• AghaKouchak, A., , Behrangi A. , , Sorooshian S. , , Hsu K. , , and Amitai E. , 2011: Evaluation of satellite-retrieved extreme precipitation rates across the central United States. J. Geophys. Res., 116, D02115, doi:10.1029/2010JD014741.

    • Search Google Scholar
    • Export Citation
  • Alfieri, L., , Salamon P. , , Pappenberger F. , , Wetterhall F. , , and Thielen J. , 2012: Operational early warning systems for water-related hazards in Europe. Environ. Sci. Policy, 21, 3549.

    • Search Google Scholar
    • Export Citation
  • Anagnostou, E. N., , Maggioni V. , , Nikolopoulos E. I. , , Meskele T. , , Hossain F. , , and Papadopoulos A. , 2010: Benchmarking high-resolution global satellite rainfall products to radar and rain-gauge rainfall estimates. IEEE Trans. Geosci. Remote Sens., 48, 16671683, doi:10.1109/TGRS.2009.2034736.

    • Search Google Scholar
    • Export Citation
  • Andreassian, V., , Perrin C. , , Michel C. , , Usartsanchez I. , , and Lavabre J. , 2001: Impact of imperfect rainfall knowledge on the efficiency and the parameters of watershed models. J. Hydrol., 250 (1–4), 206223.

    • Search Google Scholar
    • Export Citation
  • Artan, G., , Gadain H. , , Smith J. L. , , Asante K. , , Bandaragoda C. J. , , and Verdin J. P. , 2007: Adequacy of satellite derived rainfall data for streamflow modeling. Nat. Hazards, 43, 167185.

    • Search Google Scholar
    • Export Citation
  • Baroni, G., , Facchi A. , , Gandolfi C. , , Ortuani B. , , Horeschi D. , , and van Dam J. C. , 2010: Uncertainty in the determination of soil hydraulic parameters and its influence on the performance of two hydrological models of different complexity. Hydrol. Earth Syst. Sci., 14, 251270.

    • Search Google Scholar
    • Export Citation
  • Barret, E. C., , and Martin D. W. , 1981: The Use of Satellite Data in Rainfall Monitoring. Academic Press, 340 pp.

  • Behrangi, A., , Khakbaz B. , , Jaw T. C. , , AghaKouchak A. , , Hsu K. , , and Sorooshian S. , 2011: Hydrologic evaluation of satellite precipitation products over a mid-size basin. J. Hydrol., 397 (3–4), 225237.

    • Search Google Scholar
    • Export Citation
  • Bitew, M. M., , and Gebremichael M. , 2009: Evaluation through independent measurements: Complex terrain and humid tropical region in Ethiopia. Satellite Rainfall Applications for Surface Hydrology, M. Gebremichael and F. Hossain, Eds., Springer, 205–214.

  • Bitew, M. M., , and Gebremichael M. , 2011: Assessment of satellite rainfall products for streamflow simulation in medium watersheds of the Ethiopian highlands. Hydrol. Earth Syst. Sci., 15, 11471155.

    • Search Google Scholar
    • Export Citation
  • Bitew, M. M., , Gebremichael M. , , Ghebremichael L. T. , , and Bayissa Y. A. , 2012: Evaluation of high-resolution satellite rainfall products through streamflow simulation in a hydrological modeling of a small mountainous watershed in Ethiopia. J. Hydrometeor., 13, 338350.

    • Search Google Scholar
    • Export Citation
  • Borga, M., , Boscolo P. , , Zanon F. , , and Sangati M. , 2007: Hydrometeorological analysis of the 29 August 2003 flash flood in the eastern Italian Alps. J. Hydrometeor., 8, 10491067.

    • Search Google Scholar
    • Export Citation
  • Borga, M., , Gaume E. , , Creutin J. D. , , and Marchi L. , 2008: Surveying flash floods: Gauging the ungauged extremes. Hydrol. Processes, 22, 38833885.

    • Search Google Scholar
    • Export Citation
  • Borneman, R., 1988: Satellite rainfall estimating program of the NOAA/NESDIS Satellite Analysis Branch. Natl. Wea. Dig., 13 (2), 715.

    • Search Google Scholar
    • Export Citation
  • Borrell, V. E., , Dartus D. , , and Alquier M. , 2007: Forecasting flash floods in ungauged basins with satellite data. Proc. PUB Kick-Off Meeting, Brasilia, Brazil, IAHS, 249–257.

  • Brath, A., , Montanari A. , , and Toth E. , 2004: Analysis of the effects of different scenarios of historical data availability on the calibration of a spatially-distributed hydrological model. J. Hydrol., 291, 232253.

    • Search Google Scholar
    • Export Citation
  • Cabral, M. C., , Garrote L. , , Bras R. L. , , and Entekhabi D. , 1992: A kinematic model of infiltration and runoff generation in layered and sloped soils. Adv. Water Resour., 15, 311324.

    • Search Google Scholar
    • Export Citation
  • Carpenter, T. M., , Sperfslage J. A. , , Georgakakos K. P. , , Sweeney T. , , and Fread D. L. , 1999: National threshold runoff estimation utilizing GIS in support of operational flash flood warning systems. J. Hydrol., 224, 2144.

    • Search Google Scholar
    • Export Citation
  • Clark, M. P., , and Vrugt J. A. , 2006: Unraveling uncertainties in hydrologic model calibration: Addressing the problem of compensatory parameters. Geophys. Res. Lett., 33, L06406, doi:10.1029/2005GL025604.

    • Search Google Scholar
    • Export Citation
  • Collischonn, B., , Collischonn W. , , and Tucci C. E. M. , 2008: Daily hydrological modeling in the Amazon basin using TRMM rainfall estimates. J. Hydrol., 360, 207216.

    • Search Google Scholar
    • Export Citation
  • Creutin, J. D., , and Borga M. , 2003: Radar hydrology modifies the monitoring of flash-flood hazard. Hydrol. Processes, 17, 14531456.

  • Cucchi, F., , Piano S. , , Marinetti E. , , Massari G. , , Oberti S. , , and Zini L. , 2000: Studies for the realization of the hydrological map of Friuli-Venezia Giulia. Ipogea, 3, 5771.

    • Search Google Scholar
    • Export Citation
  • Delrieu, G., and Coauthors, 2005: The catastrophic flash-flood event of 8–9 September 2002 in the Gard region, France: A first case study for the Cèvennes–Vivarais Mediterranean Hydrometeorological Observatory. J. Hydrometeor., 6, 3452.

    • Search Google Scholar
    • Export Citation
  • Duan, Q., , Sorooshian S. , , and Gupta V. K. , 1992: Effective and efficient global optimization for conceptual rainfall-runoff models. Water Resour. Res., 28, 10151031.

    • Search Google Scholar
    • Export Citation
  • Garrote, L., , and Bras R. L. , 1995: A distributed model for real-time flood forecasting using digital elevation models. J. Hydrol., 167, 279306.

    • Search Google Scholar
    • Export Citation
  • Gaume E., , and Borga M. , 2008: Post-flood field investigations in upland catchments after major flash floods: Proposal of a methodology and illustrations. J. Flood Risk Manage., 1 (4), 175189.

    • Search Google Scholar
    • Export Citation
  • Georgakakos, K. P., 1986: A generalized stochastic hydrometeorological model for flood and flash-flood forecasting: 1. Formulation. Water Resour. Res., 22, 20832095, doi:10.1029/WR022i013p02083.

    • Search Google Scholar
    • Export Citation
  • Gourley, J. J., , Hong Y. , , Flamig Z. L. , , Li L. , , and Wang J. , 2010: Intercomparison of rainfall estimates from radar, satellite, gauge, and combinations for a season of record rainfall. J. Appl. Meteor. Climatol., 49, 437452.

    • Search Google Scholar
    • Export Citation
  • Gourley, J. J., , Hong Y. , , Flamig Z. L. , , Wang J. , , Vergara H. , , and Anagnostou E. N. , 2011: Hydrologic evaluation of rainfall estimates from radar, satellite, gauge, and combinations on Ft. Cobb basin, Oklahoma. J. Hydrometeor., 12, 973988.

    • Search Google Scholar
    • Export Citation
  • Gourley, J. J., , Erlingis J. M. , , Hong Y. , , and Wells E. B. , 2012: Evaluation of tool used for monitoring and forecasting flash floods in the United States. Wea. Forecasting, 27, 158173.

    • Search Google Scholar
    • Export Citation
  • Grimes, D. I. F., , and Diop M. , 2003: Satellite-based rainfall estimation for river flow forecasting in Africa. I: Rainfall estimates and hydrological forecasts. Hydrol. Sci. J., 48, 567584.

    • Search Google Scholar
    • Export Citation
  • Guetter, A. K., , Georgakakos K. P. , , and Tsonis A. A. , 1996: Hydrologic applications of satellite data: 2. Flow simulation and soil water estimates. J. Geophys. Res., 101 (D21), 26 52726 538.

    • Search Google Scholar
    • Export Citation
  • Hirpa, F. A., , Gebremichael M. , , and Hopson T. , 2010: Evaluation of high resolution satellite precipitation products over very complex terrain in Ethiopia. J. Appl. Meteor. Climatol., 49, 10441051.

    • Search Google Scholar
    • Export Citation
  • Hong, Y., , Hsu K.-L. , , Sorooshian S. , , and Gao X. , 2004: Precipitation estimation from remotely sensed imagery using an Artificial Neural Network Cloud Classification System. J. Appl. Meteor., 43, 18341852.

    • Search Google Scholar
    • Export Citation
  • Hossain, F., , and Anagnostou E. N. , 2004: Assessment of current passive microwave and infrared based satellite rainfall remote sensing for flood prediction. J. Geophys. Res., 109, D07102, doi:10.1029/2003JD003986.

    • Search Google Scholar
    • Export Citation
  • Huffman, G. J., and Coauthors, 1997: The Global Precipitation Climatology Project (GPCP) combined precipitation dataset. Bull. Amer. Meteor. Soc., 78, 520.

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

    • Search Google Scholar
    • Export Citation
  • Hughes, D. A., 2006: Comparison of satellite rainfall data with observations from gauging station networks. J. Hydrol., 327 (3–4), 399410.

    • Search Google Scholar
    • Export Citation
  • Ivanov, V. Y., , Vivoni E. R. , , Bras R. L. , , and Entekhabi D. , 2004a: Catchment hydrologic response with a fully distributed triangulated irregular network model. Water Resour. Res., 40, W11102, doi:10.1029/2004WR003218.

    • Search Google Scholar
    • Export Citation
  • Ivanov, V. Y., , Vivoni E. R. , , Bras R. L. , , and Entekhabi D. , 2004b: Preserving high-resolution surface and rainfall data in operational-scale basin hydrology: A fully-distributed physically-based approach. J. Hydrol., 298 (1–4), 80111.

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

    • Search Google Scholar
    • Export Citation
  • Krajewski, W. F., , and Smith J. A. , 2002: Radar hydrology: Rainfall estimation. Adv. Water Resour., 25, 13871394.

  • Kuligowski, R. J., 1997: An overview of National Weather Service quantitative precipitation estimates (QPE). Techniques Development Laboratory Office Note 97-4, 27 pp.

  • La Barbera, P., , Lanza L. , , and Siccardi F. , 1993: Flash flood forecasting based on multisensor information. Proc. Symp. on Extreme Hydrological Events: Precipitation, Floods and Droughts, Yokohama, Japan, IAHS, 21–32.

  • Maddox, R. A., , Zhang J. , , Gourley J. J. , , and Howard K. W. , 2002: Weather radar coverage over the contiguous United States. Wea. Forecasting, 17, 927934.

    • Search Google Scholar
    • Export Citation
  • Marchi, L., and Coauthors, 2009: Comprehensive post-event survey of a flash flood in western Slovenia: Observation strategy and lessons learned. Hydrol. Processes, 23, 37613770.

    • Search Google Scholar
    • Export Citation
  • Martina, M. L. V., , Todini E. , , and Libralon A. , 2005: A Bayesian decision approach to rainfall thresholds based flood warning. Hydrol. Earth Syst. Sci. Discuss., 2, 26632706.

    • Search Google Scholar
    • Export Citation
  • Nikolopoulos, E. I., , Anagnostou E. N. , , Hossain F. , , Gebremichael M. , , and Borga M. , 2010: Understanding the scale relationships of uncertainty propagation of satellite rainfall through a distributed hydrologic model. J. Hydrometeor., 11, 520532.

    • Search Google Scholar
    • Export Citation
  • Nikolopoulos, E. I., , Anagnostou E. N. , , Borga M. , , Vivoni E. R. , , and Papadopoulos A. , 2011: Sensitivity of a mountain basin flash flood to initial wetness conditions and rainfall variability. J. Hydrol., 402 (3–4), 165178.

    • Search Google Scholar
    • Export Citation
  • Norbiato, D., , Borga M. , , Esposti S. D. , , Gaume E. , , and Anquetin S. , 2008: Flash flood warning based on rainfall thresholds and soil moisture conditions: An assessment for gauged and ungauged basins. J. Hydrol., 362, 274290.

    • Search Google Scholar
    • Export Citation
  • Ogden, F. L., , Sharif H. O. , , Senarath S. U. S. , , Smith J. A. , , Baeck M. L. , , and Richardson J. R. , 2000: Hydrologic analysis of the Fort Collins, Colorado flash flood of 1997. J. Hydrol., 228, 82100.

    • Search Google Scholar
    • Export Citation
  • Petty, G., , and Krajewski W. F. , 1996: Satellite estimation of precipitation over land. Hydrol. Sci. J., 41, 433451.

  • Sangati, M., , and Borga M. , 2009: Influence of rainfall spatial resolution on flash flood modelling. Nat. Hazards Earth Syst. Sci., 9, 575584.

    • Search Google Scholar
    • Export Citation
  • Sapiano, M. R. P., , and Arkin P. A. , 2009: An intercomparison and validation of high resolution satellite precipitation estimates with 3-hourly gauge data. J. Hydrometeor., 10, 149166.

    • Search Google Scholar
    • Export Citation
  • Scofield, R. A., 1987: The NESDIS operational convective precipitation estimation technique. Mon. Wea. Rev., 115, 17731792.

  • Scofield, R. A., , and Oliver V. J. , 1977: A scheme for estimating convective rainfall from satellite imagery. NOAA Tech. Memo. NESS 86, 47 pp.

  • Scofield, R. A., , and Kuligowski R. J. , 2003: Status and outlook of operational satellite precipitation algorithms for extreme-precipitation events. Wea. Forecasting, 18, 10371051.

    • Search Google Scholar
    • Export Citation
  • Su, F., , Hong Y. , , and Lettenmaier D. P. , 2008: Evaluation of TRMM Multisatellite Precipitation Analysis (TMPA) and its utility in hydrologic prediction in the La Plata basin. J. Hydrometeor., 9, 622640.

    • Search Google Scholar
    • Export Citation
  • Tropeano, D., , Turconi L. , , and Sanna S. , 2004: Debris flow triggered by the 29 August 2003 cloudburst in Val Canale, eastern Italian Alps. Proc. Int. Symp. INTERPRAEVENT 2004, Riva del Garda, Italy, INTERPRAEVENT, 121–132.

  • Tsintikidis, D., , Georgakakos K. P. , , Artan G. A. , , and Tsonis A. A. , 1999: A feasibility study on mean areal rainfall estimation and hydrologic response in the Blue Nile region using METEOSAT images. J. Hydrol., 221, 97116.

    • Search Google Scholar
    • Export Citation
  • Vivoni, E. R., , Ivanov V. Y. , , Bras R. L. , , and Entekhabi D. , 2004: Generation of triangulated irregular networks based on hydrological similarity. J. Hydrol. Eng., 9, 288303.

    • Search Google Scholar
    • Export Citation
  • Vivoni, E. R., , Entekhabi D. , , Bras R. L. , , Ivanov V. Y. , , Van Horne M. P. , , Grassotti C. , , and Hoffman R. N. , 2006: Extending the predictability of hydrometeorological flood events using radar rainfall nowcasting. J. Hydrometeor., 7, 660677.

    • Search Google Scholar
    • Export Citation
  • Vivoni, E. R., , Entekhabi D. , , and Hoffman R. N. , 2007: Error propagation of radar rainfall nowcasting fields through a fully-distributed flood forecasting model. J. Appl. Meteor. Climatol., 46, 932940.

    • Search Google Scholar
    • Export Citation
  • Wilk, J., , Kniveton D. , , Andersson L. , , Layberry R. , , Todd M. C. , , Hughes D. , , Ringrose S. , , and Vanderpost C. , 2006: Estimating rainfall and water balance over the Okavango River Basin for hydrological applications. J. Hydrol., 331, 1829.

    • Search Google Scholar
    • Export Citation
  • Wu, H., , Adler R. F. , , Hong Y. , , Tian Y. , , and Policelli F. , 2012: Evaluation of global flood detection using satellite-based rainfall and a hydrologic model. J. Hydrometeor., 13, 12681284.

    • Search Google Scholar
    • Export Citation
  • Yilmaz, K. K., , Adler R. F. , , Tian Y. , , Hong Y. , , and Pierce H. F. , 2010: Evaluation of a satellite-based global flood monitoring system. Int. J. Remote Sens., 31, 37633782.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Digital elevation map of northeastern Italian region (highlighted in the inset map of Italy) showing the location of the OSMER radar and the outline boundaries of Fella River basin.

  • View in gallery

    Time series of the Fella basin-averaged rainfall during the 29 Aug 2003 flash flood derived from (a) radar, CMORPH, and PERSIANN rainfall fields at hourly scale and (b) radar and 3B42 rainfall fields at 3-hourly scale.

  • View in gallery

    Box plot showing for each satellite product the temporal distribution of mean field bias (MFBdyn) derived over the entire radar domain. Values for storm-total bias (MFB) for the radar domain and for the Fella basin are superimposed. Note that plus symbols correspond to outliers defined as values that are above (below) the 3rd (1st) quartile by 1.5 times the interquartile range.

  • View in gallery

    Radar-rainfall quantiles vs satellite rainfall quantiles for (left) CMORPH, (middle) PERSIANN, and (right) 3B42. Solid line corresponds to the slope = 1.

  • View in gallery

    (a) Radar-simulated hydrograph (reference) against discharge observations at the outlet of Fella and (b)–(d) satellite-simulated hydrographs at the outlet of Fella based on (b) 3B42, (c) CMORPH, and (d) PERSIANN product before and after the bias adjustment. Note that scales are different between (a) and (b)–(d) to improve presentation of the results.

  • View in gallery

    Simulated hydrographs based on individual calibration of each satellite product (a) before and (b),(c) after bias adjustment.

  • View in gallery

    As in Fig. 6c, but for fully saturated initial soil moisture conditions.

  • View in gallery

    Bias in simulated peak discharge as a function of bias in rainfall input for each satellite product and two different initial soil moisture conditions: (top) dry and (bottom) wet.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 73 73 10
PDF Downloads 56 56 11

Using High-Resolution Satellite Rainfall Products to Simulate a Major Flash Flood Event in Northern Italy

View More View Less
  • 1 Department of Land and Agroforest Environment, University of Padova, Padova, Italy, and Civil and Environmental Engineering, University of Connecticut, Storrs, Connecticut
  • | 2 Civil and Environmental Engineering, University of Connecticut, Storrs, Connecticut
  • | 3 Department of Land and Agroforest Environment, University of Padova, Padova, Italy
© Get Permissions
Full access

Abstract

Effective flash flood warning procedures are usually hampered by observational limitations of precipitation over mountainous basins where flash floods occur. Satellite rainfall estimates are available over complex terrain regions, offering a potentially viable solution to the observational coverage problem. However, satellite estimates of heavy rainfall rates are associated with significant biases and random errors that nonlinearly propagate in hydrologic modeling, imposing severe limitations on the use of these products in flood forecasting. In this study, the use of three quasi-global and near-real-time high-resolution satellite rainfall products for simulating flash floods over complex terrain basins are investigated. The study uses a major flash flood event that occurred during 29 August 2003 on a medium size mountainous basin (623 km2) in the eastern Italian Alps. Comparison of satellite rainfall with rainfall derived from gauge-calibrated weather radar estimates showed that although satellite products suffer from large biases they could represent the temporal variability of basin-averaged precipitation. Propagation of satellite rainfall through a distributed hydrologic model revealed that systematic error in rainfall was severely magnified when transformed to error in runoff under dry initial soil conditions. Simulation hydrographs became meaningful only after recalibrating the model for each satellite rainfall input separately. However, the unrealistic values of model parameters after recalibration show that this approach is erroneous and that model recalibration using satellite rainfall data should be treated with care. Overall, this study highlights the need for improvement of satellite rainfall retrieval algorithms in order to allow a more appropriate use of satellite rainfall products for flash flood applications.

Corresponding author address: Efthymios I. Nikolopoulos, Department of Land and Agroforest Environment, University of Padova, Campus di Agripolis Viale dell’Università, 16 35020, Legnaro, PD, Italy. E-mail: enikolop@engr.uconn.edu

Abstract

Effective flash flood warning procedures are usually hampered by observational limitations of precipitation over mountainous basins where flash floods occur. Satellite rainfall estimates are available over complex terrain regions, offering a potentially viable solution to the observational coverage problem. However, satellite estimates of heavy rainfall rates are associated with significant biases and random errors that nonlinearly propagate in hydrologic modeling, imposing severe limitations on the use of these products in flood forecasting. In this study, the use of three quasi-global and near-real-time high-resolution satellite rainfall products for simulating flash floods over complex terrain basins are investigated. The study uses a major flash flood event that occurred during 29 August 2003 on a medium size mountainous basin (623 km2) in the eastern Italian Alps. Comparison of satellite rainfall with rainfall derived from gauge-calibrated weather radar estimates showed that although satellite products suffer from large biases they could represent the temporal variability of basin-averaged precipitation. Propagation of satellite rainfall through a distributed hydrologic model revealed that systematic error in rainfall was severely magnified when transformed to error in runoff under dry initial soil conditions. Simulation hydrographs became meaningful only after recalibrating the model for each satellite rainfall input separately. However, the unrealistic values of model parameters after recalibration show that this approach is erroneous and that model recalibration using satellite rainfall data should be treated with care. Overall, this study highlights the need for improvement of satellite rainfall retrieval algorithms in order to allow a more appropriate use of satellite rainfall products for flash flood applications.

Corresponding author address: Efthymios I. Nikolopoulos, Department of Land and Agroforest Environment, University of Padova, Campus di Agripolis Viale dell’Università, 16 35020, Legnaro, PD, Italy. E-mail: enikolop@engr.uconn.edu

1. Introduction

Flash floods keep a high ranking in nature’s most devastating hazards and are responsible for significant damages and losses of life worldwide. The tremendous societal and economic impact of this hazard necessitates the development of effective flood warning systems in order to mitigate the risk. Flash flood warning systems integrate rainfall measurements, or estimates from remote sensing, with hydrological models that simulate the flood processes at basin scale, thus providing predictions of river discharges either distributed across various basin scales or as lumped values for the watershed outlet. Simpler and typically used operational flood warning schemes use the rainfall threshold approach (Martina et al. 2005; Georgakakos 1986) that is based on the derivation of the joint probability distribution of discharges at a given river section to rainfall accumulation values determined over the corresponding watershed for different antecedent soil moisture conditions and predefined runoff values that correspond to the discharge required to exceed bankfull conditions at the basin outlet (Carpenter et al. 1999). Currently, a number of flash flood warning–forecasting procedures that involve several data sources (gauges, radar, and numerical weather prediction models) and hydrologic models (lumped and distributed) exist over Europe and United States (see Alfieri et al. 2012 and Gourley et al. 2012 for an overview of these systems). Independent of the methodology or modeling approach used in the different flash flood warning systems, one thing that is common is that the accuracy of the warning is greatly impacted by the accuracy of rainfall estimates.

Rain gauge networks have been used as the primary source of precipitation estimates for over a century. However, while they provide a direct and most accurate measurement (relative to other sensors), they are associated with small sampling area (pointlike estimates). The usually low density of rain gauge networks, especially over complex terrain areas, which are particularly prone to flash floods, makes it difficult to accurately measure the dynamics of flash flood–inducing storms that are highly localized and variable in both space and time (Creutin and Borga 2003). The use of weather radar technology helped to address the coverage issue (to some extent) and advanced the precipitation monitoring by providing continuous and spatially distributed rainfall fields at high spatial (i.e., 1–4 km) and temporal resolution (i.e., 5 min to hourly). These advancements improved significantly the observational capabilities of flood warning systems, and have contributed to the predictive understanding of flash floods (Ogden et al. 2000; Delrieu et al. 2005; Vivoni et al. 2006; Borga et al. 2007; among others). However, radar-rainfall estimation is subject to a number of error sources that relate among others to radar calibration, variability of the relationship between radar reflectivity and rainfall rate, atmospheric attenuation, ground clutter, vertical reflectivity profile, and beam blockage effect (more details on the issue can be found in Krajewski and Smith 2002). The above error sources are particularly severe in real-time radar-rainfall applications (Gourley et al. 2010, 2011) and in radar observations over complex terrain areas. Furthermore, the high cost of radar systems proves a serious obstacle in providing adequate operational coverage of mountainous areas (Maddox et al. 2002) and for deploying and maintaining such systems in emerging economic regions of the world. Satellite precipitation estimates offer an alternative remote sensing data source that can compensate for some of the limitations of the other sensors. As pointed out by Scofield and Kuligowski (2003), among many other authors, satellite sensors provide global coverage, are uninhibited by mountains, and do not exhibit the spatial inconsistencies due to range-dependent errors and changes in radar beam height that affect weather radars.

The idea of using satellite rainfall retrievals for flash flood applications has been implemented at the National Environmental Satellite Data and Information Service (NESDIS) of the National Oceanic and Atmospheric Administration (NOAA) since the late 1970s, where data from the Geostationary Operational Environmental Satellite (GOES) have been used within the Interactive Flash Flood Analyzer (IFFA) system (Scofield and Oliver 1977; Scofield 1987; Borneman 1988; Kuligowski 1997). IFFA system uses satellite quantitative precipitation estimation (QPE) primarily to alert forecasters for the potential of heavy precipitation and the occurrence of flash floods (Scofield and Kuligowski 2003). While this is a very important step toward the use of satellite data in flash flood monitoring systems, to fully explore the satellite’s potential we need to directly incorporate satellite rainfall estimates into hydrologic models for simulating river flows, thus predicting the timing and magnitude of a potential flash flood. Although the idea of forcing hydrologic models with satellite QPE was suggested almost 30 years ago (Barret and Martin 1981), very few studies actually exist on the use of satellite rainfall retrievals for simulating flash floods (La Barbera et al. 1993; Borrell et al. 2007). Most of the currently published work that deals with the direct use of satellite rainfall estimates into hydrologic models involves (i) large basin scales (103–105 km2) and (ii) daily to monthly time scales (Guetter et al. 1996; Tsintikidis et al. 1999; Grimes and Diop 2003; Wilk et al. 2006; Artan et al. 2007; Su et al. 2008; Collischonn et al. 2008; Behrangi et al. 2011; Bitew and Gebremichael 2011; Wu et al. 2012). While these studies are very important and have helped to understand the potential use of satellite QPE for long-term simulations of hydrologic variables (e.g., streamflow or soil moisture), they do not provide evidence for the ability of satellite-based hydrologic predictions in the case of smaller-scale basins (<1000 km2) and short time scales (<12 h) at which flash floods take place (Borga et al. 2008).

Only a few studies exist on the subject (e.g., Bitew et al. 2012; Gourley et al. 2011; Nikolopoulos et al. 2010; Hossain and Anagnostou 2004). Nikolopoulos et al. (2010) and an earlier study by Hossain and Anagnostou (2004) used synthetic satellite data derived from a stochastic satellite rainfall error model to investigate the use of satellite rainfall estimates for flood simulations of small-scale basins (100–1200 km2) and 1–3-h time scales. They demonstrated potential in satellite-based flood applications and showed that the efficiency of model predictions depends on the product resolution and the scale of application. However, in their study synthetic satellite fields were simulated based on a long rainfall record that incorporated events of various rain intensities, and was not focused only on heavy precipitation events (representative of flash flood storms). As a result, the derived satellite rainfall error properties may not be fully representative of the satellite rainfall performance during flash floods. This hypothesis is supported by the recent work of Gourley et al. (2011), who showed that satellite hydrologic efficiency is magnitude dependent and performance is particularly limited in extreme rainfall intensity conditions as in the case of tropical storm Erin examined in their study. The present study overcomes this limitation by analyzing the hydrologic response based on actual satellite rainfall data and focusing on an extreme rainfall event that caused a major flash flood on a mountainous basin. The error analysis allows us to investigate the performance of satellite rainfall for the prediction of the basin runoff during the flash flood event.

The sudden nature of flash floods makes the use of satellite estimates for modeling and prediction an extremely challenging task. The high spatiotemporal variability of flash flood–inducing storms dictates the use of high-resolution observations in order to sufficiently capture these events. Infrared (IR)-based satellite QPE are currently available at high spatial (4 km) and temporal (15 min) resolution, which makes them suitable for monitoring storms that lead to flash floods, but their relationship to precipitation suffers from physical indirectness (Petty and Krajewski 1996), which results in low accuracy estimates. Microwave (MW)-based retrievals, on the other hand, afford a more direct inference to precipitation (Sapiano and Arkin 2009), but suffer because of infrequent sampling and spatial-resolution effects. Having recognized the strengths and limitations of both IR and MW estimates, researchers have moved toward the direction of merged (IR and MW) products that combine the potentially improved instantaneous rainfall estimates by MW with the high spatiotemporal resolution of IR observations. These recently developed merged products have opened new opportunities in global-scale flash flood applications (Yilmaz et al. 2010).

The objective of this study is to investigate the use of high-resolution merged satellite products for simulating flash floods over mountainous basins. Specifically, the work presented herein is focused on a major flash flood caused by an extreme storm event over a midsize complex terrain basin and aims to highlight issues related to 1) satellite rainfall estimation for orography-driven heavy rainfall events and 2) the incorporation of those estimates in distributed hydrologic models. Three global-scale high-resolution satellite products are used: (i) the Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis (TMPA) 3B42 version 6 (hereafter named 3B42) (Huffman et al. 2007), (ii) the Climate Prediction Center (CPC) morphing technique (CMORPH) product (Joyce et al. 2004), and (iii) the Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN) Cloud Classification System (CCS) (Hong et al. 2004). These satellite products are compared against radar-rainfall estimates in terms of precipitation estimation uncertainty and are used to force a distributed hydrologic model for flood simulation. The simulated hydrographs from satellite and radar are compared against measured hydrographs to assess the ability of satellite-based precipitation to represent the hydrologic response during this major flash flood storm case. The reader is reminded that the study is based on a single but major storm case, which may prevent us from drawing generalized conclusions, but, through a systematic investigation, presents evidence of the limitations in satellite flash flood applications that we should be aware of in developing hydrologically skillful rainfall products.

The paper is organized as follows: Section 2 describes the storm case, study area, and the data used in this study. Section 3 presents the comparison between the various satellite products and high-resolution gauge-calibrated radar-rainfall estimates (considered as ground reference). The hydrologic model simulations based on satellite and radar-rainfall forcing are analyzed in section 4 and conclusions from this study are summarized in section 5.

2. Study area and data

The area under study is within the Friuli–Venezia Giulia region in northeastern Italy (Fig. 1) and consists of highly complex terrain with elevation differences greater than 1700 m and slopes that range from 1° to 60°. The basin in which the flash flood took place is called Fella, which is a major left-hand tributary of the Tagliamento River with an area of approximately 623 km2 (Fig. 1), a mean altitude of 1140 m above sea level, and an average annual precipitation of 1920 mm (Borga et al. 2007). Land cover is dominated by broad-leaf and conifer forests and the geology of the area consists mainly of limestone with a spatial sequence of Silurian, Denovian, Triassic, Jurassic, and Creataecous formations north to south (Cucchi et al. 2000).

Fig. 1.
Fig. 1.

Digital elevation map of northeastern Italian region (highlighted in the inset map of Italy) showing the location of the OSMER radar and the outline boundaries of Fella River basin.

Citation: Journal of Hydrometeorology 14, 1; 10.1175/JHM-D-12-09.1

The heavy rainfall event that caused the Fella flash flood occurred on 29 August 2003. The storm lasted for approximately 12 h and resulted in losses of lives and damages close to 1 billion Euros (Tropeano et al. 2004) in the area of the upper Tagliamento River. The rainfall maxima were characterized by return periods in the range of 500–1000 yr for durations of 3, 6, and 12 h (Norbiato et al. 2008). The extreme rainfall amounts and the severity of flood response during this storm have made this event a case study for several past investigations on rainfall spatial dynamics and its relationship to flash flood response (Borga et al. 2007; Sangati and Borga 2009; Nikolopoulos et al. 2011).

The precipitation datasets analyzed in this study include radar and satellite rainfall fields. The radar-rainfall fields were derived from the reflectivity scans of a Doppler, dual-polarized C-band radar [Osservatorio Meteorologico Regionale (OSMER) radar station] located at Fossalon di Grado approximately 80 km south of the basin (Fig. 1). Radar-rainfall estimates were obtained based on horizontal reflectivity by applying a number of correction procedures for the following error contaminations: (i) partial beam occlusion, (ii) path attenuation, and (iii) wind drift. The RZ parameters used were a = 0.022 and b = 0.67 for R in mm h−1 and Z in mm6 m−3. These parameter values are used in this region for estimation of convective events and are similar to those used in the so-called Next Generation Weather Radar (NEXRAD) convective relationship (Ogden et al. 2000). An algorithm based on a three-step decision tree, based on Doppler velocity, clear-air echo statistics, and variance of differential reflectivity, was used to identify clutter-contaminated data in the polar volumes (see Borga et al. 2007 for details). Radar-rainfall estimates were compared with observations from 15 gauges available within the area (11 of which are within Fella). Radar–gauge statistical comparisons of hourly rainfall accumulations showed a generally good agreement with squared correlation equal to 0.73 and a 10% radar overestimation (Borga et al. 2007).

The satellite precipitation products include the CMORPH, PERSIANN, and 3B42, which are high-resolution quasi-global datasets. The 3B42 is a gauge-adjusted combination of two datasets: passive microwave (PMW) data from a variety of low-earth-orbiting satellites and IR data obtained from geosynchronous satellites. The PMW data are first calibrated using the combined TRMM Microwave Imager (TMI) and precipitation radar (PR) product and then used to calibrate the IR input. The PMW and IR are thus considered comparable with each other and are combined by using the PMW data where available and IR elsewhere (Huffman et al. 2007). The 3-hourly PMW–IR estimate is aggregated to monthly time scale and is merged with the monthly gauge data following the techniques described in Huffman et al. (1997). Finally, bias (defined as ratio) between this product and the nongauge-adjusted monthly average is used to scale the 3-hourly data and obtain the 3B42 version 6 precipitation estimates.

The CMORPH product is based on similar inputs as 3B42, but the original product is created on a 0.0727° grid and half-hourly time resolution. A histogram matching technique is used to match the different PMW records with TMI. At times and locations when PMW data are not available, the PMW estimates are propagated–interpolated using motion vectors derived from the IR data (see Joyce et al. 2004). An important feature of this combination method is that rainfall estimates do not rely on IR data. The PERSIANN-CCS algorithm extracts cloud features from IR geostationary satellite imagery and produces hourly rainfall estimates at 0.04° resolution. To account for the limitations of IR-based estimation algorithms mentioned previously, the PERSIANN-CCS algorithm segments an IR image in different cloud patches and utilizes an artificial neural network to enable the use of multiple brightness temperature–rainfall relationships in a single IR image. Table 1 summarizes the nominal spatial and temporal resolution of each precipitation dataset. Apart from the precipitation datasets, half-hourly discharge estimates were available at the outlet of Fella. More specifically, for the estimation of the hydrograph, a combination of postflood survey and hydraulic modeling was used, which limits the errors in extrapolation of the rating curves. Preflood and postflood geometry was also compared to limit the errors in discharge estimation, according to procedures discussed in Gaume and Borga (2008) and Marchi et al. (2009).

Table 1.

Nominal spatial and temporal resolution of radar and satellite precipitation data used in this study.

Table 1.

3. Rainfall error analysis

The satellite rainfall error analysis was performed over two different scales: basin scale (623 km2) and the whole radar domain (~45 239 km2). Comparative results at the basin scale are used to reveal satellite’s efficiency in estimating rainfall at hydrologic scales relevant to flash flood events. Analysis over the entire radar domain is carried out to examine if larger-scale statistics, as the ones typically used in satellite rainfall evaluation studies (see Anagnostou et al. 2010; Sapiano and Arkin 2009; among others), are representative of what we observe at the small basin scales (<1000 km2). Initially, the basin-averaged rainfall time series were calculated for each satellite dataset and compared against the radar rainfall, which has been used as the “reference” precipitation dataset throughout the analysis. Basin-averaged rainfall was calculated based on areal-weighted averaging that was applied to each product at its original spatial resolution. For the comparison over the entire radar domain, the highest-resolution radar pixels were aggregated to match the grid of each satellite product examined. Specifically, for a given satellite pixel, all radar pixels centered within the satellite pixel were aggregated to produce a radar value of the corresponding satellite pixel resolution. To match the datasets at the temporal domain, radar fields were aggregated over 1 and 3 h and were compared with the hourly PERSIANN and CMORPH, and 3-hourly 3B42, respectively. The calculated basin-averaged time series are presented in Fig. 2 and the bias (expressed as ratio of satellite to radar storm-total rainfall accumulation) with respect to the reference dataset are summarized in Table 2.

Fig. 2.
Fig. 2.

Time series of the Fella basin-averaged rainfall during the 29 Aug 2003 flash flood derived from (a) radar, CMORPH, and PERSIANN rainfall fields at hourly scale and (b) radar and 3B42 rainfall fields at 3-hourly scale.

Citation: Journal of Hydrometeorology 14, 1; 10.1175/JHM-D-12-09.1

Table 2.

Bias of total basin-averaged rainfall for satellite products before and after adjustment for mean field bias (MFB and MFBdyn).

Table 2.

Among the highest-resolution products (CMORPH and PERSIANN), PERSIANN has higher correlation (0.85) than CMORPH (0.70), but underestimates storm-total basin-averaged rainfall by more than an order of magnitude (bias = 0.08), while CMORPH underestimates by a factor of approximately 3 (bias = 0.32). The coarser product (3B42) has a bias of 0.22 and the highest correlation (0.91), but this is somewhat expected because it involves a coarser time scale. If PERSIANN and CMORPH data are aggregated to 3-h time scale the correlation coefficient increases to 0.97 and 0.9, respectively, because of the smoothing of variability due to averaging.

Although these results are based on a single event, and do not provide statistical significance, they represent an example of how satellite estimates performed over a mountainous basin for an extreme storm event that caused a major flash flood, and point out to the severe underestimation of all products. In fact, these findings are in general agreement with more extensive studies on the evaluation of satellite precipitation products over complex terrain that have shown underestimation of all three products (particularly for heavy precipitation events) over mountainous regions or highland areas (Bitew and Gebremichael 2009, 2011; Hirpa et al. 2010).

A positive outcome of this comparison is the high temporal correlation between radar and satellite rainfall estimates, which, along with the fact that all satellite products showed consistent underestimation, suggests that a bias adjustment of the satellite fields would potentially improve the representation of basin-average rainfall estimates. To investigate the benefit of a bias adjustment procedure, mean-field bias was derived for each satellite product. To achieve this, radar pixels were averaged to match the spatial resolution and orientation of each satellite data grid. This was done for the entire radar domain (120-km radius; Fig. 1) and for each radar field (at both hourly and 3-hourly scales). Mean-field bias (MFB) was calculated for the storm total and at each time step (MFBdyn) in order to compare the improvement in satellite estimates for (i) a common adjustment procedure that involves the use of a single bias factor (MFB) over satellite rainfall time series and (ii) a dynamic adjustment that uses a dynamic bias factor. Figure 3 shows the distribution (as box plot) of time-varying biases (MFBdyn) and superimposed are the storm-total biases (MFB) derived over Fella basin and over the radar domain.

Fig. 3.
Fig. 3.

Box plot showing for each satellite product the temporal distribution of mean field bias (MFBdyn) derived over the entire radar domain. Values for storm-total bias (MFB) for the radar domain and for the Fella basin are superimposed. Note that plus symbols correspond to outliers defined as values that are above (below) the 3rd (1st) quartile by 1.5 times the interquartile range.

Citation: Journal of Hydrometeorology 14, 1; 10.1175/JHM-D-12-09.1

Satellite estimates are underestimating (bias < 1) rainfall at all times while bias exhibits significant variability over time (especially for 3B42). The very low bias values shown for 3B42 may come as a surprise since this is a gauge-adjusted product and several studies on the evaluation of this product (Anagnostou et al. 2010; Sapiano and Arkin 2009; among others) report bias values closer to one. However, we should note that 3B42 is adjusted with gauge data on a monthly scale and the aforementioned studies evaluated this product over a long time record. Instead, in this study we examine a single extreme precipitation event, which cannot be represented by long-term error statistics. In fact this statement is supported by the recent work of AghaKouchak et al. (2011), who demonstrated how the error statistics for 3B42 (and also CMORPH and PERSIANN) vary with different rainfall thresholds.

An important point to note from Fig. 3 is the discrepancy between bias derived over Fella basin and bias derived considering the entire radar domain. This suggests that representation of bias in satellite rainfall products derived over large spatial domains is not adequate to represent smaller high-intensity areas (like in this case) that are of interest to flash flood hydrology. Apparently, other sources potentially related to (i) the spatial variability in the retrieval algorithm uncertainty, (ii) the local orographic enhancement of precipitation not accounted in satellite estimates, and (iii) resolution effects introduce variability in the error not captured by the mean-field bias. Comparison between MFB over Fella and values of MFBdyn (Fig. 3) shows that accounting for the spatial variability of error is more important (in the case examined) than the temporal variability. Table 2 summarizes the basin-scale storm-total bias derived for the adjusted satellite products. While there is an apparent improvement (relative to unadjusted) for all products, underestimation of basin-averaged precipitation remains severe (in the range of 50%–60%). A point to note from these results is that, except for the case of PERSIANN, the dynamic bias adjustment showed no improvement relative to the storm-total bias adjustment. This means that in the case examined, accounting for the temporal variability of error is less important than accounting for the spatial variability.

Variability of bias (over space and time) implies that bias in satellite rainfall estimates is rain-magnitude dependent. This is demonstrated in Fig. 4, where values from all rainy satellite pixels and the corresponding satellite pixel resolution radar-rainfall values are compared by plotting the quantiles of their respective distributions. Quantiles between radar-rainfall estimates and satellite products exhibit a strong linear dependence with correlation coefficients >0.9 for all products, which means that radar and satellite rainfall distributions are of the same type. However, there is an obvious offset in the slope from the 1–1 line (unbiased), demonstrating the underestimation of satellite estimates. An interesting feature to note is that for the case of CMORPH, satellite and radar quantiles are almost identical for rainfall values ≤3 mm h−1 and that a break in slope begins for higher rainfall rates. This shows that error in CMORPH estimates is magnitude dependent and this holds also for the other satellite products since quantiles do not form a single line over the entire spectrum of rainfall values.

Fig. 4.
Fig. 4.

Radar-rainfall quantiles vs satellite rainfall quantiles for (left) CMORPH, (middle) PERSIANN, and (right) 3B42. Solid line corresponds to the slope = 1.

Citation: Journal of Hydrometeorology 14, 1; 10.1175/JHM-D-12-09.1

Results from this section point to the necessity of using a dynamic scheme for bias correction, which would be a function of rainfall rate; however, this is an issue that can only be addressed in future improvements by algorithm developers. In the remainder of this paper we will focus on the modeling of the flash flood using the above satellite rainfall products and the two sets of bias-adjusted data that represent a correction methodology commonly used in many regional authorities that deal with the application of satellite rainfall products.

4. Hydrologic simulations

The hydrologic response of the Fella River basin during the flash flood event is simulated using the triangulated irregular network (TIN)-based Real-time Integrated Basin Simulator (tRIBS) model (Ivanov et al. 2004a,b; Vivoni et al. 2007). The tRIBS model is a distributed physics-based model that explicitly accounts for the spatial variability of land surface descriptors (terrain, soil, and vegetation), soil moisture, and atmospheric forcing. Infiltration is simulated in a sloped heterogeneous and anisotropic soil based on a kinematic approximation for unsaturated flow (Cabral et al. 1992; Garrote and Bras 1995). An adaptive multiple resolution approach based on TINs (Vivoni et al. 2004) is used to represent the complexity of the simulation domain. Runoff is generated at each computational element of the domain via a variety of mechanisms (infiltration excess, saturation excess, interflow, and groundwater exfiltration) depending on the soil saturation state.

Model calibration was performed using the shuffle complex evolution (SCE) optimization method (Duan et al. 1992) to minimize the mean-square error (MSE) between the observed and simulated hydrograph (at the outlet of Fella) during the flash flood event examined. While tRIBS requires a large number of parameters to be defined, we chose to calibrate only the parameters for which the flood hydrograph was most sensitive and to rely on values derived from the literature for the rest of the parameters. Several sensitivity runs (results not shown here) revealed that the parameters with the most significant impact on the flood hydrograph in terms of shape, magnitude, and timing include the (i) saturated hydraulic conductivity (Ksat), (ii) the conductivity exponential decay coefficient (f), and (iii) the anisotropy ratio (Ar, defined as the ratio of horizontal to vertical hydraulic conductivity). Model simulations were based on 0.5-km, hourly radar-rainfall fields. The length of the flash flood simulation was 30 h; the simulation started 4 h before the beginning of the storm and extended 14 h after the end of the event in order to fully capture the flood hydrograph (Fig. 5a). The calibrated model captures well the hydrologic response, matching very closely the observed peak discharge and preserving the general shape (rising–falling limb) of the hydrograph. Nikolopoulos et al. (2011) describes in more detail the model setup, calibration, and validation for the specific flash flood case in Fella. Specifically, the model was validated by comparing observed and simulated hydrographs at different subbasins of Fella exhibiting consistent overestimation of runoff with a relative error that ranged between 15% and 25% and Nash–Sutcliffe score from 0.7 to 0.87.

Fig. 5.
Fig. 5.

(a) Radar-simulated hydrograph (reference) against discharge observations at the outlet of Fella and (b)–(d) satellite-simulated hydrographs at the outlet of Fella based on (b) 3B42, (c) CMORPH, and (d) PERSIANN product before and after the bias adjustment. Note that scales are different between (a) and (b)–(d) to improve presentation of the results.

Citation: Journal of Hydrometeorology 14, 1; 10.1175/JHM-D-12-09.1

The satellite rainfall datasets (CMORPH, PERSIANN, and 3B42) before and after the bias adjustments were used to force the hydrologic model and the resulting hydrographs at the outlet of the basin were compared against the reference hydrograph that was based on the radar-rainfall input (Figs. 5b,c,d). The error in rainfall has a tremendous impact on the simulated hydrographs for all cases. The bias in total runoff and the relative error in peak discharge are summarized in Table 3 for all cases. Before the bias adjustment of the satellite rainfall products, there is an insignificant peak in the hydrographs (relative error ≥97% for all products) and even after the adjustment, only the CMORPH-based runoff simulation produces a clear signal of a flood hydrograph, which, however, underestimates peak discharge with a relative error of 89%. The Nash–Sutcliffe (N–S) score is close to 0 in all cases showing that the model has no ability in predicting discharge. Furthermore, in all cases hydrographs show a significant delay in time to peak (≥4 h). This delay is attributed to the activation of different runoff generation mechanisms owing to the lower-intensity rainfall forcing. Specifically, the reference case is associated with high rainfall amounts for which the faster overland runoff production mechanisms dominate the response, while for the satellite-based (low rainfall) cases, all of the rainfall infiltrates and runoff is produced through the slower subsurface mechanisms.

Table 3.

N–S score, bias in total runoff volume, and relative error in peak discharge (Ep) based on the comparison of radar-simulated and satellite-simulated hydrographs before and after the mean field bias adjustment. Note that simulated hydrographs were based on the radar-calibrated parameters and minus denotes underestimation.

Table 3.

Previous work on satellite-based hydrologic simulations of streamflow have shown that better results can be obtained when hydrologic models are calibrated based on satellite rainfall input (Hughes 2006; Artan et al. 2007; Bitew and Gebremichael 2011; Bitew et al. 2012). To investigate whether recalibration could potentially improve our results, the hydrologic model was calibrated for each satellite product and for both original (before adjustment) and adjusted satellite rainfall fields. The calibration of model parameters was automatic based on the SCE optimization method and following the same procedure used to calibrate the model with the reference rainfall. Figure 6 presents the simulated hydrographs for all cases and Table 4 summarizes the calculated error metrics. Results show that model recalibration brings a significant improvement in simulated hydrographs for all satellite products. For the case of unadjusted satellite input, the relative error in peak discharge in the case of CMORPH reduces to 29% (underestimation), while 3B42 and PERSIANN give a 31% and 48% lower peak. After bias adjustment and model recalibration all products perform extremely and equally well, with perhaps CMORPH outperforming the other two giving only 5% underestimation in both peak runoff and total runoff volume (for MFBdyn adjustment). A point to note is that while for rainfall, the MFBdyn adjustment did not reveal any significant advantage relative to MFB adjustment, for the case of runoff showed consistent improvement for almost all cases. For example, relative error in peak runoff improved from −23% (for MFB) to −4.8% (for MFBdyn) for the case of 3B42 and for CMORPH the N–S score improved from 0.67 (for MFB) to 0.9 (for MFBdyn). This shows that accounting for temporal variability of rainfall bias has important implications for the accuracy and predictive ability of hydrologic models.

Fig. 6.
Fig. 6.

Simulated hydrographs based on individual calibration of each satellite product (a) before and (b),(c) after bias adjustment.

Citation: Journal of Hydrometeorology 14, 1; 10.1175/JHM-D-12-09.1

Table 4.

As in Table 3, but for parameters calibrated separately for each satellite product before and after the adjustment.

Table 4.

Although at first glance these results give a very optimistic message about the use of satellite products, on second thought it is important to understand how recalibration improved the correspondence of satellite- and reference-driven hydrographs when the rainfall input is underestimated by approximately 40%–60% by the satellite products. A closer look at the calculated bias in total runoff (Table 3) reveals cases with overestimation (bias > 1), which means that while less rainfall is used as input, more runoff is produced than in the reference case. This is unrealistic and implies that recalibration has resulted essentially in “hydrologic systems” (modeling setups) representing completely different response mechanisms. In essence this means that comparing the response of these systems is like comparing the response of different watersheds. Table 5 presents the calibrated parameter values for each rainfall product (unadjusted and adjusted) to provide an idea of the variability in the estimates. As shown, variability in all parameters is large both among products and between unadjusted and adjusted cases. Because of the underestimation in satellite rainfall forcing one would expect to see hydraulic conductivity values that are lower than the reference in all cases (i.e., reduced infiltration of water). However, this is not always the case and we have values of hydraulic conductivity changing from 100 to 2 mm h−1. This is because of the interplay of the parameters that may lead to an increased runoff generation through increase of subsurface flow (e.g., high Ksat, low f, and high Ar). In these cases, contribution of interflow (infiltrated rainfall) and groundwater flow (initial water storage) end up generating more runoff than using only the infiltration-excess component (very low conductivity values).

Table 5.

Values for saturated hydraulic conductivity (Ksat, mm h−1), conductivity decay coefficient (f), and anisotropy ratio (Ar) obtained after individual calibration for each product. Note that the values shown correspond to the dominant soil class that occupies over 80% of the basin area.

Table 5.

The main point here is that bias in rainfall is compensated in the calibration parameters, causing parameters to lose their physical meaning. Even if the simulated hydrographs appear satisfactory in this case, the simulation results are not expected to be consistent under different conditions. To demonstrate this argument the simulations were repeated with the same calibrated parameters of the adjusted satellite rainfall input (for MFBdyn) but for different initial soil moisture conditions. Specifically, the hydrologic response was investigated under the scenario that initial wetness conditions over the whole basin are fully saturated. This was done within the model by setting the level of groundwater table equal to the soil surface. The results (Fig. 7) suggest exactly what was speculated. While in Fig. 5 the hydrographs have a similar behavior, once the initial conditions changed, the nonrealistic parameterization could not consistently represent the radar-derived hydrologic response. While there is a large body of literature that have investigated and demonstrated the effect of uncertainty in calibration parameters in hydrologic predictions (Andreassian et al. 2001; Brath et al. 2004; Clark and Vrugt 2006; Baroni et al. 2010; among others), many satellite-based hydrologic studies seem to overlook this importance and rely on unconditional calibration of models based on satellite rainfall input. These findings highlight the fact that recalibration should be performed under physical considerations.

Fig. 7.
Fig. 7.

As in Fig. 6c, but for fully saturated initial soil moisture conditions.

Citation: Journal of Hydrometeorology 14, 1; 10.1175/JHM-D-12-09.1

Our interpretation is that for improving satellite rainfall hydrologic applications, recalibration of model parameters to each rainfall product is not the correct path to follow. As shown earlier, errors in satellite rainfall estimates are not systematic (vary in space and time), while recalibration leads to unphysical parameters, which combined can lead to inconsistent hydrologic representation. Instead, model parameters should embed physical characteristics of the basin and the focus must be on improving satellite rainfall estimates at small basin scales. An important question along this line is “How much improvement do we need in the satellite estimates to achieve satisfactory flash flood modeling results?”

To address this question we used this case study to carry out a series of simulations for different levels of bias in the satellite products examined. The model was then forced with the satellite rainfall fields scaled by the different bias factors. The bias levels examined ranged from 0.5 to 1 (unbiased) and the exercise was performed for two different initial soil moisture conditions that correspond to dry (control initial conditions) and wet (fully saturated). Results are summarized as a function of bias in peak runoff versus bias in rainfall (Fig. 8). The following points are summarized from this exercise. Bias in rainfall magnifies as it propagates through the nonlinear transformation of rainfall to runoff. The degree of nonlinearity in the error propagation is different among products and for different initial conditions. CMORPH and 3B42 behave similarly and with a less nonlinear error propagation than PERSIANN, which exhibits a distinctly different behavior. In addition, comparison of the results for the different initial conditions suggests that nonlinearity in the error propagation decreases with increasing soil wetness. Namely, the drier the soil is, the greater the error in runoff will be due to error in rainfall. It is interesting to note that even for the case of unbiased input (bias = 1), error in simulated peak remains significant for 3B42 and CMORPH while PERSIANN performs much better with almost no bias in peak runoff for the dry conditions and an approximate 10% overestimation under fully saturated conditions. These differences are attributed to the high correlation of PERSIANN (in temporal domain) and its high spatial resolution (relative to the other products), which permits better representation of the spatial organization of rainfall over the basin. Deriving an “allowable” rainfall bias threshold in order to have meaningful simulations is not a straightforward answer since as we argued above, this relates to the satellite product and initial basin wetness conditions, and it is expected that will also dependent on the storm event and basin scale examined (Nikolopoulos et al. 2010). For this particular flash flood case examined, results suggest that only PERSIANN estimates with ≤15% bias would provide reasonable representation of the flood peak.

Fig. 8.
Fig. 8.

Bias in simulated peak discharge as a function of bias in rainfall input for each satellite product and two different initial soil moisture conditions: (top) dry and (bottom) wet.

Citation: Journal of Hydrometeorology 14, 1; 10.1175/JHM-D-12-09.1

5. Discussion and concluding remarks

This study investigated the use of high-resolution satellite precipitation products for simulating a major flash flood caused by heavy rainfall over a mountainous basin in northeastern Italian Alps. Three high-resolution global-scale satellite products—namely, 3B42, CMORPH, and PERSIANN—were compared against radar-rainfall observations for the storm event that caused the flooding. Results showed that satellite precipitation estimates suffered from large bias, but were able to represent the temporal variability of basin-averaged rainfall. The mean-field bias derived from the comparison over the whole radar domain was not able to explain the observed bias over the study basin.

The hydrologic simulations forced with satellite rainfall input could not capture the basin’s hydrologic response during the flash flood event. Even after adjusting satellite rainfall fields for the mean-field bias, the systematic error in rainfall remained significant and was magnified severely in terms of bias in runoff parameters (volume and peak runoff). Recalibration of model parameters using precipitation derived from each satellite product showed that it can account for this bias and improve the modeling results. However, this should be treated with caution because it was also shown that to match the reference hydrograph, parameter values were forced to take unrealistic values, thus losing their physical meaning and, consequently, the ability to consistently represent the hydrologic response in the basin.

Further analysis on the propagation of satellite rainfall error in runoff response revealed that systematic error magnifies nonlinearly. The degree of nonlinearity depends on the satellite product characteristics and decreases with increasing soil wetness. For the case of unbiased rainfall forcing, PERSIANN outperforms 3B42 and CMORPH. Overall, for the basin scale examined it was shown that a target of less than 15% bias is required for the highest-spatial-resolution product (PERSIANN) to obtain meaningful simulations of the flood response. 3B42 and CMORPH gave large errors even in the unbiased case, which suggests that the resolution effect is significant for the basin scale examined. We therefore believe that satellite-based flash flood applications should rely on the improvement of the high-resolution satellite rainfall estimates through 1) more accurate retrieval algorithms that would potentially account for surface emissivity and storm structure (spatial and vertical) effects and/or 2) advanced error-correction schemes that can rely on statistical (e.g., characterize the spatial variability of error) or physical considerations (e.g., through high-resolution numerical weather prediction) of the satellite error.

Through investigation of a single case study, this work, provided an understanding about the use of satellite rainfall estimates for modeling flash floods. Positive and negative aspects on the use of currently available quasi-global satellite rainfall estimates were highlighted. While generalized conclusions cannot be drawn, we hope however to have pointed out important issues that require particular attention in the context of satellite rainfall and hydrology. Satellite rainfall products were examined only from the point of rainfall estimation/representation and no discussion, or comparison, was made on the generation of the products in real time (i.e., latency issue). We treated satellite products from a retrospective point of view as we believe that careful evaluation and identification of issues related to satellite-based hydrologic predictions should be the first step before moving into the real-time operational use. We therefore acknowledge that many of the proposed avenues for improving satellite estimates (e.g., dynamic error correction schemes to account for space/time variability of error) are currently very difficult to be carried out in real-time operation.

Acknowledgments

This work was supported by a NASA Precipitation Measurement Mission Award (NNX07AE31G). We thankfully acknowledge the three anonymous reviewers for their very constructive comments that helped us improve our paper.

REFERENCES

  • AghaKouchak, A., , Behrangi A. , , Sorooshian S. , , Hsu K. , , and Amitai E. , 2011: Evaluation of satellite-retrieved extreme precipitation rates across the central United States. J. Geophys. Res., 116, D02115, doi:10.1029/2010JD014741.

    • Search Google Scholar
    • Export Citation
  • Alfieri, L., , Salamon P. , , Pappenberger F. , , Wetterhall F. , , and Thielen J. , 2012: Operational early warning systems for water-related hazards in Europe. Environ. Sci. Policy, 21, 3549.

    • Search Google Scholar
    • Export Citation
  • Anagnostou, E. N., , Maggioni V. , , Nikolopoulos E. I. , , Meskele T. , , Hossain F. , , and Papadopoulos A. , 2010: Benchmarking high-resolution global satellite rainfall products to radar and rain-gauge rainfall estimates. IEEE Trans. Geosci. Remote Sens., 48, 16671683, doi:10.1109/TGRS.2009.2034736.

    • Search Google Scholar
    • Export Citation
  • Andreassian, V., , Perrin C. , , Michel C. , , Usartsanchez I. , , and Lavabre J. , 2001: Impact of imperfect rainfall knowledge on the efficiency and the parameters of watershed models. J. Hydrol., 250 (1–4), 206223.

    • Search Google Scholar
    • Export Citation
  • Artan, G., , Gadain H. , , Smith J. L. , , Asante K. , , Bandaragoda C. J. , , and Verdin J. P. , 2007: Adequacy of satellite derived rainfall data for streamflow modeling. Nat. Hazards, 43, 167185.

    • Search Google Scholar
    • Export Citation
  • Baroni, G., , Facchi A. , , Gandolfi C. , , Ortuani B. , , Horeschi D. , , and van Dam J. C. , 2010: Uncertainty in the determination of soil hydraulic parameters and its influence on the performance of two hydrological models of different complexity. Hydrol. Earth Syst. Sci., 14, 251270.

    • Search Google Scholar
    • Export Citation
  • Barret, E. C., , and Martin D. W. , 1981: The Use of Satellite Data in Rainfall Monitoring. Academic Press, 340 pp.

  • Behrangi, A., , Khakbaz B. , , Jaw T. C. , , AghaKouchak A. , , Hsu K. , , and Sorooshian S. , 2011: Hydrologic evaluation of satellite precipitation products over a mid-size basin. J. Hydrol., 397 (3–4), 225237.

    • Search Google Scholar
    • Export Citation
  • Bitew, M. M., , and Gebremichael M. , 2009: Evaluation through independent measurements: Complex terrain and humid tropical region in Ethiopia. Satellite Rainfall Applications for Surface Hydrology, M. Gebremichael and F. Hossain, Eds., Springer, 205–214.

  • Bitew, M. M., , and Gebremichael M. , 2011: Assessment of satellite rainfall products for streamflow simulation in medium watersheds of the Ethiopian highlands. Hydrol. Earth Syst. Sci., 15, 11471155.

    • Search Google Scholar
    • Export Citation
  • Bitew, M. M., , Gebremichael M. , , Ghebremichael L. T. , , and Bayissa Y. A. , 2012: Evaluation of high-resolution satellite rainfall products through streamflow simulation in a hydrological modeling of a small mountainous watershed in Ethiopia. J. Hydrometeor., 13, 338350.

    • Search Google Scholar
    • Export Citation
  • Borga, M., , Boscolo P. , , Zanon F. , , and Sangati M. , 2007: Hydrometeorological analysis of the 29 August 2003 flash flood in the eastern Italian Alps. J. Hydrometeor., 8, 10491067.

    • Search Google Scholar
    • Export Citation
  • Borga, M., , Gaume E. , , Creutin J. D. , , and Marchi L. , 2008: Surveying flash floods: Gauging the ungauged extremes. Hydrol. Processes, 22, 38833885.

    • Search Google Scholar
    • Export Citation
  • Borneman, R., 1988: Satellite rainfall estimating program of the NOAA/NESDIS Satellite Analysis Branch. Natl. Wea. Dig., 13 (2), 715.

    • Search Google Scholar
    • Export Citation
  • Borrell, V. E., , Dartus D. , , and Alquier M. , 2007: Forecasting flash floods in ungauged basins with satellite data. Proc. PUB Kick-Off Meeting, Brasilia, Brazil, IAHS, 249–257.

  • Brath, A., , Montanari A. , , and Toth E. , 2004: Analysis of the effects of different scenarios of historical data availability on the calibration of a spatially-distributed hydrological model. J. Hydrol., 291, 232253.

    • Search Google Scholar
    • Export Citation
  • Cabral, M. C., , Garrote L. , , Bras R. L. , , and Entekhabi D. , 1992: A kinematic model of infiltration and runoff generation in layered and sloped soils. Adv. Water Resour., 15, 311324.

    • Search Google Scholar
    • Export Citation
  • Carpenter, T. M., , Sperfslage J. A. , , Georgakakos K. P. , , Sweeney T. , , and Fread D. L. , 1999: National threshold runoff estimation utilizing GIS in support of operational flash flood warning systems. J. Hydrol., 224, 2144.

    • Search Google Scholar
    • Export Citation
  • Clark, M. P., , and Vrugt J. A. , 2006: Unraveling uncertainties in hydrologic model calibration: Addressing the problem of compensatory parameters. Geophys. Res. Lett., 33, L06406, doi:10.1029/2005GL025604.

    • Search Google Scholar
    • Export Citation
  • Collischonn, B., , Collischonn W. , , and Tucci C. E. M. , 2008: Daily hydrological modeling in the Amazon basin using TRMM rainfall estimates. J. Hydrol., 360, 207216.

    • Search Google Scholar
    • Export Citation
  • Creutin, J. D., , and Borga M. , 2003: Radar hydrology modifies the monitoring of flash-flood hazard. Hydrol. Processes, 17, 14531456.

  • Cucchi, F., , Piano S. , , Marinetti E. , , Massari G. , , Oberti S. , , and Zini L. , 2000: Studies for the realization of the hydrological map of Friuli-Venezia Giulia. Ipogea, 3, 5771.

    • Search Google Scholar
    • Export Citation
  • Delrieu, G., and Coauthors, 2005: The catastrophic flash-flood event of 8–9 September 2002 in the Gard region, France: A first case study for the Cèvennes–Vivarais Mediterranean Hydrometeorological Observatory. J. Hydrometeor., 6, 3452.

    • Search Google Scholar
    • Export Citation
  • Duan, Q., , Sorooshian S. , , and Gupta V. K. , 1992: Effective and efficient global optimization for conceptual rainfall-runoff models. Water Resour. Res., 28, 10151031.

    • Search Google Scholar
    • Export Citation
  • Garrote, L., , and Bras R. L. , 1995: A distributed model for real-time flood forecasting using digital elevation models. J. Hydrol., 167, 279306.

    • Search Google Scholar
    • Export Citation
  • Gaume E., , and Borga M. , 2008: Post-flood field investigations in upland catchments after major flash floods: Proposal of a methodology and illustrations. J. Flood Risk Manage., 1 (4), 175189.

    • Search Google Scholar
    • Export Citation
  • Georgakakos, K. P., 1986: A generalized stochastic hydrometeorological model for flood and flash-flood forecasting: 1. Formulation. Water Resour. Res., 22, 20832095, doi:10.1029/WR022i013p02083.

    • Search Google Scholar
    • Export Citation
  • Gourley, J. J., , Hong Y. , , Flamig Z. L. , , Li L. , , and Wang J. , 2010: Intercomparison of rainfall estimates from radar, satellite, gauge, and combinations for a season of record rainfall. J. Appl. Meteor. Climatol., 49, 437452.

    • Search Google Scholar
    • Export Citation
  • Gourley, J. J., , Hong Y. , , Flamig Z. L. , , Wang J. , , Vergara H. , , and Anagnostou E. N. , 2011: Hydrologic evaluation of rainfall estimates from radar, satellite, gauge, and combinations on Ft. Cobb basin, Oklahoma. J. Hydrometeor., 12, 973988.

    • Search Google Scholar
    • Export Citation
  • Gourley, J. J., , Erlingis J. M. , , Hong Y. , , and Wells E. B. , 2012: Evaluation of tool used for monitoring and forecasting flash floods in the United States. Wea. Forecasting, 27, 158173.

    • Search Google Scholar
    • Export Citation
  • Grimes, D. I. F., , and Diop M. , 2003: Satellite-based rainfall estimation for river flow forecasting in Africa. I: Rainfall estimates and hydrological forecasts. Hydrol. Sci. J., 48, 567584.

    • Search Google Scholar
    • Export Citation
  • Guetter, A. K., , Georgakakos K. P. , , and Tsonis A. A. , 1996: Hydrologic applications of satellite data: 2. Flow simulation and soil water estimates. J. Geophys. Res., 101 (D21), 26 52726 538.

    • Search Google Scholar
    • Export Citation
  • Hirpa, F. A., , Gebremichael M. , , and Hopson T. , 2010: Evaluation of high resolution satellite precipitation products over very complex terrain in Ethiopia. J. Appl. Meteor. Climatol., 49, 10441051.

    • Search Google Scholar
    • Export Citation
  • Hong, Y., , Hsu K.-L. , , Sorooshian S. , , and Gao X. , 2004: Precipitation estimation from remotely sensed imagery using an Artificial Neural Network Cloud Classification System. J. Appl. Meteor., 43, 18341852.

    • Search Google Scholar
    • Export Citation
  • Hossain, F., , and Anagnostou E. N. , 2004: Assessment of current passive microwave and infrared based satellite rainfall remote sensing for flood prediction. J. Geophys. Res., 109, D07102, doi:10.1029/2003JD003986.

    • Search Google Scholar
    • Export Citation
  • Huffman, G. J., and Coauthors, 1997: The Global Precipitation Climatology Project (GPCP) combined precipitation dataset. Bull. Amer. Meteor. Soc., 78, 520.

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

    • Search Google Scholar
    • Export Citation
  • Hughes, D. A., 2006: Comparison of satellite rainfall data with observations from gauging station networks. J. Hydrol., 327 (3–4), 399410.

    • Search Google Scholar
    • Export Citation
  • Ivanov, V. Y., , Vivoni E. R. , , Bras R. L. , , and Entekhabi D. , 2004a: Catchment hydrologic response with a fully distributed triangulated irregular network model. Water Resour. Res., 40, W11102, doi:10.1029/2004WR003218.

    • Search Google Scholar
    • Export Citation
  • Ivanov, V. Y., , Vivoni E. R. , , Bras R. L. , , and Entekhabi D. , 2004b: Preserving high-resolution surface and rainfall data in operational-scale basin hydrology: A fully-distributed physically-based approach. J. Hydrol., 298 (1–4), 80111.

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

    • Search Google Scholar
    • Export Citation
  • Krajewski, W. F., , and Smith J. A. , 2002: Radar hydrology: Rainfall estimation. Adv. Water Resour., 25, 13871394.

  • Kuligowski, R. J., 1997: An overview of National Weather Service quantitative precipitation estimates (QPE). Techniques Development Laboratory Office Note 97-4, 27 pp.

  • La Barbera, P., , Lanza L. , , and Siccardi F. , 1993: Flash flood forecasting based on multisensor information. Proc. Symp. on Extreme Hydrological Events: Precipitation, Floods and Droughts, Yokohama, Japan, IAHS, 21–32.

  • Maddox, R. A., , Zhang J. , , Gourley J. J. , , and Howard K. W. , 2002: Weather radar coverage over the contiguous United States. Wea. Forecasting, 17, 927934.

    • Search Google Scholar
    • Export Citation
  • Marchi, L., and Coauthors, 2009: Comprehensive post-event survey of a flash flood in western Slovenia: Observation strategy and lessons learned. Hydrol. Processes, 23, 37613770.

    • Search Google Scholar
    • Export Citation
  • Martina, M. L. V., , Todini E. , , and Libralon A. , 2005: A Bayesian decision approach to rainfall thresholds based flood warning. Hydrol. Earth Syst. Sci. Discuss., 2, 26632706.

    • Search Google Scholar
    • Export Citation
  • Nikolopoulos, E. I., , Anagnostou E. N. , , Hossain F. , , Gebremichael M. , , and Borga M. , 2010: Understanding the scale relationships of uncertainty propagation of satellite rainfall through a distributed hydrologic model. J. Hydrometeor., 11, 520532.

    • Search Google Scholar
    • Export Citation
  • Nikolopoulos, E. I., , Anagnostou E. N. , , Borga M. , , Vivoni E. R. , , and Papadopoulos A. , 2011: Sensitivity of a mountain basin flash flood to initial wetness conditions and rainfall variability. J. Hydrol., 402 (3–4), 165178.

    • Search Google Scholar
    • Export Citation
  • Norbiato, D., , Borga M. , , Esposti S. D. , , Gaume E. , , and Anquetin S. , 2008: Flash flood warning based on rainfall thresholds and soil moisture conditions: An assessment for gauged and ungauged basins. J. Hydrol., 362, 274290.

    • Search Google Scholar
    • Export Citation
  • Ogden, F. L., , Sharif H. O. , , Senarath S. U. S. , , Smith J. A. , , Baeck M. L. , , and Richardson J. R. , 2000: Hydrologic analysis of the Fort Collins, Colorado flash flood of 1997. J. Hydrol., 228, 82100.

    • Search Google Scholar
    • Export Citation
  • Petty, G., , and Krajewski W. F. , 1996: Satellite estimation of precipitation over land. Hydrol. Sci. J., 41, 433451.

  • Sangati, M., , and Borga M. , 2009: Influence of rainfall spatial resolution on flash flood modelling. Nat. Hazards Earth Syst. Sci., 9, 575584.

    • Search Google Scholar
    • Export Citation
  • Sapiano, M. R. P., , and Arkin P. A. , 2009: An intercomparison and validation of high resolution satellite precipitation estimates with 3-hourly gauge data. J. Hydrometeor., 10, 149166.

    • Search Google Scholar
    • Export Citation
  • Scofield, R. A., 1987: The NESDIS operational convective precipitation estimation technique. Mon. Wea. Rev., 115, 17731792.

  • Scofield, R. A., , and Oliver V. J. , 1977: A scheme for estimating convective rainfall from satellite imagery. NOAA Tech. Memo. NESS 86, 47 pp.

  • Scofield, R. A., , and Kuligowski R. J. , 2003: Status and outlook of operational satellite precipitation algorithms for extreme-precipitation events. Wea. Forecasting, 18, 10371051.

    • Search Google Scholar
    • Export Citation
  • Su, F., , Hong Y. , , and Lettenmaier D. P. , 2008: Evaluation of TRMM Multisatellite Precipitation Analysis (TMPA) and its utility in hydrologic prediction in the La Plata basin. J. Hydrometeor., 9, 622640.

    • Search Google Scholar
    • Export Citation
  • Tropeano, D., , Turconi L. , , and Sanna S. , 2004: Debris flow triggered by the 29 August 2003 cloudburst in Val Canale, eastern Italian Alps. Proc. Int. Symp. INTERPRAEVENT 2004, Riva del Garda, Italy, INTERPRAEVENT, 121–132.

  • Tsintikidis, D., , Georgakakos K. P. , , Artan G. A. , , and Tsonis A. A. , 1999: A feasibility study on mean areal rainfall estimation and hydrologic response in the Blue Nile region using METEOSAT images. J. Hydrol., 221, 97116.

    • Search Google Scholar
    • Export Citation
  • Vivoni, E. R., , Ivanov V. Y. , , Bras R. L. , , and Entekhabi D. , 2004: Generation of triangulated irregular networks based on hydrological similarity. J. Hydrol. Eng., 9, 288303.

    • Search Google Scholar
    • Export Citation
  • Vivoni, E. R., , Entekhabi D. , , Bras R. L. , , Ivanov V. Y. , , Van Horne M. P. , , Grassotti C. , , and Hoffman R. N. , 2006: Extending the predictability of hydrometeorological flood events using radar rainfall nowcasting. J. Hydrometeor., 7, 660677.

    • Search Google Scholar
    • Export Citation
  • Vivoni, E. R., , Entekhabi D. , , and Hoffman R. N. , 2007: Error propagation of radar rainfall nowcasting fields through a fully-distributed flood forecasting model. J. Appl. Meteor. Climatol., 46, 932940.

    • Search Google Scholar
    • Export Citation
  • Wilk, J., , Kniveton D. , , Andersson L. , , Layberry R. , , Todd M. C. , , Hughes D. , , Ringrose S. , , and Vanderpost C. , 2006: Estimating rainfall and water balance over the Okavango River Basin for hydrological applications. J. Hydrol., 331, 1829.

    • Search Google Scholar
    • Export Citation
  • Wu, H., , Adler R. F. , , Hong Y. , , Tian Y. , , and Policelli F. , 2012: Evaluation of global flood detection using satellite-based rainfall and a hydrologic model. J. Hydrometeor., 13, 12681284.

    • Search Google Scholar
    • Export Citation
  • Yilmaz, K. K., , Adler R. F. , , Tian Y. , , Hong Y. , , and Pierce H. F. , 2010: Evaluation of a satellite-based global flood monitoring system. Int. J. Remote Sens., 31, 37633782.

    • Search Google Scholar
    • Export Citation
Save