perrin C. , and Andréassian V. , 2003: ANN output updating of lumped conceptual rainfall/runoff forecasting models. J. Amer. Water Resour. Assoc., 39, 12691279.

    • Search Google Scholar
    • Export Citation
  • Arduino, G., Reggiani P. , and Todini E. , 2005: Recent advances in flood forecasting and flood risk assessment. Hydrol. Earth Syst. Sci., 9, 280284.

    • Search Google Scholar
    • Export Citation
  • Bartholmes, J., and Todini E. , 2005: Coupling meteorological and hydrological models for flood forecasting. Hydrol. Earth Syst. Sci., 9, 333346.

    • Search Google Scholar
    • Export Citation
  • Berthet, L., Andreassian V. , Perrin C. , and Javelle P. , 2009: How crucial is it to account for the antecedent moisture conditions in flood forecasting? Comparison of event-based and continuous approaches on 178 catchments. Hydrol. Earth Syst. Sci., 13, 819831.

    • Search Google Scholar
    • Export Citation
  • Bogner, K., and Kalas M. , 2008: Erro-correction methods and evaluation of an ensemble based hydrological forecasting system for the Upper Danube catchment. Atmos. Sci. Lett., 9, 95102.

    • Search Google Scholar
    • Export Citation
  • Bohn, T. J., Sonessa M. Y. , and Lettenmaier D. P. , 2010: Seasonal hydrologic forecasting: Do multimodel ensemble averages always yield improvements in forecast skill? J. Hydrometeor., 11, 13581372.

    • Search Google Scholar
    • Export Citation
  • Bougeault, P., and Coauthors, 2010: The THORPEX Interactive Grand Global Ensemble (TIGGE). Bull. Amer. Meteor. Soc., 91, 10591072.

  • Boughton, W., 2005: Catchment water balance modelling in Australia 1960–2004. Agric. Water Manage., 71, 91116.

  • Bowler, N. E., Pierce C. E. , and Seed A. W. , 2006: STEPS: A probabilistic precipitation forecasting scheme which merges an extrapolation nowcast with downscaled NWP. Quart. J. Roy. Meteor. Soc., 132, 21272155.

    • Search Google Scholar
    • Export Citation
  • Bright, D. R., and Mullen S. L. , 2002: Short-range ensemble forecasts of precipitation during the Southwest monsoon. Wea. Forecasting, 17, 10801100.

    • Search Google Scholar
    • Export Citation
  • Buizza, R., 2008: The value of probabilistic prediction. Atmos. Sci. Lett., 9, 3642.

  • Buizza, R., Miller M. , and Palmer T. N. , 1999: Stochastic representation of model uncertainties in the ECMWF Ensemble Prediction System. Quart. J. Roy. Meteor. Soc., 125, 28872908.

    • Search Google Scholar
    • Export Citation
  • Carpenter, T. M., and Georgakakos K. P. , 2004: Impacts of parametric and radar rainfall uncertainty on the ensemble streamflow simulations of a distributed hydrologic model. J. Hydrol., 298, 202221.

    • Search Google Scholar
    • Export Citation
  • Carpenter, T. M., and Georgakakos K. P. , 2006: Discretization scale dependencies of the ensemble flow range versus catchment area relationship in distributed hydrologic modelling. J. Hydrol., 328, 242257.

    • Search Google Scholar
    • Export Citation
  • Catchlove, R., Leahy C. , Seed A. , and Malone T. , 2005: An investigation into the use of real-time radar rainfall and comparison with point gauge estimates for flood forecasting. Proc. 29th Hydrology and Water Resources Symp., Canberra, Australia, Engineers Australia, 633–640.

    • Search Google Scholar
    • Export Citation
  • Cerlini, P. B., Emanuel K. A. , and Todini E. , 2005: Orographic effects on convective precipitation and space–time rainfall variability: Preliminary results. Hydrol. Earth Syst. Sci., 9, 285299.

    • Search Google Scholar
    • Export Citation
  • Clark, M. P., and Hay L. E. , 2004: Use of medium-range numerical weather prediction model output to produce forecasts of streamflow. J. Hydrometeor., 5, 1532.

    • Search Google Scholar
    • Export Citation
  • Cloke, H. I., and Pappenberger F. , 2009: Ensemble flood forecast: A review. J. Hydrol., 375, 613626.

  • Collier, C. G., 2009: On the propagation of uncertainty in weather radar estimates of rainfall through hydrological models. Meteor. Appl., 16, 3540.

    • Search Google Scholar
    • Export Citation
  • Collier, C. G., and Robbins G. L. , 2008: Using stochastic state-space modelling to specify and reduce hydrograph error. Hydrol. Processes, 22, 11731179.

    • Search Google Scholar
    • Export Citation
  • Cornford, D., 2004: A Bayesian state space modelling approach to probabilistic quantitative precipitation forecasting. J. Hydrol., 288, 92104.

    • Search Google Scholar
    • Export Citation
  • Croley, T. E., II, and Hartmann H. C. , 1987: Near real-time forecasting of large lake supplies. J. Water Resour. Plann. Manage., 113, 810823.

    • Search Google Scholar
    • Export Citation
  • Damrath, U., Doms G. , Frühwald D. , Heise E. , Richter B. , and Steppeler J. , 2000: Operational quantitative precipitation forecasting at the German Weather Service. J. Hydrol., 239, 260285.

    • Search Google Scholar
    • Export Citation
  • Davies, T., Cullen M. J. P. , Malcolm A. J. , Mawson M. H. , Staniforth A. , White A. A. , and Wood N. , 2005: A new dynamical core for the Met Office’s global and regional modelling of the atmosphere. Quart. J. Roy. Meteor. Soc., 131, 17591782.

    • Search Google Scholar
    • Export Citation
  • Davolio, S., Miglietta M. M. , Diomede T. , Marsigli C. , Morgillo A. , and Moscatello A. , 2008: A meteo-hydrological prediction system based on a multi-model approach for precipitation forecasting. Nat. Hazards Earth Syst. Sci., 8, 143159.

    • Search Google Scholar
    • Export Citation
  • Day, G. N., 1985: Extended streamflow forecasting using NWSRFS. J. Water Resour. Plann. Manage., 111, 157170.

  • Demargne, J., Mullusky M. , Werner K. , Adams T. , Lindsey S. , Schwein N. , Marosi W. , and Welles E. , 2009: Application of forecast verification science to operational river forecasting in the U.S. National Weather Service. Bull. Amer. Meteor. Soc., 90, 779784.

    • Search Google Scholar
    • Export Citation
  • Demeritt, D., Cloke H. , Pappenberger F. , Thielen J. , Bartholmes J. , and Ramos M.-H. , 2007: Ensemble predictions and perceptions of risk, uncertainty, and error in flood forecasting. Environ. Hazards, 7, 115127.

    • Search Google Scholar
    • Export Citation
  • de Roo, A. P. J., Wesseling C. G. , and Van Deursen W. P. A. , 2000: Physically based river basin modelling within a GIS: The LISFLOOD model. Hydrol. Processes, 14, 19811992.

    • Search Google Scholar
    • Export Citation
  • Diomede, T., and Coauthors, 2008: Discharge prediction based on multi-model precipitation forecasts. Meteor. Atmos. Phys., 101, 245265.

    • Search Google Scholar
    • Export Citation
  • Ebert, E. E., and McBride J. L. , 2000: Verification of precipitation in weather systems: Determination of systematic errors. J. Hydrol., 239, 179202.

    • Search Google Scholar
    • Export Citation
  • Ebert, E. E., Damrath U. , Wergen W. , and Baldwin M. E. , 2003: The WGNE assessment of short-term quantitative precipitation forecasts. Bull. Amer. Meteor. Soc., 84, 481492.

    • Search Google Scholar
    • Export Citation
  • Ebert, E. E., Janowiak J. E. , and Kidd C. , 2007: Comparison of near-real-time precipitation estimates from satellite observations and numerical models. Bull. Amer. Meteor. Soc., 88, 4764.

    • Search Google Scholar
    • Export Citation
  • Friederichs, P., and Hense A. , 2007: Statistical downscaling of extreme precipitation events using censored quantile regression. Mon. Wea. Rev., 135, 23652378.

    • Search Google Scholar
    • Export Citation
  • Ganguly, A. R., and Bras R. L. , 2003: Distributed quantitative precipitation forecasting using information from radar and numerical weather prediction models. J. Hydrometeor., 4, 11681180.

    • Search Google Scholar
    • Export Citation
  • Gebhardt, C., Theis S. , Krahe P. , and Renner V. , 2008: Experimental ensemble forecasts of precipitation based on a convection-resolving model. Atmos. Sci. Lett., 9, 6772.

    • Search Google Scholar
    • Export Citation
  • Georgakakos, K. P., and Hudlow M. D. , 1984: Quantitative precipitation forecast techniques for use in hydrologic forecasting. Bull. Amer. Meteor. Soc., 65, 11861200.

    • Search Google Scholar
    • Export Citation
  • Germann, U., Berenguer M. , Sempere-Torres D. , and Zappa M. , 2009: REAL-Ensemble radar precipitation estimation for hydrology in a mountainous region. Quart. J. Roy. Meteor. Soc., 135, 445456.

    • Search Google Scholar
    • Export Citation
  • Ghile, Y. B., and Schulze R. E. , 2010: Evaluation of three numerical weather prediction models for short and medium range agrohydrological applications. Water Resour. Manage., 24, 10051028.

    • Search Google Scholar
    • Export Citation
  • Golding, B. W., 2000: Quantitative precipitation forecasting in the UK. J. Hydrol., 239, 286305.

  • Golding, B. W., 2009: Long lead time flood warnings: Reality or fantasy? Meteor. Appl., 16, 312.

  • Gouweleeuw, B. T., Thielen J. , Franchello G. , de Roo A. P. J. , and Buizza R. , 2005: Flood forecasting using medium-range probabilistic weather prediction. Hydrol. Earth Syst. Sci., 9, 365380.

    • Search Google Scholar
    • Export Citation
  • Habets, F., LeMoigne P. , and Noilhan J. , 2004: On the utility of operational precipitation forecasts to served as input for streamflow forecasting. J. Hydrol., 293, 270288.

    • Search Google Scholar
    • Export Citation
  • Hagedorn, R., Doblas-Reyes F. J. , and Palmer T. N. , 2005: The rationale behind the success of multi-model ensembles in seasonal forecasting. I. Basic concept. Tellus, 57A, 219233.

    • Search Google Scholar
    • Export Citation
  • Hapuarachchi, P., Wang Q. J. , and Pagano T. C. , 2011: A review of advances in flash flood forecasting. Hydrol. Processes, doi:10.1002/hyp.8040, in press.

    • Search Google Scholar
    • Export Citation
  • Hartmann, H. C., Pagano T. C. , Sorooshian S. , and Bales R. , 2002: Confidence builders: Evaluating seasonal climate forecasts from user perspectives. Bull. Amer. Meteor. Soc., 83, 683698.

    • Search Google Scholar
    • Export Citation
  • Hay, L. E., and Clark M. P. , 2003: Use of statistically and dynamically downscaled atmospheric model output for hydrologic simulations in three mountainous basins in the western United States. J. Hydrol., 282, 5675.

    • Search Google Scholar
    • Export Citation
  • He, Y., Wetterhall F. , Cloke H. L. , Pappenberger F. , Wilson M. , Freer J. , and McGregor G. , 2009: Tracking the uncertainty in flood alerts driven by grand ensemble weather predictions. Meteor. Appl., 16, 91101.

    • Search Google Scholar
    • Export Citation
  • Hou, D., Mitchell K. , Toth Z. , Lohmann D. , and Wei H. , 2009: The effects of large-scale atmospheric uncertainty on streamflow predictability. J. Hydrometeor., 10, 717733.

    • Search Google Scholar
    • Export Citation
  • Houtekamer, P. L., Lefaivre L. , Derome J. , Ritchie H. , and Mitchell H. L. , 1996: A system simulation approach to ensemble prediction. Mon. Wea. Rev., 124, 12251242.

    • Search Google Scholar
    • Export Citation
  • Jasper, K., Gurtz J. , and Lang H. , 2002: Advanced flood forecasting in Alpine watersheds by coupling meteorological observations and forecasts with a distributed hydrological model. J. Hydrol., 267, 4052.

    • Search Google Scholar
    • Export Citation
  • Jaun, S., and Ahrens B. , 2009: Evaluation of a probabilistic hydrometeorological forecast system. Hydrol. Earth Syst. Sci., 13, 10311043.

    • Search Google Scholar
    • Export Citation
  • Kaufmann, P., Schubiger F. , and Binder P. , 2003: Precipitation forecasting by a mesoscale numerical weather prediction (NWP) model: Eight years of experience. Hydrol. Earth Syst. Sci., 7, 812832.

    • Search Google Scholar
    • Export Citation
  • Klein, W. H., and Glahn H. R. , 1974: Forecasting local weather by means of model output statistics. Bull. Amer. Meteor. Soc., 55, 12171227.

    • Search Google Scholar
    • Export Citation
  • Kobold, M., and Sušelj K. , 2005: Precipitation forecasts and their uncertainty as input into hydrological models. Hydrol. Earth Syst. Sci., 9, 322332.

    • Search Google Scholar
    • Export Citation
  • Koh, T.-Y., and Ng J. S. , 2009: Improved diagnostics for NWP verification in the tropics. J. Geophys. Res., 114, D12102, doi:10.1029/2008JD011179.

    • Search Google Scholar
    • Export Citation
  • Komma, J., Reszler C. , Blöschl G. , and Haiden T. , 2007: Ensemble prediction of floods—Catchment non-linearity and forecast probability. Nat. Hazards Earth Syst. Sci., 7, 431444.

    • Search Google Scholar
    • Export Citation
  • Krishnamurti, T. N., Kistawal C. M. , LaRow T. E. , Bachiochi D. R. , Zhang Z. , Williford C. E. , Gadgil S. , and Surendran S. , 1999: Improved weather and seasonal climate forecasts from multimodel superensemble. Science, 285, 15481550.

    • Search Google Scholar
    • Export Citation
  • Krzysztofowicz, R., 1999: Bayesian theory of probabilistic forecasting via deterministic hydrologic model. Water Resour. Res., 35, 27392750.

    • Search Google Scholar
    • Export Citation
  • Krzysztofowicz, R., 2002: Bayesian system for probabilistic river stage forecasting. J. Hydrol., 268, 1640.

  • Landman, W. A., Mason S. J. , Tyson P. D. , and Tennant W. J. , 2001: Statistical downscaling of GCM simulations to streamflow. J. Hydrol., 252, 221236.

    • Search Google Scholar
    • Export Citation
  • Lindström, G., Johansson B. , Persson M. , Gardelin M. , and Bergström S. , 1997: Development and test of the distributed HBV-96 hydrological model. J. Hydrol., 201, 272288.

    • Search Google Scholar
    • Export Citation
  • Lorenz, E., 1969: The predictability of a flow which contains many scales of motion. Tellus, 21A, 289307.

  • Lu, C., Yuan H. , Tollerud E. I. , and Wang N. , 2010: Scale-dependent uncertainties in global QPFs and QPEs from NWP model and satellite fields. J. Hydrometeor., 11, 139155.

    • Search Google Scholar
    • Export Citation
  • Madsen, H., and Skotner C. , 2005: Adaptive state updating in real-time river flow forecasting—A combined filtering and error forecasting procedure. J. Hydrol., 308, 302312.

    • Search Google Scholar
    • Export Citation
  • Mascaro, G., Vivoni E. R. , and Deidda R. , 2010: Implications of ensemble quantitative precipitation forecast errors on distributed streamflow forecasting. J. Hydrometeor., 11, 6986.

    • Search Google Scholar
    • Export Citation
  • McGregor, J. L., 2005: C-CAM: Geometric aspects and dynamical formulation. CSIRO Tech. Rep. 70, 41 pp.

  • Mittermaier, M. P., 2007: Improving short-range high-resolution model precipitation forecast skill using time-lagged ensembles. Quart. J. Roy. Meteor. Soc., 133, 14871500.

    • Search Google Scholar
    • Export Citation
  • Moore, R. J., 2007: The PDM rainfall-runoff model. Hydrol. Earth Syst. Sci., 11, 483499.

  • Moore, R. J., and Bell V. A. , 2001: Comparison of rainfall-runoff models for flood forecasting. Part 1: Literature review of models. Environment Agency R&D Tech. Rep. W241, 94 pp.

    • Search Google Scholar
    • Export Citation
  • Nijssen, B., and Coauthors, 2003: Simulation of high latitude hydrological processes in the Torne–Kalix basin: PILPS Phase 2(e): 2: Comparison of model results with observations. Global Planet. Change, 38, 3153.

    • Search Google Scholar
    • Export Citation
  • Olson, D. A., Junker N. W. , and Korty B. , 1995: Evaluation of 33 years of quantitative precipitation forecasting at the NMC. Wea. Forecasting, 10, 498511.

    • Search Google Scholar
    • Export Citation
  • Olsson, J., and Lindström G. , 2008: Evaluation and calibration of operational hydrological ensemble forecasts in Sweden. J. Hydrol., 350, 1424.

    • Search Google Scholar
    • Export Citation
  • Oudin, L., Perrin C. , Mathevet T. , Andréassian V. , and Michel C. , 2006: Impact of biased and randomly corrupted inputs on the efficiency and the parameters of watershed models. J. Hydrol., 320, 6283.

    • Search Google Scholar
    • Export Citation
  • Pagano, T. C., Garen D. , and Sorooshian S. , 2004: Evaluation of official western U.S. seasonal water supply outlooks, 1922–2002. J. Hydrometeor., 5, 896909.

    • Search Google Scholar
    • Export Citation
  • Pagano, T. C., Hapuarachchi P. , and Wang Q. J. , 2009a: Continuous soil moisture accounting and routing modelling to support short lead-time streamflow forecasting. CSIRO Tech. Rep., 53 pp.

    • Search Google Scholar
    • Export Citation
  • Pagano, T. C., Hapuarachchi P. , and Wang Q. J. , 2009b: Development and testing of a multi-model rainfall-runoff streamflow forecasting application. CSIRO Tech. Rep., 40 pp.

    • Search Google Scholar
    • Export Citation
  • Pagano, T. C., Hapuarachchi P. , and Wang Q. J. , 2010: Continuous rainfall-runoff model comparison and short-term daily streamflow forecast skill evaluation. CSIRO Tech. Rep., 70 pp.

    • Search Google Scholar
    • Export Citation
  • Palmer, T. N., Barkmeijer J. , Buizza R. , and Petroliagis T. , 1997: The ECMWF ensemble prediction system. Meteor. Appl., 4, 301304.

  • Pappenberger, F., Beve K. J. , Hunter N. M. , Bates P. D. , Gouweleeuw B. T. , Thielen J. , and de Roo A. P. J. , 2005: Cascading model uncertainty from medium range weather forecasts (10 days) through a rainfall-runoff model to flood inundation predictions within the European Flood Forecasting System (EFFS). Hydrol. Earth Syst. Sci., 9, 381393.

    • Search Google Scholar
    • Export Citation
  • Pappenberger, F., Bartholmes J. , Thielen J. , Cloke H. L. , Buizza R. , and de Roo A. , 2008: New dimensions in early flood warning across the globe using grand-ensemble weather predictions. Geophys. Res. Lett., 35, L10404, doi:10.1029/2008GL033837.

    • Search Google Scholar
    • Export Citation
  • Park, Y.-Y., Buizza R. , and Leutbecher M. , 2008: TIGGE: Preliminary results on comparing and combining ensembles. Quart. J. Roy. Meteor. Soc., 134, 20292050.

    • Search Google Scholar
    • Export Citation
  • Perrin, C., Michel C. , and Andréassian V. , 2004: Improvement of a parsimonious model for streamflow simulation. J. Hydrol., 279, 275289.

    • Search Google Scholar
    • Export Citation
  • Rabuffetti, D., and Barbero S. , 2005: Operational hydro-meteorological warning and real-time flood forecasting: The Piemonte Region case study. Hydrol. Earth Syst. Sci., 9, 457466.

    • Search Google Scholar
    • Export Citation
  • Richard, E., Cosma S. , Benoit R. , Binder P. , Buzzi A. , and Kaufmann P. , 2003: Intercomparison of mesoscale meteorological models for precipitation forecasting. Hydrol. Earth Syst. Sci., 7, 799811.

    • Search Google Scholar
    • Export Citation
  • Roberts, N. M., Cole S. J. , Forbes R. M. , Moore R. J. , and Boswell D. , 2009: Use of high-resolution NWP rainfall and river flow forecasts for advance warning of the Carlisle flood, north-west England. Meteor. Appl., 16, 2334.

    • Search Google Scholar
    • Export Citation
  • Roulin, E., 2007: Skill and relative economic value of medium-range hydrological ensemble predictions. Hydrol. Earth Syst. Sci., 11, 725737.

    • Search Google Scholar
    • Export Citation
  • Roulin, E., and Vannitsem S. , 2005: Skill of medium-range hydrological ensemble predictions. J. Hydrometeor., 6, 729744.

  • Sattler, K., and Feddersen H. , 2005: Limited-area short-range ensemble predictions targeted for heavy rain in Europe. Hydrol. Earth Syst. Sci., 9, 300312.

    • Search Google Scholar
    • Export Citation
  • Schaake, J., 1978: The National Weather Service extended streamflow prediction techniques: Description and applications during 1977. Proc. Third Annual Climate Diagnostics Workshop, Miami, FL, National Oceanic and Atmospheric Administration, 31.1–31.19.

    • Search Google Scholar
    • Export Citation
  • Schaake, J., Franz K. , Bradley A. , and Buizza R. , 2006: The Hydrologic Ensemble Prediction Experiment (HEPEX). Hydrol. Earth Syst. Sci. Discuss., 3, 33213332.

    • Search Google Scholar
    • Export Citation
  • Schaake, J., Hamill T. M. , Buizza R. , and Clark M. , 2007: HEPEX: The Hydrological Ensemble Prediction Experiment. Bull. Amer. Meteor. Soc., 88, 15411547.

    • Search Google Scholar
    • Export Citation
  • Schulze, R. E., 2000: Transcending scales of space and time in impact studies of climate and climate change on agrohydrological responses. Agric. Ecosyst. Environ., 82, 185212.

    • Search Google Scholar
    • Export Citation
  • Sene, K., 2010: Hydrometeorology: Forecasting and Applications. Springer, 355 pp.

  • Seo, D.-J., Koren V. , and Cajina N. , 2003: Real-time variational assimilation of hydrologic and hydrometeorological data into operational hydrologic forecasting. J. Hydrometeor., 4, 627641.

    • Search Google Scholar
    • Export Citation
  • Shi, X., Wood A. W. , and Lettenmaier D. P. , 2008: How essential is hydrological model calibration to seasonal streamflow forecasting? J. Hydrometeor., 9, 13501363.

    • Search Google Scholar
    • Export Citation
  • Simmons, A. J., and Hollingsworth A. , 2002: Some aspects of the improvement in skill of numerical weather prediction. Quart. J. Roy. Meteor. Soc., 128, 647677.

    • Search Google Scholar
    • Export Citation
  • Smith, M. B., Georgakakos K. P. , and Liang X. , 2004: The distributed model intercomparison project (DMIP). J. Hydrol., 298, 13.

  • Srikanthan, R., Elliott J. F. , and Adams G. , 1994: A review of real-time flood forecasting methods. Cooperative Research Center for Catchment Hydrology Tech. Rep. 94/2, 120 pp.

    • Search Google Scholar
    • Export Citation
  • Stensrud, D. J., Brooks H. E. , Du J. , Tracton M. S. , and Rogers E. , 1999: Using ensembles for short-range forecasting. Mon. Wea. Rev., 127, 433446.

    • Search Google Scholar
    • Export Citation
  • Stensrud, D. J., Bao J.-W. , and Warner T. T. , 2000: Using initial condition and model physics perturbations in short-range ensemble simulations of mesoscale convective systems. Mon. Wea. Rev., 128, 20772107.

    • Search Google Scholar
    • Export Citation
  • Thielen, J., Schaake J. , Hartman R. , and Buizza R. , 2008: Aims, challenges and progress of the Hydrological Ensemble Prediction Experiment (HEPEX) following the third HEPEX workshop held in Stresa 27 to 29 June 2007. Atmos. Sci. Lett., 9, 2935.

    • Search Google Scholar
    • Export Citation
  • Thielen, J., Bartholmes J. , Ramos M.-H. , and de Roo A. , 2009: The European Flood Alert System—Part 1: Concept and development. Hydrol. Earth Syst. Sci., 13, 125140.

    • Search Google Scholar
    • Export Citation
  • Thirel, G., Rousset-Regimbeau F. , Martin E. , and Habets F. , 2008: On the impact of short-range meteorological forecasts for ensemble streamflow predictions. J. Hydrometeor., 9, 13011317.

    • Search Google Scholar
    • Export Citation
  • Toth, Z., and Kalnay E. , 1993: Ensemble forecasting at NMC: The generation of perturbations. Bull. Amer. Meteor. Soc., 74, 23172330.

  • Toth, Z., and Kalnay E. , 1997: Ensemble forecasting at NCEP and the breeding method. Mon. Wea. Rev., 125, 32973319.

  • Trenberth, K. E., Ed., 1992: Climate System Modeling. Cambridge University Press, 820 pp.

  • Tucci, C. E. M., Coolischonn W. , Clarke R. T. , Pax A. R. , and Allasia D. , 2008: Short- and long-term flow forecasting in the Rio Grande watershed (Brazil). Atmos. Sci. Lett., 9, 5356.

    • Search Google Scholar
    • Export Citation
  • van Andel, S. J., Lobbrecht A. H. , and Price R. K. , 2008: Rijnland case study: Hindcast experiment for anticipatory water-system control. Atmos. Sci. Lett., 9, 5760.

    • Search Google Scholar
    • Export Citation
  • Vasiloff, S. V., and Coauthors, 2007: Improving QPE and very short term QPF: An initiative for a community-wide integrated approach. Bull. Amer. Meteor. Soc., 88, 18991911.

    • Search Google Scholar
    • Export Citation
  • Velázquez, J. A., Petit T. , Lavoie A. , Boucher M.-A. , Turcotte R. , Fortin V. , and Anctil F. , 2009: An evaluation of the Canadian global meteorological ensemble prediction system for short term hydrological forecasting. Hydrol. Earth Syst. Sci., 13, 22212231.

    • Search Google Scholar
    • Export Citation
  • Venugopal, V., Foufoula-Georgiou E. , and Sapozhnikov V. , 1999: A space-time downscaling model for rainfall. J. Geophys. Res., 104 (D16), 19 70519 721.

    • Search Google Scholar
    • Export Citation
  • Verbunt, M., Walser A. , Gurtz J. , Montani A. , and Schar C. , 2007: Probabilistic flood forecasting with a limited-area ensemble prediction system: Selected case studies. J. Hydrometeor., 8, 897909.

    • Search Google Scholar
    • Export Citation
  • Vitart, F., and Coauthors, 2008: The new VarEPS-monthly forecasting system: A first step towards seamless prediction. Quart. J. Roy. Meteor. Soc., 134, 17891799.

    • Search Google Scholar
    • Export Citation
  • Welles, E., Sorooshian S. , Carter G. , and Olsen B. , 2007: Hydrologic verification: A call for action and collaboration. Bull. Amer. Meteor. Soc., 88, 503511.

    • Search Google Scholar
    • Export Citation
  • Werner, K., Brandon D. , Clark M. , and Gangopadhyay S. , 2004: Climate index weighting schemes for NWP ESP-based seasonal volume forecasts. J. Hydrometeor., 5, 10761090.

    • Search Google Scholar
    • Export Citation
  • Werner, K., Brandon D. , Clark M. , and Gangopadhyay S. , 2005: Incorporating medium-range weather model output into the ensemble streamflow prediction system of the National Weather Service. J. Hydrometeor., 6, 101114.

    • Search Google Scholar
    • Export Citation
  • Werner, M., and Cranston M. , 2009: Understanding the value of radar rainfall nowcasts in flood forecasting and warning in flashy catchments. Meteor. Appl., 16, 4155.

    • Search Google Scholar
    • Export Citation
  • Werner, M., Cranston M. , Harrison T. , Whitfield D. , and Schellekens J. , 2009: Recent developments in operational flood forecasting in England, Wales and Scotland. Meteor. Appl., 16, 1322.

    • Search Google Scholar
    • Export Citation
  • Whitaker, J. S., Wei X. , and Vitart F. R. , 2006: Improving week-2 forecasts with multimodel reforecast ensembles. Mon. Wea. Rev., 134, 22792284.

    • Search Google Scholar
    • Export Citation
  • Wood, A. W., Leung L. R. , Sridhar V. , and Lettenmaier D. P. , 2004: Hydrologic implications of dynamical and statistical approaches to downscaling climate model outputs. Climatic Change, 62, 189216.

    • Search Google Scholar
    • Export Citation
  • Xuan, Y., Cluckie I. D. , and Wang Y. , 2009: Uncertainty analysis of hydrological ensemble forecasts in a distributed model utilising short-range rainfall prediction. Hydrol. Earth Syst. Sci., 13, 293303.

    • Search Google Scholar
    • Export Citation
  • Yilmaz, K. K., Hogue T. S. , Hsu K.-L. , Sorooshian S. , Gupta H. V. , and Wagener T. , 2005: Intercomparison of rain gauge, radar, and satellite-based precipitation estimates with emphasis on hydrologic forecasting. J. Hydrometeor., 6, 497517.

    • Search Google Scholar
    • Export Citation
  • Younis, J., Ramos M.-H. , and Thielen J. , 2008: EFAS forecasts for the March–April 2006 flood in the Czech part of the Elbe River Basin—A case study. Atmos. Sci. Lett., 9, 8894.

    • Search Google Scholar
    • Export Citation
  • Yuan, H., Mullen S. L. , Gao X. , Sorooshian S. , Du J. , and Juang H.-M. H. , 2007: Short-range probabilistic quantitative precipitation forecasts over the southwest United States by the RSM ensemble system. Mon. Wea. Rev., 135, 16851698.

    • Search Google Scholar
    • Export Citation
  • Zappa, M., and Coauthors, 2008: MAP D-PHASE: Real-time demonstration of hydrological ensemble prediction systems. Atmos. Sci. Lett., 9, 8087.

    • Search Google Scholar
    • Export Citation
  • Zipser, E., 1990: Rainfall predictability: When will extrapolation-based algorithms fail? Preprints, Eighth Conf. on Hydrometeorology, Kananaskis Park, AB, Canada, Amer. Meteor. Soc., 138–142.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    NWP/radar/satellite-driven streamflow forecast system. In this figure, “extrapolation” means a future forecast based on an extrapolation of recent measured patterns, such as is used in STEPS (described in later sections).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 1050 843 71
PDF Downloads 763 624 36

A Review of Quantitative Precipitation Forecasts and Their Use in Short- to Medium-Range Streamflow Forecasting

View More View Less
  • 1 Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China
  • | 2 Land and Water Division, Commonwealth Scientific and Industrial Research Organisation, Highett, Victoria, Australia
© Get Permissions
Full access

Abstract

Unknown future precipitation is the dominant source of uncertainty for many streamflow forecasts. Numerical weather prediction (NWP) models can be used to generate quantitative precipitation forecasts (QPF) to reduce this uncertainty. The usability and usefulness of NWP model outputs depend on the application time and space scales as well as forecast lead time. For streamflow nowcasting (very short lead times; e.g., 12 h), many applications are based on measured in situ or radar-based real-time precipitation and/or the extrapolation of recent precipitation patterns. QPF based on NWP model output may be more useful in extending forecast lead time, particularly in the range of a few days to a week, although low NWP model skill remains a major obstacle. Ensemble outputs from NWP models are used to articulate QPF uncertainty, improve forecast skill, and extend forecast lead times. Hydrologic prediction driven by these ensembles has been an active research field, although operational adoption has lagged behind. Conversely, relatively little study has been done on the hydrologic component (i.e., model, parameter, and initial condition) of uncertainty in the streamflow prediction system. Four domains of research are identified: selection and evaluation of NWP model–based QPF products, improved QPF products, appropriate hydrologic modeling, and integrated applications.

Corresponding author address: Thomas Pagano, CSIRO, P.O. Box 56, Highett VIC 3190, Australia. E-mail: thomas.pagano@csiro.au

Abstract

Unknown future precipitation is the dominant source of uncertainty for many streamflow forecasts. Numerical weather prediction (NWP) models can be used to generate quantitative precipitation forecasts (QPF) to reduce this uncertainty. The usability and usefulness of NWP model outputs depend on the application time and space scales as well as forecast lead time. For streamflow nowcasting (very short lead times; e.g., 12 h), many applications are based on measured in situ or radar-based real-time precipitation and/or the extrapolation of recent precipitation patterns. QPF based on NWP model output may be more useful in extending forecast lead time, particularly in the range of a few days to a week, although low NWP model skill remains a major obstacle. Ensemble outputs from NWP models are used to articulate QPF uncertainty, improve forecast skill, and extend forecast lead times. Hydrologic prediction driven by these ensembles has been an active research field, although operational adoption has lagged behind. Conversely, relatively little study has been done on the hydrologic component (i.e., model, parameter, and initial condition) of uncertainty in the streamflow prediction system. Four domains of research are identified: selection and evaluation of NWP model–based QPF products, improved QPF products, appropriate hydrologic modeling, and integrated applications.

Corresponding author address: Thomas Pagano, CSIRO, P.O. Box 56, Highett VIC 3190, Australia. E-mail: thomas.pagano@csiro.au

1. Introduction

a. Motivation and organization

Recent advances in weather measurement and forecasting have created opportunities to improve streamflow forecasts. The accuracy of weather forecasts has steadily improved over the years, but it has been challenging to integrate quantitative precipitation forecasts/forecasting (QPF) into flood forecasting operations. This review investigates the current status of the application of QPF as a forcing for hydrologic models. The objectives of the study are to identify current achievements and problems in the application of numerical weather prediction (NWP) model outputs and to identify frontiers for new research.

The work begins in the next section with a review of NWP models and their accuracy in predicting precipitation. This is followed by introductions to ensemble QPF methods and streamflow forecasting techniques. Section 2 describes integrated systems that use NWP model outputs to force hydrologic models. It begins with deterministic forecasts, separating very short-range forecasts from short- to medium-range forecasts. Next, issues of spatial scale and initial conditions are addressed, and operational integrated systems are identified. The remainder of the work is on the characterization of uncertainty and the role of ensembles. The study concludes with discussion and a set of recommendations.

b. NWP models

NWP models use current weather conditions as input to atmospheric models to predict the evolution of weather systems. These models represent the atmosphere as a dynamic fluid and solve for its behavior through the use of mechanics and thermodynamics.

NWP models have improved since the 1940s because of advantages in digital computing and improvements in measurement technology, including weather satellites and extensive radiosonde and radar networks (Trenberth 1992). Progress has been made in the past 50 years in weather modeling and, as a result, forecast skill has improved (Buizza et al. 1999).

Forecasting is difficult because the atmosphere is a nonlinear, chaotic system (Lorenz 1969). A subtle change in the initial and boundary layer conditions of a circulation system could result in unpredictable outcomes. NWP models have their deficiencies in describing atmospheric physical and chemical processes. In addition, unavoidable random errors in atmospheric model parameters make it difficult for NWP models to simulate atmospheric properties accurately (Buizza et al. 1999). NWP models perform worse in the Southern Hemisphere than in the Northern (mostly because of a previous lack of data in the Southern Hemisphere), although the difference has been narrowing. Simmons and Hollingsworth (2002) attribute much of the improvement in the Southern Hemisphere to the increase in global satellite data and effective data assimilation techniques. NWP models also have limited skill in the tropics (Krishnamurti et al. 1999) because of 1) the limited data in this region for model initialization and 2) difficulties in simulating cumulus convection (Koh and Ng 2009).

QPF has proven to be one of the most difficult challenges in NWP modeling because of enormous variability in space and time of the variables affecting the precipitation production process (Golding 2000; Ebert et al. 2003). QPF provides the total amount of expected liquid precipitation, and its skill is largely dependent on location, season, intensity, and storm type.

NWP models are generally good at predicting precipitation generated from synoptic frontal weather systems (as opposed to convective systems; Olson et al. 1995). The European Centre for Medium-Range Weather Forecasts (ECMWF) NWP model performs better during winter than in summer (Buizza et al. 1999). Precipitation intensity can be a forecasting challenge; Kobold and Sušelj (2005) found that ECMWF underestimated by 60% the 27–28 June 1997 precipitation events in a Slovenia catchment and ECMWF could not describe precipitation variability properly.

Furthermore, although the model resolutions have increased, NWP models still struggle to accurately forecast finescale weather systems such as local–regional convective systems (e.g., thunderstorms) and orographic process, particularly at lead times beyond 12–48 h (Golding 2000; Cerlini et al. 2005; Kaufmann et al. 2003; Richard et al. 2003). In addition to subgrid-scale issues, NWP models have challenges with displacement; that is, the intensity and shape of a storm may be correct but the location of the storm is wrong. Ebert and McBride (2000) found that displacement was the dominant source of QPF error. For extreme events (>100 mm day−1), however, intensity was the dominant source of error. Poor QPF skill has hindered hydrologic applications, particularly streamflow forecasting operations. Many researchers have had to process and adjust NWP model output–based QPF to improve the reliability in the hydrologic prediction application (e.g., Damrath et al. 2000; Landman et al. 2001; Wood et al. 2004). Such adjustments are discussed further in section 2.

c. Ensemble QPF

NWP model output ensembles have been generated since the early 1990s (e.g., ECMWF ensembles started in 1993) and probabilistic weather forecasts have been used to articulate forecast uncertainty. It is believed that NWP ensemble prediction systems exhibit greater forecast skill than any single NWP model control run or deterministic model run; ensembles increase forecast accuracy and allow for skilful predictions at longer lead times (Buizza et al. 1999; Demeritt et al. 2007). Also, NWP model–based probability forecasts issued on consecutive days are usually more consistent than single-valued forecasts (Buizza 2008).

In ensemble forecasting, one or more dimensions of forecast uncertainty are explored through the use of scenarios. There are many ways to categorize ensembles. One main distinction is the number of models used. Some ensembles combine deterministic forecasts from multiple models (Hagedorn et al. 2005). Other ensembles come from a single model with perturbed initial conditions, boundary conditions, or parameters, among others (Toth and Kalnay 1993). A collection of ensembles from individual models has been called a “grand” or “super” ensemble (Krishnamurti et al. 1999; Park et al. 2008). It is also possible to use forecasts (ensemble or deterministic) from other lead times for the same target period to form what has been called “lagged ensembles” (Mittermaier 2007).

Currently many forecasting centers are issuing regional and/or global ensemble weather forecasts. Table 1 lists the centers that provide ensemble global NWP model outputs, which are now incorporated in The Observing System Research and Predictability Experiment (THORPEX) Interactive Grand Global Ensemble (TIGGE) program (Bougeault et al. 2010).

Table 1.

Centers that provide global ensemble forecast (Park et al. 2008 and THORPEX–TIGGE; http://tigge.ucar.edu/documentation.htm; see also http://dss.ucar.edu/datasets/ds330.2/docs/tiggedocumentation.pdf).

Table 1.

A few examples of research efforts in ensemble weather forecasting are Palmer et al. (1997), Sattler and Feddersen (2005), Yuan et al. (2007), and Gebhardt et al. (2008). The ensemble members must represent the probability distribution of the state of atmosphere and improve the ensemble forecast skill compared to the control forecast. Toth and Kalnay (1997) found that only a minimal improvement in ensemble forecasts is obtained beyond 20 members. However, they also found that there was still much to be gained in the temporal and spatial relationship between spread and error by increasing the size of the ensemble beyond 40 members. As shown in Table 1, different forecast centers provide different ensemble sizes, and the minimum number is 15 [Center for Weather Forecasting and Climate Studies (CPTEC) and Chinese Meteorological Administration (CMA)].

Another category of available ensemble QPF guidance is from data extrapolation and blended/hybrid products. For example, radar, satellite, NWP model forecasts, and in situ rain gauge data can be blended statistically. This can improve the reliability of the forecasts in the very short term, provide the complete geographic coverage of satellite and NWP model output, and improve the spatial–temporal resolution of NWP model output (Vasiloff et al. 2007).

Persistence- and extrapolation-based forecasts can be better than NWP models at shorter lead times. At longer lead times, the initial conditions wash out and NWP models become the most reliable source for information (Zipser 1990). Blended products ensure that the best available information is used at each lead time. Therefore, blending is an important component of very short lead time (<12 h ahead) “nowcasting” forecasting systems.

Nowcasting is a technique for very short-range forecasting that maps the current weather and then uses an estimate of its speed and direction of movement to forecast the weather a short period ahead (usually 0–7-h lead time) based on current or most recent observations. Examples of such blended nowcasting QPF are the U.K. Short Term Ensemble Prediction System (STEPS; Bowler et al. 2006) and Ganguly and Bras (2003) for research purposes.

STEPS is a multisource blended system developed at the UKMO in collaboration with the Australian Bureau of Meteorology. In STEPS, the precipitation distribution is separated into different sizes of precipitation features. Large events are more predictable and small events have a shorter lifetime. Information from an NWP model is used to extrapolate or propagate larger precipitation features, whereas the smaller events are filled using a statistical method. Many scenarios are generated to represent various moving speeds for large events and different random statistics for small events. Active research has been undertaken to extend the forecast lead time by integrating radar and satellite data (Golding 2000).

d. Streamflow forecasting methods

Myriad authors have documented streamflow modeling methods employed in operational environments (e.g., Srikanthan et al. 1994; de Roo et al. 2000; Moore and Bell 2001; Boughton 2005; Pagano et al. 2009a,b; Hapuarachchi et al. 2011). A brief summary is provided here to provide a context for later discussions. The focus is on conceptual hydrologic models, although other models (physical, statistical, routing, etc.) are useful and described in more detail in Srikanthan et al. (1994).

Many conceptual hydrologic models accept precipitation and potential evapotranspiration (PET) as forcings, calculate losses to actual evapotranspiration, simulate the changes in surface and deeper soil moisture, and produce runoff. Oudin et al. (2006) found that, while sensitive to systematic errors in PET, hydrologic models are relatively insensitive to random errors in PET; it is often sufficient to assume that PET over the forecast horizon is the same as the long-term historical average. Therefore, the main streamflow forecasting challenge is to estimate the moisture in catchment soils by using recent measured precipitation and predicted future precipitation.

As mentioned earlier, some (“continuous”) models produce runoff by simulating soil moisture and its effects on runoff production efficiency. In comparison, “event-based” models rely on the operator to estimate the status of soil moisture and the anticipated losses (Berthet et al. 2009). For example, runoff may be some fraction of precipitation, and this fraction varies from event to event. Although their ability to simulate streamflow may differ, in theory, event-based and continuous models should be equally capable of using NWP model output as forcings.

The spatial dimension in hydrologic models is represented in one of three ways. The catchment may be lumped, in that a single time series of catchment average precipitation forces the model. It may be semidistributed in that the subbasins are represented as irregularly shaped but hydrologically homogeneous. Distributed models also relate model parameters to fields of soil characteristics and other physical properties, commonly on an evenly spaced grid (which is most similar to NWP models). In the latter two cases, runoff is aggregated to the catchment outlet using routing methods.

In operational environments, streamflow forecasters rely on individual or combined use of real-time precipitation and streamflow gauges, radar and weather satellite data, and NWP model forecasts. Around the world, the most common forecasting method uses a combination of real-time precipitation and streamflow gauge data; forecasts are often issued in a deterministic mode.

The most common operational ensemble streamflow forecasting method is to assume a range of possible scenarios of future precipitation. The U.S. National Weather Service (NWS) has generated ensembles using the Ensemble Streamflow Prediction (ESP) method for three decades (Schaake 1978). According to Day (1985), “The ESP procedure assumes that meteorological events that occurred in the past are representative of events that may occur in the future. Each year of historical meteorological data is assumed to be a possible representation of the future and is used to simulate a streamflow trace.” The standard ESP method assumes that a full range of historical meteorological possibilities is equally likely, although there are methods to indicate that certain scenarios are more likely than others (e.g., Werner et al. 2004).

The ESP procedure has been widely used for seasonal forecasting, although it can be applied to models that run on any length time step. It has been particularly successful in springtime in snowmelt-dominated regions where the initial conditions (snowpack and soil moisture) largely determine the expected volume of flow and the climate scenarios (e.g., temperature) provide a range of possibilities for timing and rate of flow. ESP has been applied in nonsnowmelt-dominated areas, mostly in combination with seasonal climate (e.g., precipitation) forecasts to determine the likelihood of each ensemble trace (Croley and Hartmann 1987). In these nonsnowmelt-dominated areas, ESP has been used for short-range (i.e., <15-day lead time) operational forecasting. For example, four of the nine NWS nonmountainous River Forecast Centers use short-range ESP, sometimes conditioning the ensembles on weather forecasts.

There are fewer peer-reviewed published studies of streamflow forecast accuracy (particularly operational forecast accuracy) than comparable studies of QPF accuracy (Welles et al. 2007). Welles et al. (2007) and Pagano et al. (2004) described the accuracy of short and seasonal operational streamflow forecasts in the United States and both found unknown future precipitation to be an important component, especially at longer lead times. It is difficult to generalize some “national” single-valued measure of skill as it depends on catchment size, in situ network records (length and quality), forcing data representativeness, geology, and many other factors. Regardless, the availability of operational streamflow forecast accuracy information is growing (Demargne et al. 2009). The research literature does contain some comparisons of hydrologic models, but these commonly assume that precipitation is known with perfect certainty and therefore can make statements about the ability of models to simulate but not necessarily forecast streamflow (e.g., Pagano et al. 2010).

Figure 1 shows the various components of a complete short- to medium-range streamflow forecasting system. This figure emphasizes that NWP model output and the hydrologic model are only two components of a system with many parts (these components are detailed in Srikanthan et al. 1994).

Fig. 1.
Fig. 1.

NWP/radar/satellite-driven streamflow forecast system. In this figure, “extrapolation” means a future forecast based on an extrapolation of recent measured patterns, such as is used in STEPS (described in later sections).

Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1347.1

2. Integrated forecast systems

The state of the art in the integration of NWP model output and streamflow forecasts in the research literature and in the operational environment is described in this section. First, the integration of deterministic QPF with hydrologic models is presented in order of forecast horizon length (short to long). The forecast ranges are based on World Meteorological Organization definitions (http://www.wmo.int/pages/prog/www/DPS/GDPS-Supplement5-AppI-4.html): nowcasting range is 0–2 h ahead, very short-range forecasting is 0–12 h, short-range forecasting is 12–72 h, and medium-range forecasting is 72–240 h. Next, the limitations of deterministic forecasts are identified and a discussion of the complete chain of uncertainty in hydrologic forecasts is given. Finally, some examples of how various groups have attempted to quantify uncertainty along this chain by using ensembles are provided. Some issues surrounding the integration of NWP model output and hydrologic models are universal across time scales; some issues also apply to both deterministic and ensemble forecasts.

a. Deterministic forecasts

1) Very short-range forecasts

Real-time flood and streamflow forecasting is mostly done deterministically. Real-time measurements from individual or integrated precipitation gauges, stream gauges, weather radars, satellites, and high-resolution deterministic NWP model forecasts are input to fast-running (i.e., finishing in seconds to minutes) hydrologic models to simulate the streamflow at a forecast point (<12 h). Although real-time forecasts are issued by operational flood and streamflow forecast centers around the world, it remains an active research field because of the uncertainties associated with the hydrologic model input data. Hapuarachchi et al. (2011) provided a comprehensive review of flash flood forecasting practices and challenges.

First, it is necessary to mention that some systems do not rely on NWP models while generating very short-term (0–12-h lead times) streamflow forecasts. For the sake of brevity, only one such non-NWP system is discussed. The Scottish Environment Protection Agency (SEPA) has an operational streamflow forecasting system forced with a combination of radar and in situ precipitation data (Werner et al. 2009). SEPA uses both measured and forecast radar precipitation for flash flood forecasting through the U.K. Nimrod system. Nimrod is an automatic system for processing radar reflectivity images and generating quantitative precipitation estimates by correcting bright band, orographic enhancement, and ground truthing based on radar gauge comparisons. It also generates forecasts by spatially advecting recently measured precipitation features. Werner and Cranston (2009) selected three small catchments in the SEPA system and used the UKMO Nimrod system for flash flood hindcasting with a lead time of 12 h, employing the lumped Probability Distributed Moisture (PDM; Moore 2007) hydrologic model for the period of 2004–08. Werner and Cranston’s (2009) results show that Nimrod’s precipitation predictions are uncertain and sometimes biased, but there is considerable benefit in their use for flood forecasting compared to not using precipitation forecasts.

Roberts et al. (2009) and Jasper et al. (2002) pointed out that radar coverage is often incomplete because radar scans are too high to see low-level precipitation, or because of ground clutter and beam blockage in rugged terrain. In this situation, high-resolution NWP model simulation can be used to fill the gaps, making such blended NWP model/radar systems more robust than single-source systems. For systems that use deterministic NWP model/radar and other information, an example is the U.K. real-time flood forecasts issued by the Environment Agency in collaboration with the UKMO (Golding 2009). Real-time precipitation, river level/flow, sea level, and measured radar data are used to initialize the hydrologic model. Six-hour-ahead radar extrapolation and raw output of forecast precipitation from UKMO NWP models are used as QPF for the hydrologic forecasting models. The skilful lead time of the hydrologic forecasting system varies from 1–12 h across the United Kingdom depending on the distance to the coastal area. Japanese hydrologists also combine NWP model output products and radar-based precipitation information to provide streamflow forecasts at 6 h lead times.

The Mesoscale Alpine Programme’s Demonstration of Probabilistic Hydrological and Atmospheric Simulation of Flood Events in the Alpine Region (MAP D-PHASE) project issued streamflow forecasts using radar information and deterministic NWP model forecasts for the summer and autumn of 2007. France established the National Hydrometeorological Service (SCHAPI) to develop techniques for merging short-range forecasts from radar-based techniques and NWP models, and to assess the ensemble precipitation forecast uncertainty for nowcasting. Other similar forecasting systems include the U.S. Gridded Flash Flood Guidance system, Central American Flash Flood Guidance system, Northern Austria Flash Flood Forecast system, Thailand Decision Support System for Flash Flood Warning, the Geospatial Data Exchange System (GEOREX) flood forecasting system of Malaysia, and systems in place in Germany and the Republic of South Africa (see more details in Hapuarachchi et al. 2011).

2) Short- to medium-range forecasts

For short- to medium-range forecasts (12–240 h), the integration of NWP and hydrologic models has been a focus of the research community and there has been relatively less operational adoption. For example, Clark and Hay (2004) used 40 years of eight-day atmospheric forecasts over the contiguous United States from the National Centers for Environmental Prediction (NCEP) reanalysis project to assess the possibilities for using the Medium-Range Forecast (NCEP–MRF) model output to aid predictions of streamflow. They concluded that NCEP–MRF output must be preprocessed before it is used for hydrologic predictions because of the systematic precipitation biases (exceeding 100% of the mean) and temperature biases (3°C).

They used a model output statistic (MOS) technique to downscale the NCEP forecasts to station locations and a forward screening multiple linear regression model to improve forecasts of precipitation and temperature. After preprocessing, Clark and Hay’s temperature forecasts were improved while precipitation forecasts were mixed. Eight-day streamflow forecasts produced using the MOS-corrected NCEP–MRF output were compared with those produced using the climatic ESP technique. They found that MOS-corrected streamflow forecasts had the most skill in snowmelt-dominated basins where temperature is the major factor. In rainfall-dominated basins, MOS-corrected streamflow forecasts were no better than those from the ESP method, primarily because the original forecasts of precipitation were poor.

Ghile and Schulze (2010) studied the effects of NWP model spatial resolution by comparing the Conformal-Cubic Atmospheric Model (C-CAM; McGregor 2005), Unified Model (UM; Davies et al. 2005), and NCEP–MRF over the Mgeni catchment in South Africa. The UM was deterministic and had resolution and lead time of 12 km and 2 days, respectively, and NCEP–MRF was an ensemble run at 2.5° (~250 km) and 14 days. They used the Agricultural Catchments Research Unit (ACRU) agrohydrological model (Schulze 2000) for streamflow simulation. Ghile and Schulze found that, without downscaling, C-CAM and UM were fair at predicting the general characteristics of precipitation but had significant biases in the mean (determined by showing that measured and forecast precipitation had positive correlation but negative root-mean-square error skill scores). Skill is the forecast performance relative to an unskilled reference; in this case, the baseline was daily measured precipitation values randomly selected from the historical record. The low skill of the precipitation forecasts was deemed unsuitable for operational hydrologic forecasting. In particular, the NCEP–MRF ensemble approach is highly positively biased and found to be “completely unskillful” for both high and low precipitation values. They suggested that, for operational use, these large-scale forecasts should be downscaled to a finer resolution based on the use of a statistical or dynamical precipitation downscaling model.

The need for downscaling and processing of precipitation forecasts is echoed in other studies. Hay and Clark (2003) statistically and dynamically downscaled NCEP model output for hydrologic simulations in the western United States. Five categories of data were used in the study: 1) NCEP output at 210-km resolution (NCEP); 2) dynamically downscaled (DDS) NCEP output using a regional climate model (RegCM2) at 52-km resolution; 3) statistically downscaled (SDS) NCEP output, which is based on empirical relations between features reliably simulated in global-scale models at gridbox scales (e.g., 500-hpa geopotential height) and surface predictands at subgrid scales (e.g., precipitation occurrence and amounts); 4) spatially averaged measured data used to calibrate the hydrologic model (Best-Sta); and 5) spatially averaged measured data derived from stations located within the area of RegCM2 model output, excluding data used in category 4 (All-Sta). SDS is a general name for a class of statistical model postprocessing techniques that includes the MOS and “perfect prog” methods commonly used in the NWP modeling community (Klein and Glahn 1974). Hay and Clark found that SDS-based simulations of daily runoff were comparable to the simulations using Best-Sta. In comparison, NCEP, DDS, and All-Sta time series showed little skill on a daily basis. They concluded that it is important to identify the causes and removal of systematic biases in DDS NWP model output.

3) Issues of spatial scale and initial conditions

The UKMO UM is now operationally run with a horizontal resolution of 4 km. Within the 4-km domain is a 1.5-km-resolution window that moves around the forecasting domain, tracking features of interest (Roberts et al. 2009). Also, 12-km horizontal resolution NWP model outputs have been available since 2005 for the United Kingdom and 25-km resolution for the globe. In the study by Roberts et al. (2009), UM grids of 12-, 4-, and 1-km horizontal resolutions were fed into the PDM model for hindcasting a flood in 2005 in northwest England. The results show the benefit of increased resolution in the UM and in coupling the high-resolution precipitation forecasts to the PDM.

Because of the computational expensiveness of high-resolution model runs, model resolution generally decreases with NWP model forecast lead time and size of the domain, as was shown in the UM example in the previous paragraph. To date, commonly available regional NWP model forecast products are still at relatively coarse resolutions (generally, around 10–20-km horizontal resolution)—for example, ECMWF and NCEP forecasts.

Habets et al. (2004) used QPF from two French NWP models as inputs to a hydrologic model. The QPF used in their study underwent statistical correction or regional adaptation to correct errors. By comparing forecasts in which the initial state was randomized versus forecasts in which the initial state was obtained by using recent measured precipitation and temperature, they found that streamflow forecasts in this region were very sensitive to the initialization of soil moisture and snowpack. They found systematic biases in precipitation amount (on the order of ~20%) and phase (rain versus snow).

4) Operational deterministic flood warning systems

Rabuffetti and Barbero (2005) described the development and implementation of a real-time flood forecasting system with a lead time of 48 h in the Piemonte region of Italy. The forecasting system is composed of survey, warning, alarm, and emergency phases. The operational FloodWatch, a geographical information system (GIS)-based decision support system for real-time flood survey and forecasting, was established in 2000, and is a 24-h operating service. The information that feeds into the system includes telemetered meteorological and hydrologic gauge data, two weather radars, weather forecasts at local and global scales, and hydrologic modeling for flood forecasting on the main river network. They find that NWP model outputs allow warnings with lead times needed by the civil protection agencies; use of NWP model output also introduces many sources of uncertainty so that the deterministic simulations need careful interpretation.

Many studies reported that NWP model outputs introduce significant uncertainty into hydrologic modeling. Georgakakos and Hudlow (1984), Damrath et al. (2000), Ebert and McBride (2000), Habets et al. (2004), Ebert et al. (2007), and Lu et al. (2010) all stated that QPF quality needs to be improved to provide reliable hydrologic prediction. Errors in QPF location, timing, and intensity hampered the direct QPF application to hydrologic prediction. Ensemble forecasts attempt to quantify this uncertainty and this is the focus of the remainder of this section.

b. Identifying and quantifying the chain of uncertainty

This subsection describes the various sources of uncertainty in streamflow forecasts. The purpose is to show that the common strategy of using ensemble forcings is not a complete representation of uncertainty.

The full spectrum of hydrologic prediction uncertainty includes several aspects: 1) model input data such as precipitation and temperature uncertainty; 2) model initial conditions such as soil moisture, snow, and river flow uncertainty; and 3) hydrologic model uncertainty due to its physical representations and parameter uncertainty. To account for the full range of uncertainty, Schaake et al. (2007) listed the main elements of the hydrologic ensemble prediction system; these are 1) an atmospheric ensemble preprocessor, 2) a data assimilator, 3) an ensemble of hydrologic models, 4) a hydrologic ensemble processor, and 5) a forecast product generator.

There have been studies about the individual aspects of uncertainty along this chain. Many studies have been done on the accuracy of NWP model forecasts in predicting precipitation, such as Houtekamer et al. (1996), Stensrud et al. (1999, 2000), Ebert and McBride (2000), Bright and Mullen (2002), Sattler and Feddersen (2005), and Ebert et al. (2007).

Several authors have then carried through NWP model uncertainty into the hydrologic forecasts, such as Kobold and Sušelj (2005), Bartholmes and Todini (2005), Gouweleeuw et al. (2005), Pappenberger et al. (2005), Collier (2009), Xuan et al. (2009) and Mascaro et al. (2010). In all cases, the hydrologic model results were very sensitive to the forcing data.

Other authors have attempted to isolate the uncertainty components. Krzysztofowicz (1999, 2002) separated forecast uncertainty into “input” (mainly precipitation) and “hydrology” (a blend of model, parameter, and others) uncertainty and then integrated the two uncertainty components into “forecast” uncertainty. A Bayesian postprocessor was used to analyze the output component error associated with particular data input types (Collier and Robbins 2008). Carpenter and Georgakakos (2004, 2006) analyzed the influences of precipitation input and hydrologic model parametric uncertainty on the flow simulation uncertainty. They found that, when using radar data, flow uncertainty is closely related to the catchment scale. Collier (2009) found that the error in weather radar precipitation data propagates through a fully distributed model. To account for the short-term precipitation errors, Collier (2009) used short-range ensemble precipitation forecasts.

Efforts to quantify hydrologic (i.e., model parameter, initial condition, and physical representation) uncertainty have had mixed results. For example, Shi et al. (2008) found that an objectively calibrated Variable Infiltration Capacity (VIC) model and noncalibrated but bias-corrected VIC model simulation produced the similar forecast skill scores for seasonal streamflow forecasts. However, without bias correcting, Yilmaz et al. (2005) reported that hydrologic model forecasts forced with satellite-based precipitation data were significantly worse when the hydrologic model parameters were not calibrated to the forcing dataset. Bohn et al. (2010) found in the context of seasonal hydrologic forecasting, multi(hydrologic)-model averaging may be no more effective in reducing forecast errors than applying a monthly bias correction to a single model in snow-dominated basins.

The key objective of investigating the uncertainty chain is to identify the sources of dominant uncertainty and to extract the useful forecast information from the uncertain system and effectively communicate forecast confidence to end users. Unfortunately, the dominant uncertainty is often basin specific or time- or space-scale dependent. Part of the issue is how few studies have attempted to analyze the total system error, especially the components of hydrologic error (e.g., Krzysztofowicz 1999; Pappenberger et al. 2005). At times it seems like the only reliable conclusion from the literature is that QPF uncertainty is the dominant source in a vast array of hydrologic forecasting contexts.

To complicate the picture, hydrologists have several tools for compensating for uncertainties within the system. In short-range streamflow forecasting, data assimilation procedures are often implemented to reduce hydrologic uncertainty. These procedures seek to reduce the difference between the simulated and measured by changing the input forcing (input updating), changing how the model simulates flow (state updating and parameter updating), or simply by postprocessing the output (output updating; Madsen and Skotner 2005). Madsen and Skotner (2005) updated the model parameters for streamflow forecasting, and Bogner and Kalas (2008) used a dynamic linear model to update the model states. Seo et al. (2003) is an example of output, input, and state updating, among many others. Output updating is cost effective and is a commonly used procedure (Sene 2010).

Although such model corrections are effective during forecasting, they can lead to unexpected interactions. For example, output updating may be calibrated on the simulation performance of the hydrologic model. If the real-time forecasts use state updating but this was not used in the historical simulation, are the statistics for the output updating still reliable? Furthermore, forecast errors may be due to deficient models (e.g., poor representation of certain processes); model improvement may be more direct and effective than data assimilation at reducing errors.

c. Ensemble forecasts

This section focuses on the use of NWP model–based ensembles to quantify uncertainty in hydrologic forecasts. In many of these studies a deterministic forecast is included as a control or baseline. The recent research literature has many studies of ensemble streamflow forecasting. Cloke and Pappenberger (2009) is an excellent contemporary review.

1) Multimodel ensembles

Ensembles made from a collection of deterministic forecasts from multiple models is occasionally called a “poor man’s ensemble.” Examples of streamflow forecasts using poor man’s ensembles are Jasper et al. (2002), Diomede et al. (2008), and Davolio et al. (2008). Jasper et al. (2002) used five high-resolution mesoscale NWP model outputs, rain gauge–measured precipitation data, and radar data to drive a distributed hydrology model and evaluated the NWP model–driven streamflow forecast performance. Although NWP model–simulated precipitation was spatially downscaled to the hydrologic model resolution, their results showed that NWP model simulations still need improvement in order to be used for short-term streamflow forecasting in fast-responding mountainous Alpine watersheds in Switzerland. They also found that NWP model performance is not consistent over time and changes with location and storm type.

As of this writing, poor man’s ensemble–based streamflow forecasting applications have not been implemented operationally. Cloke and Pappenberger (2009) suggest that this is because different models have different structural errors and cannot be combined easily.

2) Single-model ensembles

Ensembles produced by perturbing NWP model initial conditions have been used in many studies—for example, Gouweleeuw et al. (2005), Roulin and Vannitsem (2005), Werner et al. (2005), Roulin (2007), Verbunt et al. (2007), Olsson and Lindström (2008), Thirel et al. (2008), Tucci et al. (2008), van Andel et al. (2008), Younis et al. (2008), Zappa et al. (2008), Hou et al. (2009), Jaun and Ahrens (2009), Velázquez et al. (2009), and Xuan et al. (2009). Most of the above studies use ECMWF’s ensemble (50 ensemble members and one control run).

Essentially, all the studies agree that NWP model–based ensembles provided valuable added information compared to the streamflow forecast based on deterministic QPF. However, the common problem in the ensemble forecast system identified by many studies is the inaccuracy of the NWP model forecast itself. Many studies conclude that the errors in precipitation forecasts still dominate the analysis (e.g., Bartholmes and Todini 2005), even when using ensemble NWP model outputs. Ensemble precipitation forecast distribution underdispersion (i.e., overconfidence) is also a common complaint. Charges of probabilistic underdispersion have been leveled at ensemble hydrologic forecasts as well, so it may be a universal challenge in ensemble forecasting.

3) Grand ensembles

A single ensemble of NWP model forecasts from a single forecast center can be insufficient as it addresses only some of the uncertainties inherent in NWP model output (Pappenberger et al. 2008). By contrast, the grand ensemble from multiple forecast centers can incorporate more sources of uncertainty from multiple NWP models and data assimilation, thus representing a better probability distribution of forecasts.

The examples of the grand-ensemble application are Komma et al. (2007), Pappenberger et al. (2008), and He et al. (2009). The reports on the value of grand ensembles have been mixed. He et al. (2009) found that the ensemble spread is generally large and implied significant uncertainty. However, Pappenberger et al. (2008) stated that the grand ensemble captured peaks that would have been missed by single ensembles and thus provided more reliable results. In general, studies demonstrated a promising role for use of grand ensembles in streamflow forecasting.

4) Data extrapolation ensembles

It is known that there is uncertainty when combining radar, satellite, and rain gauge data with NWP model outputs as well as extrapolations of radar/satellite fields; ensembles can be used to quantify this uncertainty. Several studies have addressed ensembles generated from radar data (e.g., Cornford 2004; Germann et al. 2009) but the authors are unaware of any examples of operational streamflow forecasts that use ensemble precipitation from a blend of radar and NWP model outputs.

Germann et al. (2009) generated an ensemble radar precipitation product and fed this ensemble to a semidistributed hydrologic model for flash flooding in a mountainous Alpine catchment. A real-time experiment in the study showed that the accuracy of radar-driven runoff is comparable to that of rain gauge–driven runoff. When the precipitation variability increases across the catchment (e.g., for convective systems), it is expected that radar-based ensembles should outperform the rain gauges. In contrast, although Catchlove et al. (2005) were not using radar ensembles, they found direct use of radar rainfall estimates (without bias adjustments) led to very large biases in hydrologic simulations (i.e., a 100% error in river rise during the peak of a flood).

d. Operational ensemble forecasting and relevant programs

Although ensemble hydrologic prediction is an active research field, operational ensemble streamflow forecasting is not widespread. Indeed, operational hydrologists underutilize deterministic NWP model outputs as well. This section briefly describes operational practice in using NWP model outputs. Much of the innovation is being driven by a few international programs that are also given attention here.

The recent review paper by Cloke and Pappenberger (2009) outlined the current research and operational practices in NWP model–driven ensemble streamflow forecasting. Their review paper listed a few streamflow forecast centers (Cloke and Pappenberger 2009, their Table 1) that had operational or preoperational ensemble streamflow forecasting systems. This list is continually updated at a Web site (http://www.ecmwf.int/staff/florian_pappenberger/heps_review.html). The review did not include a few systems in the United States [e.g., the NWS Experimental Ensemble Forecast System (XEFS) and Community Hydrologic Prediction System (CHPS)] that are still under development. New developments in using NWP model output for ensemble short-term streamflow prediction at the NWS are summarized by Adams and Ostrowski (2010).

To our knowledge, the only two fully operational systems where NWP model–based ensembles are routinely and widely used to generate streamflow forecasts are the European Flood Alert System (EFAS; Thielen et al. 2009) in the Joint Research Centre of European Commission and the Swedish Meteorological and Hydrological Institute (SMHI; Olsson and Lindström 2008). All other systems are preoperational or use ensemble QPF episodically and/or for a limited subset of catchments.

EFAS incorporates deterministic NWP model output from ECMWF (10-day lead time) and the German Meteorological Service (DWD; high-resolution 7-day lead time). It also uses a full set of 50 ensemble members from ECMWF and a 16-ensemble-member run from the Consortium for Small-Scale Modeling Limited-Area Ensemble Prediction System (COSMO-LEPS). The NWP model results force a spatially distributed version of a hybrid conceptual–physical hydrologic model LISFLOOD (de Roo et al. 2000). Ensemble members are run at a 24-h time step, and deterministic forecasts are run at a 1-h time step.

Aimed at increasing preparedness for floods in transnational European river basins by providing local water authorities with medium-range probabilistic flood forecasting, EFAS complements other European national flood forecast systems with extended lead times, information upstream and in neighboring areas, and sharing of data across national boundaries. The problems identified in EFAS are that 1) local water authorities found that EFAS results were hard to interpret because of lack of expertise, 2) there is a heavy computational burden to run large ensembles, and 3) EFAS has heavy emphasis on NWP model uncertainty with no consideration of hydrologic model uncertainty (Theilen et al. 2009).

At SHMI, the operational medium-range hydrologic ensemble forecasting service was established in July 2004 (Olsson and Lindström 2008). ECMWF ensemble predictions of precipitation and temperature are inputs to the Hydrologiska Byråns Vattenbalansavdelning (HBV-96; Lindström et al. 1997) conceptual hydrologic model. Around 50 catchments that are evenly distributed throughout Sweden to represent different geography and catchment sizes were selected for forecasting. The focus, however, is on small- to medium-sized catchments that respond rapidly to meteorological changes. Forecasts are updated autoregressively (i.e., the model error for a certain day in the forecast is estimated as a function of the model error at the start of the forecast).

An international initiative to foster hydrologic ensemble prediction science is worthy of mention: the Hydrologic Ensemble Prediction Experiment (HEPEX). HEPEX is an open “international effort that brings together hydrological and meteorological communities to develop advanced probabilistic hydrologic forecast techniques that use emerging weather and climate ensemble forecasts” (Schaake et al. 2006; Thielen et al. 2008). The objective of HEPEX is reliable quantification of hydrologic forecast uncertainty.

3. Discussion and recommendations

This section discusses the challenges in using NWP model output and proposes research topics that may benefit operations. Cloke and Pappenberger (2009) discussed the problems and challenges in short- to medium-range hydrologic ensemble prediction. They identified seven challenges, which are paraphrased here:

  1. Current NWP model output skill is low; resolutions are too coarse and too few ensemble members exist.

  2. We do not understand the total uncertainty in the system and so will have trouble with extremes.

  3. Hydrologic data assimilation is underutilized.

  4. Studies rely on too few case studies.

  5. There is not enough computer power.

  6. There is inexperience with ensembles in operational flood forecasting environments.

  7. Communicating uncertainty and probabilistic forecasts is difficult.

When computational power is a limitation, deterministic NWP model forecasts are often a reasonable alternative in order to extend forecast lead time. In this case, challenges 1–4 are still relevant. Another issue that also exists for ensemble systems, but is not emphasized by Cloke and Pappenberger, is how to couple the coarse NWP model [e.g., the 12-km Australian Community Climate and Earth-System Simulator (ACCESS)] forecasts to a smaller-scale hydrologic model (e.g., 5 km). While it might not be necessary to rescale NWP model deterministic forecasts (though bias correction may still be needed), ensemble NWP model output is often at a coarser scale than hydrologic models and would need both downscaling and bias correction. A variety of downscaling methods exist (e.g., Venugopal et al. 1999).

In light of the problems and challenges that were identified by previous and present studies, there are four domains of research involved with the successful application of NWP model output to streamflow forecasting. These are described as follows.

a. Select and evaluate NWP model–based QPF products

1) Select two of these four: Resolution, lead time, ensembles, and long records

NWP models are computationally expensive. Forecasting centers’ archived and forecast data are balanced by spatial resolution, lead times, and number of ensemble members. Most centers give a range of products, each targeting one or two of the above objectives at a time (e.g., high-resolution forecasts at short lead time turning into lower-resolution forecasts at longer lead times). No group achieves all four objectives at once, although ECMWF comes the closest with its long retrospective records of greater-than-5-day-ahead high-resolution deterministic forecasts and low-resolution ensemble forecasts. As an alternative to high-resolution model output, researchers have used retrospective NWP model forecasts from very coarse (250 km) models and postprocessed the results to make them more reliable than the state-of-the-art ECMWF unadjusted forecasts (Whitaker et al. 2006).

2) Evaluate NWP model forecast accuracy

One of the main reasons that NWP model ensemble forecasts would be used in the hydrologic prediction system is to address the input forcing uncertainty. When deterministic weather forecasts are the only option because of computational limitations, the uncertainty in NWP model forecasts can be examined by comparing them to meteorological measurements. Before applying NWP model forecasts to the hydrologic prediction system, there should be a quantitative report about the NWP model performance in frequency, intensity, and location of precipitation forecasts. Many case studies in the literature did not provide this kind of information. Hartmann et al. (2002) compares how forecasters measure performance (i.e., national summary statistics) with the perspectives of users, including water managers (i.e., measures that target certain locations, seasons, lead times, and types of events). Hydrologists are particularly interested in evaluating NWP model outputs using their own set of operationally available measurements. Short records of NWP model forecasts and lack of subdaily data may, however, pose a challenge to model output evaluation, bias correction, and spatial downscaling evaluation.

3) Use bias adjustment and postprocessing as necessary

NWP model output evaluation can be done on regional- to large-scale gridded datasets, but the forecasts should also be compared to the measurements that were used to calibrate the hydrologic model. Ultimately, the hydrologic model must be calibrated with a dataset that is homogeneous with the dataset that is being used to drive the model in real time. This data may be on a different spatial resolution than the NWP model forecasts. Downscaling and bias correction will likely be necessary to ensure reasonable hydrologic forecast skill. Either statistical (e.g., Hay and Clark 2003; Friederichs and Hense 2007; Mascaro et al. 2010) or dynamical (e.g., Hay and Clark 2003; Jaun and Ahrens 2009) methods can be chosen to preprocess QPF. It will be important to understand the uncertainties the method introduces (e.g., Krzysztofowicz 1999). Some conceptual hydrologic models (e.g., Perrin et al. 2004) have ways of artificially closing the water balance and this might cause compounding interaction with QPF bias correction. The benefits of these adjustments can be measured by forcing the hydrologic model with raw NWP model output and comparing the results to forcing the hydrologic model with bias-corrected and downscaled precipitation and measured streamflow data.

b. Develop and use enhanced QPF products

1) Utilize a blended QPF for nowcasting and very short-range forecasting

Persistence- and extrapolation-based precipitation forecasts appear more skilful than NWP model output at the shortest lead times. It is also difficult to achieve adequate forecast dispersion near the start of a NWP model run. The blended products discussed in section 1 offer the greatest potential for QPF skill in the short range. The literature has examples of using NWP model–based ensembles and using radar ensembles, but none use both in combination with forecasting streamflow. The benefits of each source of information can be measured by running the blended systems in a variety of configurations (e.g., with/without radar, with/without NWP model output, etc.).

2) Demonstrate both deterministic and ensemble forecasts

Deterministic forecasting is a common operational hydrologic practice. An operational center attempting to use NWP model output may select the deterministic QPF because of its high spatial resolution. However, when computational capability allows, and adequate bias adjustment and downscaling are available, it will be important to also demonstrate the benefits of using ensemble QPF. Further postprocessing may be necessary to ensure that the ensemble forecasts are probabilistically reliable; that is, making them not over- or underdispersive.

c. Prepare the hydrologic modeling system

1) Select a hydrologic model

The choice of hydrologic model depends on the forecast lead time, catchment size, and the characteristics of runoff (Arduino et al. 2005). It will also depend on the operational forecasting center’s experience and legacy systems. Physics-based distributed hydrologic models are more likely to represent the cause–effect relationships leading to changing runoff behavior (Arduino et al. 2005). In the distributed model, flood inundation information across the catchment could be more easily derived. However, the distributed model is usually parameter intensive and there is uncertainty associated with parameters. Furthermore, physics-based distributed models still trail behind conceptual lumped models on most performance measures evaluating simulated streamflow at the catchment outlet (Smith et al. 2004). Pagano et al. (2009a) evaluated many hydrologic models in order to determine their ability to simulate streamflow and their potential for operational implementation in Australia. Many models performed equally well, although Genie Rural á 4 Parametres Journalier (GR4J; Perrin et al. 2004) stood out for its combination of performance and parsimony, having only four parameters.

2) Consider hydrologic uncertainty

Parameter uncertainty could be examined by using Pareto sets that are generated during the model calibration (e.g., Pappenberger et al. 2005). If the calibration is trying to satisfy many objectives (e.g., the fit to high flows, low flows, etc.), the Pareto set is the collection of parameter sets that perform the best for a given mixture of objectives. It is unknown if certain parameter sets behave equally well with different forcings such as measured and NWP model–simulated precipitation. Model structure uncertainty could be examined using multimodel ensemble approaches (e.g., Nijssen et al. 2003). The simplest way to investigate the initial condition uncertainty is to generate model initial conditions based on the historical climate data for particular days and examine the ensemble of hydrologic model simulations using various initial conditions. To compare the uncertainties from all aspects, there should be a common reference (measured streamflow).

3) Remove systematic hydrologic biases

A hydrologic model forced with measured precipitation may be able to simulate measured streamflow well in the calibration period, but may produce biased results in other periods. If bias is present and it is systematic, it should be removed by postprocessing.

4) Consider the effects of hydrologic data assimilation

Data assimilation can be used to improve hydrologic forecasting skill. Output updating (e.g., Anctil et al. 2003) may be the method of choice because it is cost effective and widely used. In this method, an error-correction forecast model is built based on the measured hydrologic model residuals; predictions of the anticipated error are used to adjust the output of the hydrologic model. It is an open question if output-updating methods calibrated on simulated flow will still be valid for models forced with NWP model output. Furthermore, can output updating correct biases introduced by the NWP model (while also not conflicting with the bias adjustment mentioned in the previous recommendation)?

d. Integrate NWP models and hydrologic forecast systems

1) Perform holistic evaluation of integrated forecast system

As an experiment, the NWP model–based QPF resulting from the hydrologic system forced by advanced QPF can be input to the hydrologic model with post processing and bias adjustment to form an integrated hydrologic forecast system. The system should be used to hindcast streamflow in selected basins with various sizes and other characteristics. The results can be evaluated by comparing measured streamflow with simulations driven by atmospheric measurements and forecasts.

2) Test methods for blending ensemble forecasts across time scales

Many of the forecast centers and research studies used a single source of NWP model output. Usually, NWP model output is available at a variety of resolutions and lead times. The integrated hydrologic forecasting system should be forced with a seamless time series that includes the best information from all sources and is internally consistent. A few examples combine medium and seasonal forecasts such as Vitart et al. (2008), but it would be useful to know if these methods are relevant at short time scales.

Acknowledgments

This work is part of the Water Information Research and Development Alliance (WIRADA) between CSIRO’s Water for a Healthy Country Flagship and the Bureau of Meteorology.

REFERENCES

  • Adams, T., and Ostrowski J. , 2010: Short lead-time hydrologic ensemble forecasts from numerical weather prediction model ensembles. Proc. World Environmental and Water Resources Congress, Providence, RI, American Society of Civil Engineers, 2294–2304.

    • Search Google Scholar
    • Export Citation
  • Anctil, F., Perrin C. , and Andréassian V. , 2003: ANN output updating of lumped conceptual rainfall/runoff forecasting models. J. Amer. Water Resour. Assoc., 39, 12691279.

    • Search Google Scholar
    • Export Citation
  • Arduino, G., Reggiani P. , and Todini E. , 2005: Recent advances in flood forecasting and flood risk assessment. Hydrol. Earth Syst. Sci., 9, 280284.

    • Search Google Scholar
    • Export Citation
  • Bartholmes, J., and Todini E. , 2005: Coupling meteorological and hydrological models for flood forecasting. Hydrol. Earth Syst. Sci., 9, 333346.

    • Search Google Scholar
    • Export Citation
  • Berthet, L., Andreassian V. , Perrin C. , and Javelle P. , 2009: How crucial is it to account for the antecedent moisture conditions in flood forecasting? Comparison of event-based and continuous approaches on 178 catchments. Hydrol. Earth Syst. Sci., 13, 819831.

    • Search Google Scholar
    • Export Citation
  • Bogner, K., and Kalas M. , 2008: Erro-correction methods and evaluation of an ensemble based hydrological forecasting system for the Upper Danube catchment. Atmos. Sci. Lett., 9, 95102.

    • Search Google Scholar
    • Export Citation
  • Bohn, T. J., Sonessa M. Y. , and Lettenmaier D. P. , 2010: Seasonal hydrologic forecasting: Do multimodel ensemble averages always yield improvements in forecast skill? J. Hydrometeor., 11, 13581372.

    • Search Google Scholar
    • Export Citation
  • Bougeault, P., and Coauthors, 2010: The THORPEX Interactive Grand Global Ensemble (TIGGE). Bull. Amer. Meteor. Soc., 91, 10591072.

  • Boughton, W., 2005: Catchment water balance modelling in Australia 1960–2004. Agric. Water Manage., 71, 91116.

  • Bowler, N. E., Pierce C. E. , and Seed A. W. , 2006: STEPS: A probabilistic precipitation forecasting scheme which merges an extrapolation nowcast with downscaled NWP. Quart. J. Roy. Meteor. Soc., 132, 21272155.

    • Search Google Scholar
    • Export Citation
  • Bright, D. R., and Mullen S. L. , 2002: Short-range ensemble forecasts of precipitation during the Southwest monsoon. Wea. Forecasting, 17, 10801100.

    • Search Google Scholar
    • Export Citation
  • Buizza, R., 2008: The value of probabilistic prediction. Atmos. Sci. Lett., 9, 3642.

  • Buizza, R., Miller M. , and Palmer T. N. , 1999: Stochastic representation of model uncertainties in the ECMWF Ensemble Prediction System. Quart. J. Roy. Meteor. Soc., 125, 28872908.

    • Search Google Scholar
    • Export Citation
  • Carpenter, T. M., and Georgakakos K. P. , 2004: Impacts of parametric and radar rainfall uncertainty on the ensemble streamflow simulations of a distributed hydrologic model. J. Hydrol., 298, 202221.

    • Search Google Scholar
    • Export Citation
  • Carpenter, T. M., and Georgakakos K. P. , 2006: Discretization scale dependencies of the ensemble flow range versus catchment area relationship in distributed hydrologic modelling. J. Hydrol., 328, 242257.

    • Search Google Scholar
    • Export Citation
  • Catchlove, R., Leahy C. , Seed A. , and Malone T. , 2005: An investigation into the use of real-time radar rainfall and comparison with point gauge estimates for flood forecasting. Proc. 29th Hydrology and Water Resources Symp., Canberra, Australia, Engineers Australia, 633–640.

    • Search Google Scholar
    • Export Citation
  • Cerlini, P. B., Emanuel K. A. , and Todini E. , 2005: Orographic effects on convective precipitation and space–time rainfall variability: Preliminary results. Hydrol. Earth Syst. Sci., 9, 285299.

    • Search Google Scholar
    • Export Citation
  • Clark, M. P., and Hay L. E. , 2004: Use of medium-range numerical weather prediction model output to produce forecasts of streamflow. J. Hydrometeor., 5, 1532.

    • Search Google Scholar
    • Export Citation
  • Cloke, H. I., and Pappenberger F. , 2009: Ensemble flood forecast: A review. J. Hydrol., 375, 613626.

  • Collier, C. G., 2009: On the propagation of uncertainty in weather radar estimates of rainfall through hydrological models. Meteor. Appl., 16, 3540.

    • Search Google Scholar
    • Export Citation
  • Collier, C. G., and Robbins G. L. , 2008: Using stochastic state-space modelling to specify and reduce hydrograph error. Hydrol. Processes, 22, 11731179.

    • Search Google Scholar
    • Export Citation
  • Cornford, D., 2004: A Bayesian state space modelling approach to probabilistic quantitative precipitation forecasting. J. Hydrol., 288, 92104.

    • Search Google Scholar
    • Export Citation
  • Croley, T. E., II, and Hartmann H. C. , 1987: Near real-time forecasting of large lake supplies. J. Water Resour. Plann. Manage., 113, 810823.

    • Search Google Scholar
    • Export Citation
  • Damrath, U., Doms G. , Frühwald D. , Heise E. , Richter B. , and Steppeler J. , 2000: Operational quantitative precipitation forecasting at the German Weather Service. J. Hydrol., 239, 260285.

    • Search Google Scholar
    • Export Citation
  • Davies, T., Cullen M. J. P. , Malcolm A. J. , Mawson M. H. , Staniforth A. , White A. A. , and Wood N. , 2005: A new dynamical core for the Met Office’s global and regional modelling of the atmosphere. Quart. J. Roy. Meteor. Soc., 131, 17591782.

    • Search Google Scholar
    • Export Citation
  • Davolio, S., Miglietta M. M. , Diomede T. , Marsigli C. , Morgillo A. , and Moscatello A. , 2008: A meteo-hydrological prediction system based on a multi-model approach for precipitation forecasting. Nat. Hazards Earth Syst. Sci., 8, 143159.

    • Search Google Scholar
    • Export Citation
  • Day, G. N., 1985: Extended streamflow forecasting using NWSRFS. J. Water Resour. Plann. Manage., 111, 157170.

  • Demargne, J., Mullusky M. , Werner K. , Adams T. , Lindsey S. , Schwein N. , Marosi W. , and Welles E. , 2009: Application of forecast verification science to operational river forecasting in the U.S. National Weather Service. Bull. Amer. Meteor. Soc., 90, 779784.

    • Search Google Scholar
    • Export Citation
  • Demeritt, D., Cloke H. , Pappenberger F. , Thielen J. , Bartholmes J. , and Ramos M.-H. , 2007: Ensemble predictions and perceptions of risk, uncertainty, and error in flood forecasting. Environ. Hazards, 7, 115127.

    • Search Google Scholar
    • Export Citation
  • de Roo, A. P. J., Wesseling C. G. , and Van Deursen W. P. A. , 2000: Physically based river basin modelling within a GIS: The LISFLOOD model. Hydrol. Processes, 14, 19811992.

    • Search Google Scholar
    • Export Citation
  • Diomede, T., and Coauthors, 2008: Discharge prediction based on multi-model precipitation forecasts. Meteor. Atmos. Phys., 101, 245265.

    • Search Google Scholar
    • Export Citation
  • Ebert, E. E., and McBride J. L. , 2000: Verification of precipitation in weather systems: Determination of systematic errors. J. Hydrol., 239, 179202.

    • Search Google Scholar
    • Export Citation
  • Ebert, E. E., Damrath U. , Wergen W. , and Baldwin M. E. , 2003: The WGNE assessment of short-term quantitative precipitation forecasts. Bull. Amer. Meteor. Soc., 84, 481492.

    • Search Google Scholar
    • Export Citation
  • Ebert, E. E., Janowiak J. E. , and Kidd C. , 2007: Comparison of near-real-time precipitation estimates from satellite observations and numerical models. Bull. Amer. Meteor. Soc., 88, 4764.

    • Search Google Scholar
    • Export Citation
  • Friederichs, P., and Hense A. , 2007: Statistical downscaling of extreme precipitation events using censored quantile regression. Mon. Wea. Rev., 135, 23652378.

    • Search Google Scholar
    • Export Citation
  • Ganguly, A. R., and Bras R. L. , 2003: Distributed quantitative precipitation forecasting using information from radar and numerical weather prediction models. J. Hydrometeor., 4, 11681180.

    • Search Google Scholar
    • Export Citation
  • Gebhardt, C., Theis S. , Krahe P. , and Renner V. , 2008: Experimental ensemble forecasts of precipitation based on a convection-resolving model. Atmos. Sci. Lett., 9, 6772.

    • Search Google Scholar
    • Export Citation
  • Georgakakos, K. P., and Hudlow M. D. , 1984: Quantitative precipitation forecast techniques for use in hydrologic forecasting. Bull. Amer. Meteor. Soc., 65, 11861200.

    • Search Google Scholar
    • Export Citation
  • Germann, U., Berenguer M. , Sempere-Torres D. , and Zappa M. , 2009: REAL-Ensemble radar precipitation estimation for hydrology in a mountainous region. Quart. J. Roy. Meteor. Soc., 135, 445456.

    • Search Google Scholar
    • Export Citation
  • Ghile, Y. B., and Schulze R. E. , 2010: Evaluation of three numerical weather prediction models for short and medium range agrohydrological applications. Water Resour. Manage., 24, 10051028.

    • Search Google Scholar
    • Export Citation
  • Golding, B. W., 2000: Quantitative precipitation forecasting in the UK. J. Hydrol., 239, 286305.

  • Golding, B. W., 2009: Long lead time flood warnings: Reality or fantasy? Meteor. Appl., 16, 312.

  • Gouweleeuw, B. T., Thielen J. , Franchello G. , de Roo A. P. J. , and Buizza R. , 2005: Flood forecasting using medium-range probabilistic weather prediction. Hydrol. Earth Syst. Sci., 9, 365380.

    • Search Google Scholar
    • Export Citation
  • Habets, F., LeMoigne P. , and Noilhan J. , 2004: On the utility of operational precipitation forecasts to served as input for streamflow forecasting. J. Hydrol., 293, 270288.

    • Search Google Scholar
    • Export Citation
  • Hagedorn, R., Doblas-Reyes F. J. , and Palmer T. N. , 2005: The rationale behind the success of multi-model ensembles in seasonal forecasting. I. Basic concept. Tellus, 57A, 219233.

    • Search Google Scholar
    • Export Citation
  • Hapuarachchi, P., Wang Q. J. , and Pagano T. C. , 2011: A review of advances in flash flood forecasting. Hydrol. Processes, doi:10.1002/hyp.8040, in press.

    • Search Google Scholar
    • Export Citation
  • Hartmann, H. C., Pagano T. C. , Sorooshian S. , and Bales R. , 2002: Confidence builders: Evaluating seasonal climate forecasts from user perspectives. Bull. Amer. Meteor. Soc., 83, 683698.

    • Search Google Scholar
    • Export Citation
  • Hay, L. E., and Clark M. P. , 2003: Use of statistically and dynamically downscaled atmospheric model output for hydrologic simulations in three mountainous basins in the western United States. J. Hydrol., 282, 5675.

    • Search Google Scholar
    • Export Citation
  • He, Y., Wetterhall F. , Cloke H. L. , Pappenberger F. , Wilson M. , Freer J. , and McGregor G. , 2009: Tracking the uncertainty in flood alerts driven by grand ensemble weather predictions. Meteor. Appl., 16, 91101.

    • Search Google Scholar
    • Export Citation
  • Hou, D., Mitchell K. , Toth Z. , Lohmann D. , and Wei H. , 2009: The effects of large-scale atmospheric uncertainty on streamflow predictability. J. Hydrometeor., 10, 717733.

    • Search Google Scholar
    • Export Citation
  • Houtekamer, P. L., Lefaivre L. , Derome J. , Ritchie H. , and Mitchell H. L. , 1996: A system simulation approach to ensemble prediction. Mon. Wea. Rev., 124, 12251242.

    • Search Google Scholar
    • Export Citation
  • Jasper, K., Gurtz J. , and Lang H. , 2002: Advanced flood forecasting in Alpine watersheds by coupling meteorological observations and forecasts with a distributed hydrological model. J. Hydrol., 267, 4052.

    • Search Google Scholar
    • Export Citation
  • Jaun, S., and Ahrens B. , 2009: Evaluation of a probabilistic hydrometeorological forecast system. Hydrol. Earth Syst. Sci., 13, 10311043.

    • Search Google Scholar
    • Export Citation
  • Kaufmann, P., Schubiger F. , and Binder P. , 2003: Precipitation forecasting by a mesoscale numerical weather prediction (NWP) model: Eight years of experience. Hydrol. Earth Syst. Sci., 7, 812832.

    • Search Google Scholar
    • Export Citation
  • Klein, W. H., and Glahn H. R. , 1974: Forecasting local weather by means of model output statistics. Bull. Amer. Meteor. Soc., 55, 12171227.

    • Search Google Scholar
    • Export Citation
  • Kobold, M., and Sušelj K. , 2005: Precipitation forecasts and their uncertainty as input into hydrological models. Hydrol. Earth Syst. Sci., 9, 322332.

    • Search Google Scholar
    • Export Citation
  • Koh, T.-Y., and Ng J. S. , 2009: Improved diagnostics for NWP verification in the tropics. J. Geophys. Res., 114, D12102, doi:10.1029/2008JD011179.

    • Search Google Scholar
    • Export Citation
  • Komma, J., Reszler C. , Blöschl G. , and Haiden T. , 2007: Ensemble prediction of floods—Catchment non-linearity and forecast probability. Nat. Hazards Earth Syst. Sci., 7, 431444.

    • Search Google Scholar
    • Export Citation
  • Krishnamurti, T. N., Kistawal C. M. , LaRow T. E. , Bachiochi D. R. , Zhang Z. , Williford C. E. , Gadgil S. , and Surendran S. , 1999: Improved weather and seasonal climate forecasts from multimodel superensemble. Science, 285, 15481550.

    • Search Google Scholar
    • Export Citation
  • Krzysztofowicz, R., 1999: Bayesian theory of probabilistic forecasting via deterministic hydrologic model. Water Resour. Res., 35, 27392750.

    • Search Google Scholar
    • Export Citation
  • Krzysztofowicz, R., 2002: Bayesian system for probabilistic river stage forecasting. J. Hydrol., 268, 1640.

  • Landman, W. A., Mason S. J. , Tyson P. D. , and Tennant W. J. , 2001: Statistical downscaling of GCM simulations to streamflow. J. Hydrol., 252, 221236.

    • Search Google Scholar
    • Export Citation
  • Lindström, G., Johansson B. , Persson M. , Gardelin M. , and Bergström S. , 1997: Development and test of the distributed HBV-96 hydrological model. J. Hydrol., 201, 272288.

    • Search Google Scholar
    • Export Citation
  • Lorenz, E., 1969: The predictability of a flow which contains many scales of motion. Tellus, 21A, 289307.

  • Lu, C., Yuan H. , Tollerud E. I. , and Wang N. , 2010: Scale-dependent uncertainties in global QPFs and QPEs from NWP model and satellite fields. J. Hydrometeor., 11, 139155.

    • Search Google Scholar
    • Export Citation
  • Madsen, H., and Skotner C. , 2005: Adaptive state updating in real-time river flow forecasting—A combined filtering and error forecasting procedure. J. Hydrol., 308, 302312.

    • Search Google Scholar
    • Export Citation
  • Mascaro, G., Vivoni E. R. , and Deidda R. , 2010: Implications of ensemble quantitative precipitation forecast errors on distributed streamflow forecasting. J. Hydrometeor., 11, 6986.

    • Search Google Scholar
    • Export Citation
  • McGregor, J. L., 2005: C-CAM: Geometric aspects and dynamical formulation. CSIRO Tech. Rep. 70, 41 pp.

  • Mittermaier, M. P., 2007: Improving short-range high-resolution model precipitation forecast skill using time-lagged ensembles. Quart. J. Roy. Meteor. Soc., 133, 14871500.

    • Search Google Scholar
    • Export Citation
  • Moore, R. J., 2007: The PDM rainfall-runoff model. Hydrol. Earth Syst. Sci., 11, 483499.

  • Moore, R. J., and Bell V. A. , 2001: Comparison of rainfall-runoff models for flood forecasting. Part 1: Literature review of models. Environment Agency R&D Tech. Rep. W241, 94 pp.

    • Search Google Scholar
    • Export Citation
  • Nijssen, B., and Coauthors, 2003: Simulation of high latitude hydrological processes in the Torne–Kalix basin: PILPS Phase 2(e): 2: Comparison of model results with observations. Global Planet. Change, 38, 3153.

    • Search Google Scholar
    • Export Citation
  • Olson, D. A., Junker N. W. , and Korty B. , 1995: Evaluation of 33 years of quantitative precipitation forecasting at the NMC. Wea. Forecasting, 10, 498511.

    • Search Google Scholar
    • Export Citation
  • Olsson, J., and Lindström G. , 2008: Evaluation and calibration of operational hydrological ensemble forecasts in Sweden. J. Hydrol., 350, 1424.

    • Search Google Scholar
    • Export Citation
  • Oudin, L., Perrin C. , Mathevet T. , Andréassian V. , and Michel C. , 2006: Impact of biased and randomly corrupted inputs on the efficiency and the parameters of watershed models. J. Hydrol., 320, 6283.

    • Search Google Scholar
    • Export Citation
  • Pagano, T. C., Garen D. , and Sorooshian S. , 2004: Evaluation of official western U.S. seasonal water supply outlooks, 1922–2002. J. Hydrometeor., 5, 896909.

    • Search Google Scholar
    • Export Citation
  • Pagano, T. C., Hapuarachchi P. , and Wang Q. J. , 2009a: Continuous soil moisture accounting and routing modelling to support short lead-time streamflow forecasting. CSIRO Tech. Rep., 53 pp.

    • Search Google Scholar
    • Export Citation
  • Pagano, T. C., Hapuarachchi P. , and Wang Q. J. , 2009b: Development and testing of a multi-model rainfall-runoff streamflow forecasting application. CSIRO Tech. Rep., 40 pp.

    • Search Google Scholar
    • Export Citation
  • Pagano, T. C., Hapuarachchi P. , and Wang Q. J. , 2010: Continuous rainfall-runoff model comparison and short-term daily streamflow forecast skill evaluation. CSIRO Tech. Rep., 70 pp.

    • Search Google Scholar
    • Export Citation
  • Palmer, T. N., Barkmeijer J. , Buizza R. , and Petroliagis T. , 1997: The ECMWF ensemble prediction system. Meteor. Appl., 4, 301304.

  • Pappenberger, F., Beve K. J. , Hunter N. M. , Bates P. D. , Gouweleeuw B. T. , Thielen J. , and de Roo A. P. J. , 2005: Cascading model uncertainty from medium range weather forecasts (10 days) through a rainfall-runoff model to flood inundation predictions within the European Flood Forecasting System (EFFS). Hydrol. Earth Syst. Sci., 9, 381393.

    • Search Google Scholar
    • Export Citation
  • Pappenberger, F., Bartholmes J. , Thielen J. , Cloke H. L. , Buizza R. , and de Roo A. , 2008: New dimensions in early flood warning across the globe using grand-ensemble weather predictions. Geophys. Res. Lett., 35, L10404, doi:10.1029/2008GL033837.

    • Search Google Scholar
    • Export Citation
  • Park, Y.-Y., Buizza R. , and Leutbecher M. , 2008: TIGGE: Preliminary results on comparing and combining ensembles. Quart. J. Roy. Meteor. Soc., 134, 20292050.

    • Search Google Scholar
    • Export Citation
  • Perrin, C., Michel C. , and Andréassian V. , 2004: Improvement of a parsimonious model for streamflow simulation. J. Hydrol., 279, 275289.

    • Search Google Scholar
    • Export Citation
  • Rabuffetti, D., and Barbero S. , 2005: Operational hydro-meteorological warning and real-time flood forecasting: The Piemonte Region case study. Hydrol. Earth Syst. Sci., 9, 457466.

    • Search Google Scholar
    • Export Citation
  • Richard, E., Cosma S. , Benoit R. , Binder P. , Buzzi A. , and Kaufmann P. , 2003: Intercomparison of mesoscale meteorological models for precipitation forecasting. Hydrol. Earth Syst. Sci., 7, 799811.

    • Search Google Scholar
    • Export Citation
  • Roberts, N. M., Cole S. J. , Forbes R. M. , Moore R. J. , and Boswell D. , 2009: Use of high-resolution NWP rainfall and river flow forecasts for advance warning of the Carlisle flood, north-west England. Meteor. Appl., 16, 2334.

    • Search Google Scholar
    • Export Citation
  • Roulin, E., 2007: Skill and relative economic value of medium-range hydrological ensemble predictions. Hydrol. Earth Syst. Sci., 11, 725737.

    • Search Google Scholar
    • Export Citation
  • Roulin, E., and Vannitsem S. , 2005: Skill of medium-range hydrological ensemble predictions. J. Hydrometeor., 6, 729744.

  • Sattler, K., and Feddersen H. , 2005: Limited-area short-range ensemble predictions targeted for heavy rain in Europe. Hydrol. Earth Syst. Sci., 9, 300312.

    • Search Google Scholar
    • Export Citation
  • Schaake, J., 1978: The National Weather Service extended streamflow prediction techniques: Description and applications during 1977. Proc. Third Annual Climate Diagnostics Workshop, Miami, FL, National Oceanic and Atmospheric Administration, 31.1–31.19.

    • Search Google Scholar
    • Export Citation
  • Schaake, J., Franz K. , Bradley A. , and Buizza R. , 2006: The Hydrologic Ensemble Prediction Experiment (HEPEX). Hydrol. Earth Syst. Sci. Discuss., 3, 33213332.

    • Search Google Scholar
    • Export Citation
  • Schaake, J., Hamill T. M. , Buizza R. , and Clark M. , 2007: HEPEX: The Hydrological Ensemble Prediction Experiment. Bull. Amer. Meteor. Soc., 88, 15411547.

    • Search Google Scholar
    • Export Citation
  • Schulze, R. E., 2000: Transcending scales of space and time in impact studies of climate and climate change on agrohydrological responses. Agric. Ecosyst. Environ., 82, 185212.

    • Search Google Scholar
    • Export Citation
  • Sene, K., 2010: Hydrometeorology: Forecasting and Applications. Springer, 355 pp.

  • Seo, D.-J., Koren V. , and Cajina N. , 2003: Real-time variational assimilation of hydrologic and hydrometeorological data into operational hydrologic forecasting. J. Hydrometeor., 4, 627641.

    • Search Google Scholar
    • Export Citation
  • Shi, X., Wood A. W. , and Lettenmaier D. P. , 2008: How essential is hydrological model calibration to seasonal streamflow forecasting? J. Hydrometeor., 9, 13501363.

    • Search Google Scholar
    • Export Citation
  • Simmons, A. J., and Hollingsworth A. , 2002: Some aspects of the improvement in skill of numerical weather prediction. Quart. J. Roy. Meteor. Soc., 128, 647677.

    • Search Google Scholar
    • Export Citation
  • Smith, M. B., Georgakakos K. P. , and Liang X. , 2004: The distributed model intercomparison project (DMIP). J. Hydrol., 298, 13.

  • Srikanthan, R., Elliott J. F. , and Adams G. , 1994: A review of real-time flood forecasting methods. Cooperative Research Center for Catchment Hydrology Tech. Rep. 94/2, 120 pp.

    • Search Google Scholar
    • Export Citation
  • Stensrud, D. J., Brooks H. E. , Du J. , Tracton M. S. , and Rogers E. , 1999: Using ensembles for short-range forecasting. Mon. Wea. Rev., 127, 433446.

    • Search Google Scholar
    • Export Citation
  • Stensrud, D. J., Bao J.-W. , and Warner T. T. , 2000: Using initial condition and model physics perturbations in short-range ensemble simulations of mesoscale convective systems. Mon. Wea. Rev., 128, 20772107.

    • Search Google Scholar
    • Export Citation
  • Thielen, J., Schaake J. , Hartman R. , and Buizza R. , 2008: Aims, challenges and progress of the Hydrological Ensemble Prediction Experiment (HEPEX) following the third HEPEX workshop held in Stresa 27 to 29 June 2007. Atmos. Sci. Lett., 9, 2935.

    • Search Google Scholar
    • Export Citation
  • Thielen, J., Bartholmes J. , Ramos M.-H. , and de Roo A. , 2009: The European Flood Alert System—Part 1: Concept and development. Hydrol. Earth Syst. Sci., 13, 125140.

    • Search Google Scholar
    • Export Citation
  • Thirel, G., Rousset-Regimbeau F. , Martin E. , and Habets F. , 2008: On the impact of short-range meteorological forecasts for ensemble streamflow predictions. J. Hydrometeor., 9, 13011317.