Runoff calculations using different precipitation inputs

It was possible to apply the TACD model with very good success: The Reff (Q) amounted to 0.94 and 0.85 and Reff (log Q) amounted to 0.99 and 0.90 during the calibration period for the Brugga basin and the St. Wilhelmer Talbach basin, respectively. The statistical measures for the model validation period showed also clearly the suitability of the TACD for runoff calculation: Reff (Q) amounted to 0.80 and 0.85 and Reff (log Q) amounted to 0.83 and 0.87 for the Brugga basin and the St. Wilhelmer Talbach basin, respectively. It is important to note that the model was calibrated and validated using ground station data. Radar data were only available for the three selected events.

Comparing the model performance for the three investigated events (Figure 16.4, Table 16.3) demonstrates the importance of the spatial and temporal distribution of the precipitation input for flood modelling. In general, the simulated discharge using the IDW-elevation method matches the observed discharge better than the simulated discharge using radar data as precipitation input data. The falling limbs are not modelled well, independent of the precipitation input. This failing in capturing accurately

Table 16.3 Observed and simulated peak discharges of the three investigated events [m3 s-1]

Event

Brugga

St. Wilhelmer Talbach

Observed

IDW-elev.

Radar

Observed IDW-elev. Radar

I

1.4

1.2

0.9

0.6 0.4 0.4

II

8.0

7.4

8.0

4.2 2.9 3.2

III

6.9

7.3

8.4

2.3 2.5 3.3

the falling limbs is exceptional for the three examined events from the validation period; others were modelled much better as indicated in the high statistical measures of model efficiency values of more than 0.8 (see above).

Event I is a typical small event during summer times (Figure 16.4a). This type of event might occur several times during a year. The basin precipitation calculated by radar data is lower compared to that using the IDW-elevation method (Table 16.2). This resulted in a smaller simulated event (Table 16.3). Remarkable are the large differences in the timing of the simulated peak discharges. The differences in the temporal variability of rainfall cause an earlier peak discharge using radar data than the IDW method.

Event II occurred after a two-week rainless low-flow period and was followed by a second event for which no radar data were available (Figure 16.4b). The radar data had higher basin precipitation values than the IDW-elevation method (Table 16.2); this resulted in a

Figure 16.5 Results of simulating event II at the outlet of the sub-basin St. Wilhelmer Talbach using the TACD model and applying the two different precipitation scenarios

larger simulated flood. Hence, in this case, simulation results are better using radar data (Table 16.3). Also, the subsequent event is simulated differently even if the same rainfall input was applied. This is caused by the higher antecedent moisture conditions for this event during the model run using the higher radar rainfall input for the first event.

For event III, again the radar data led to higher basin precipitation that resulted in a larger simulated flood (Figure 16.4c). However, the 2 mm more precipitation using radar data (Table 16.2) could not explain the larger simulated response alone. Thus, the temporal distribution with larger intensities for the radar rainfall (Figure 16.2b) and the spatial pattern seems to be responsible for the differently simulated runoff. The radar data had much higher rain intensities in an area in the southwest of the Brugga basin with 55-75 mm for the whole of the event, compared to 45-52 mm for the precipitation input using the IDW-elevation method. At this part of the basin, a lot of fast responding areas, in particular saturated areas, are located (Uhlenbrook etal. 2003b), which caused the increased runoff response modelled by the process-oriented model.

The simulation of the event II at the sub-basin St. Wilhelmer Talbach (Figure 16.5) highlights the power of rainfall radar data for spatially very unequally distributed events. The largest rainfall occurred within the St. Wilhelmer Talbach (Figure 16.3b) and could not be observed adequately with the existing monitoring network (see Figure 16.1). This resulted in a significant underestimation of the flood at the outlet of the St.

Wilhelmer Talbach basin. The radar observed the event much better and, consequently, the TACD model was able to simulate the flood discharge more closely to the observed discharge. Again, the different simulations of the second event using the two rainfall scenarios are caused by the different antecedent moisture conditions owing to the different rainfall inputs of the preceding event. Unfortunately, no radar data were available for the second peak of event II. But it is plausible that the underestimation of this convective event is also caused by an inappropriate rainfall input.

The comparison of the simulated peak discharges (Table 16.3) shows that compared to the measured peaks, both input data sets result in similar responses: either both data sets under- or overestimate the peak. When high rainfall intensities at areas with quick runoff responses (according to the spatial delineation runoff generation areas used in the TACD model, see Section 16.3.2) are determined (e.g. during event III), the more extreme rainfall values associated with radar data cause higher errors in peak discharge simulation. Because of the more smoothed rainfall pattern using ground station data and the IDW-elevation method for regionalisation, the deviations of the peak discharge predictions are smaller. However, if high intensity rain cells were not recorded by the ground station network, in particular, at locations with many direct runoff generation areas (the St. Wilhelmer Talbach sub-basin during event II, Figure 16.3), the simulations using radar data outperform those using the IDW-elevation method also for the whole Brugga basin (Table 16.3).

16.5 DISCUSSION

The importance of an adequate representation of the spatial and temporal distribution of convective precipitation for runoff modelling can only be examined by a well-validated process-oriented model. Owing to previous investigations (see Section 16.3), the TACD model is qualified to study the impact of precipitation variabilities in the mountainous Brugga basin. Its large significance was clearly demonstrated. These findings are in line with other studies in heterogeneous catchments (Woods etal. 2000; Ogden etal. 2000; Syed etal. 2003). However, these insights are restricted to the examined convective storm events. Events that are (partly) produced by snow melt were not considered, but a similar importance of the precipitation input (rain and snow melt) is very likely.

Small differences in total basin rainfall can cause large differences in the simulated hydrographs due to non-linear feedback mechanisms, as shown for event I. However, depending on the position and the movement of the rain cell and the runoff travel times to the basin outlet, the importance of the rain distribution can differ. This was learnt by comparing the Brugga basin and its sub-basin.

The contribution of spatial highly resoluted radar data is the ability of capturing highly localised rain cells, which are not well represented by a ground station network. This was clearly demonstrated for event II at the St. Wilhelmer Talbach basin. The rain cell was too small to be captured adequately by the gauging network, even if this can be attributed as dense with up to seven stations within or near the examined basin. However, the use of radar can also lead to runoff simulations, which are worse than the simulations using the ordinary gauging network only (see simulations for the Brugga basin, Figure 16.4). This is caused mainly by two reasons: (i) The model was not trained (calibrated) with input from radar data, but with input using the IDW-elevation method. This is caused by the fact that the model run continuously for several months and radar data were only available for the three events. (ii) Because of the technical problems in 1998 (see Section 16.2.2) the radar data resolution is relatively coarse and it tends to overestimate high rain intensities. Owing to non-linearities during flood formation processes (e.g. Grayson etal. 1997) and the respective mathematical description in the TACD model, relatively small differences in precipitation input can cause large differences for the modelled flood. The latter was shown clearly for event I.

The used radar device is located at the highest part of the catchment (Figure 16.1). However, this is not the case in many mountainous regions and the radar measurement can be affected by factors like beam blocking as demonstrated, for instance, by Andrieu et al. (1997). They showed that beam blocking can be corrected using digital terrain models and took into account vertical variations in radar reflectivity for providing satisfactory range-dependent corrections. But these data are not available in many studies. Thus, the comparison and adjustment of the radar data with highly resoluted ground station data is essential (e.g. Creutin et al. 1997). In this study, also ground stations that are not situated directly within or near the catchment were used for radar data adjustment. This made it possible to consider a wider range of rain intensities for data adjustment, and hence, made the calibration more robust.

For the used adjustment method (Section 16.2.3), only the parameters a and p were calibrated to optimise the Zradar/RZ/R relation and no additional adjustment factor was considered. Further investigations regarding the correlation between the parameters a and p and spatial factors like the distance from radar device or the height of the radar beam above the ground did not lead to conclusive results. This seems to have been caused by the heterogeneity of the convective events. Therefore, a spatial mean but event-dependent Zradar/RZ/R relation was used. The use of a standard Zradar/RZ/R relation (e.g. Marshall etal. 1955; Dyck and Peschke 1995) without calibrating a and p would not have led to an acceptable data adjustment and consequently to large runoff modelling errors. This was also concluded by Quirmbach et al. (1999), who showed a significant underestimation of flood events using a standard Zradar/RZ/R relation. This is caused by the event-dependent raindrop size distribution as well as by precipitation characteristics (e.g. Smith and Krajewski 1993; Pessoa etal. 1993; Haase and Crewell 2000). In general, the used method for radar data adjustment was found to be practical and efficient as it considered the total amount and the distribution of rainfall intensity during the event. It led to a-values below 100, which were lower that the a-values ofthe standard Zradar/RZ/R relation of the German weather survey, but were found in the earlier literature (e.g. Hirayama et al. 1997).

A good overview about limitations and shortcomings connected with observation and transformation of radar data and ongoing research to improve weather radar measurements is given, for example, by Terblanche et al. (2001). In recent years, numerous uncertainties during data observation and transformation were investigated (e.g. Grecu and Krajewski 2000). In addition, methods for improvement of data quality were developed (e.g. Geor-gakakos 2000), and the precipitation rate dependence on the raindrop size distribution was examined (Uijlenhoet and Strieker 1999). Lange et al. (2003) discuss the shortcomings of using C-band radar products. In general, the use of operational available radar data for hydrologieal applications is still controversial, also because of the labour-intensive radar data management and adjustment. But the results of this study highlight also the potential of radar data, in particular, for convective storm events with large spatial and temporal heterogeneities. However, the prerequisites to utilise these data are a reliable ground station network and a distributed, process-oriented and well-validated rainfall-runoff model.

16.6 CONCLUSIONS AND OUTLOOK

The following three questions were addressed in this study:

1. What might be the contribution of highly resoluted radar data to capture spatial and temporal variability of convective precipitation events in mountainous areas?

Detecting rainfall patterns in mountainous catchments is complicated because of the difficulties both in maintaining a sufficient network of ground stations and the variability of precipitation caused by luv and lee effects, and hence, difficulties in detecting highly localised intense rain cells. Highly resoluted rainfall radar data can help significantly in capturing the spatial and temporal variability of precipitation in mountainous areas, in particular, for very heterogeneous events, or if a sufficient number of good ground stations is not available, for example, in smaller sub-catchments. Additionally, higher short-term intensities are measured using radar data, which can be important for triggering certain runoff generation processes. However, the location of the radar station, the topography of the basin, the characteristics of the respective event, the specific problems of radar measurements in mountainous catchments and availability of representative ground stations at least near the catchment for radar data adjustment need to be considered regarding the potential of highly resoluted rainfall radar data. Although the quality of operational radar data can be - event dependent - low, they offer useful information about rainfall patterns and maximum intensities.

2. What is the significance of the distribution of basin precipitation during convective cells for flood modelling in mountainous catchments?

The importance of considering adequately the spatial and temporal variability of basin precipitation during convective cells for flood modelling has been demonstrated clearly in this study. Basin precipitation is a major order control on flood modelling in mountainous, meso-scale basins, and the errors in flood predictions can be large if incorrect basin precipitation values are used. Lumped precipitation values can lead to wrong runoff generation predictions locally and, consequently, to uncertain discharge predictions at ungauged sites or sub-basins. However, there are many site- and event-specific circumstances that make general statements regarding the impact of rainfall distributions on runoff simulations difficult.

3. How can operational available radar data be used within a detailed hydrological model in an appropriate way?

The results showed that an improvement of the runoff simulation by incorporating radar data is only possible if an extensive data disaggregation, correction, and adjustment is performed. Therefore, suitable ground station data are indispensable. Considering the experiences with the radar product used in this study let us conclude that a nowcasting of basin precipitation using standard Zradar/Rz/R relations without an event-dependent adjustment of the data seems to be hazardous and will result in uncertain discharge predictions.

The results of this study suggest a number of new avenues for research. Firstly, so far the precipitation input was used in a very detailed spatial resolution (50 x 50 m2). In a next step, different aggregations and their influence on the runoff modelling need to be examined. Therefore, the used process-based catchment model and the nested basin structure of the test site are suitable. Secondly, the parameters a and p were adjusted for each event separately but averaged spatially. It should be investigated if even better simulation results can be obtained by a ground station or better elevation-specific, and thus spatially variable, determination of these two fitting parameters. These results should be compared with calibrating the Zradar/RZ/R relation using spatially variable adjustment factors and standard Zradar/RZ/R relations. Thirdly, the significance of the spatial and temporal variability of precipitation for flood modelling should be compared for different magnitudes of events, different scaled basins, and different landscapes. The scale at which the basin precipitation dominates the extent of flood generation in comparison with other influences, such as the physiographic basin characteristics, should be examined.

16.7 ACKNOWLEDGEMENTS

The detailed radar data were been provided from the German Weather Service (DWD). The State Institute for Environmental Protection Baden-Wurttemberg

(Landesanstalt fur Umweltschutz (LfU) BadenWürttemberg) made the precipitation ground station data available. In addition, the federal environmental survey (Umweltbündesamt, UBA) provided the rainfall data from the station Schauinsland. The Gewaesserdirektion Waldshut, Germany, measured the runoff data. The input of Günter Gassler during analysis of the radar data and during extensive discussions is gratefully acknowledged. Jens Lange (University of Freiburg, Germany) has provided a code for converting the radar data. Also his helpful comments are very much appreciated. Parts from the converting program from Jens Lange were combined with a reading program from Daniel Sacher, J. Lang Datenservice. We would also like to thank Daniel Sacher. The study was supported by the German Research Foundation (Deutsche Forschungsgemeinschaft, Bonn) grant no. Le 698/12-1.

REFERENCES

Adamowski K, MuirJ (1989) A Kalman-filter modelling of space-time rainfall using radar and rain gauge observations. Can. J. Civil Eng. 16(5): 767-773. Anderson SP, Dietrich WE, Montgomery DR, Torres R, Conrad ME, Loague K (1997) Subsurface flow paths in a steep, unchanneled catchment. Water Resour. Res., 33(12): 2637-2654.

Andrieu H, Creutin JD, Delrieu G, Faure D (1997) Use of weather radar for the hydrology of mountainous area. Part I: radar measurement interpretation. J. Hydrol. 193(1-4): 1 -25.

Arnaud P, Bouvier C, Cisneros L, Dominquez R (2002) Influence of spatial variability on flood prediction. J. Hydrol. 260:216-230.

Bergstrom S (1992) The HBV Model-Its Structure and

Applications. SMHI, RH, 4., Norrkoping, Schweden. Beven KJ (2002) The future of distributed hydrological modelling, Special Issue of the Journal Hydrological Processes, 16. John Wiley & Sons, Chichester, UK. Beven KJ, Lamb R, Quinn P, Romanowicz R, Freer J (1995) TOPMODEL. In: VP Singh (Ed.), Computer Models of Watershed Hydrology. Water Resources Publications, Colorado, CO.

Boyle DP, Gupta HV, Sorooshian S, Koren V, Zhang Z, Smith M (2001) Toward improved streamflow forecasts: Value of semidistributed modeling. Water Resour. Res. 37(11): 2749-2759. Carpenter TM, Georgakakos KP, Sperfslagea JA (2001) On the parametric and NEXRAD-radar sensitivities of a distributed hydrologic model suitable for operational use. J. Hydrol. 253: 169-193.

Ciach GJ, Krajewski W (1999) On the estimation of radar rainfall error variance. Adv. Water Resour. 22(6): 585-595. Creutin JD, Andrieu H, Faure D (1997) Use of weather radar for the hydrology of mountainous area. Part II: radar measurement validation. J. Hydrol. 193(1-4): 26-44.

Delrieu G, Serrar S, Guardo E, Creutin JD (1999) Rain measurement in hilly terrain with X-band weather radar systems: Accuracy of path-integrated attenuation estimates derived from mountain returns. J.Atmos. Ocean. Tech. 16(4): 405-416.

DWD Deutscher Wetterdienst (German Weather Service) (1997) AKORD -Anwenderkoordinierte Organisation von Radar-Daten. Produktkatalog. Deutscher Wetterdienst Geschaftsfeld Hydrometeorologie, Offenbach, Germany.

Dyck S, Peschke G (1995) Grundlagen der Hydrologie. Verlag für Bauwesen, Berlin, Germany, p. 408.

Faeh A, Scherrer S, NaefF (1997) A combined field and numerical approach to investigate flow processes in natural macroporous soils under extreme precipitation. Hydrol. Earth Syst. Sci. 4: 787-800.

Faures JM, Goodrich DC, Woolhiser DA, Sorooshian S (1995) Impact of small-scale spatial variability on runoff modelling. J. Hydrol. 173: 309-326.

Georgakakos KP (2000) Covariance propagation and updating in the context of real-time radar data assimilation by quantitative precipitation forecast models. J. Hydrol. 239(1-4): 115-129.

Goodrich DC, Faures J-M, Woolhiser DA, Lane LJ, Sorooshian S (1995) Measurement and analysis of small-scale convective storm rainfall variability. J. Hydrol. 173(1-4): 283-308.

Grayson RB, Western AW, Chiew FHS, Bloschl G (1997) Preferred states in spatial soil moisture patterns: Local and non-local controls. Water Rescour. Res. 33(12): 2897-2908.

Grecu M, Krajewski W (2000) Simulation study of the effects of model uncertainty in variational assimilation of radar data on rainfall forecasting. J. Hydrol. 239(1-4): 85-96.

GuntnerA, Uhlenbrook S, Seibert J, Leibundgut C (1999) Multi-criterial validation of TOPMODEL in a mountainous catchment. Hydrol. Process. 13: 1603-1620.

Gurtz J, Baltensweiler A, Lang H, Menzel L, Schulla J (1997) Auswirkungen von Klimatischen Variationen auf Wasserhaushalt und Abfluß im Flußgebiet des Rheins. Vdf Hochschulverlag AG an der ETH Zürich, Zurich, Schweiz.

Haase G, Crewell S (2000) Simulation of radar reflectivities using a mesoscale weather forecast model. Water Resour. Res. 36(8): 2221-2231.

Hinton MJ, Schiff SL, English MC (1994) Examining the contributions of glacial till water to storm runoff using two-and three-component hydrograph separations. Water Resour. Res. 30: 983-993.

HirayamaD, FujitaM, NakatsugawaM (1997) The identification of optimum Z-R relation based on runoff analysis. In: BJ Braga, O Massambani (Eds.), Weather Radar Technologies for Water Resources Management IRTCUD/University of Sao Paulo, Brazil, South America and IHP-UNESCO.

Hoeg S, Uhlenbrook S, Leibundgut C (2000) Hydrograph separation in a mountainous catchment - combining hydro-chemical and isotopic tracers. Hydrol. Process. 14: 1199-1216.

Hornberger GM, Germann PF, Beven KJ (1991) Throughflow and solute transport in an isolated sloping block in a forested catchment. J. Hydrol. 124: 81-99.

Jasper K, Gurtz J, Lang H (2002) Advanced flood forecasting in Alpine watersheds by coupling meteorological observations and forecasts with a distributed hydrological model. J.Hydrol. 267(1-2): 40-52.

Lange J (1999) A non-calibrated rainfall-runoff model for large arid catchments, Nahal Zin, Israel. PhD Thesis, Freiburger Schriften zur Hydrologie, Band 9, University of Freiburg, Institute of Hydrology, Freiburg.

Lange J, Leibundgut C, Greenbaum N, Schick AP (1999) A non-calibrated rainfall-runoff model for large, arid catchments. Water Resour. Res. 35(7): 2161-2172.

Lange J, Wagner A, TetzlaffD (2003) Hochwassersimulation in kleinen Einzugsgebieten: Eignung von Niederschlagsradar und Auswirkung versiegelter Flachen (Flood simulation in small catchments: Applicability of rainfall radar data and effects of impervious areas). Tag der Hydrologie 2003, ATV-DVWK, HB Kleeberg (Ed.), Vol. 1, pp. 123-131.

Karssenberg D (2002) The value of environmental modelling languages for building distributed hydrological models. Hydrol. Process. 16: 2751-2766.

KorenVI, Finnerty BD, Schaake JC, Smith MB, Seo DJ, Duan QY (1999) Scale dependencies of hydrological models to spatial variability of precipitation. J.Hydrol. 217:285 -302.

Krajewski WF, Ventakataramann L, Georgakakos KP, Jain SC (1991) A Monte Carlo study of rainfall sampling effect on a distributed catchment model. Water Resour. Res. 27(1): 119-128.

Marshall JS, Hitschfeld W, Gunn KLS (1955) Advances in radar weather. Adv. Geophys. 2: 1-56.

Maul-Kotter B, Spies S, Einfalt T (2001) Qualitätskriterien für Radardaten in der hydrologischen Simulation. Hydrol. Wasserbewirt. 45(6): 236-243.

McDonnell JJ (1990) A rationale for old water discharge through macropores in a steep, humid catchment. Water Resour. Res. 26: 2821 -2832.

McDonnell JJ, TanakaT. (2001) Hydrology and biogeochem-istry of forested catchments. Special Issue of the Journal Hydrological Processes, 15, John Wiley & Sons, Chichester, UK, pp. 1673-2055.

Mecklenburg S, Joss J, Schmid W (2000) Improving the nowcasting of precipitation in an Alpine region with enhanced radar echo tracking algorithm. J. Hydrol. 239: 46-68.

Mehlhorn J, Armbruster F, Uhlenbrook S, Leibundgut C (1998) Determination of the Geomorphological Instantaneous unit Hydrograph Using Tracer Experiments in a Headwater Basin. IAHS Publication No. 248, pp. 327-336.

Merot Ph, Ezzehar B, Walter C, Aurousseau P (1995) Mapping waterlogging of soils using digital terrain models. Hydrol. Process. 9: 27-34.

Michaud JD, Sorooshian S (1994) Effects of rainfall-sampling errors on simulations of desert flash floods. Water Resour. Res. 30(10): 2765-2775.

Milly PCD, Eagleson PS (1988) Effects of storm scale on surface runoff volume. Water Resour. Res. 24(4): 620-624.

Mosley MP (1982) Subsurface flow velocities through selected forest soils, south island, New Zealand. J. Hydrol. 55: 65 -92.

Nash JE, Sutcliffe JV (1970) River flow forecasting through conceptual models, 1. A discussion of principles. J. Hydrol. 10: 282-290.

Obled C, Cluckie ID, Collier CG (1991) Reflections on Rainfall Information Requirements for Operational Rainfall-Runoff Modelling, Hydrological Applications of Weather Radar. Ellis Horwood, pp. 469-482.

Obled C, Wendling J, Beven K (1994) The sensitivity of hydrological models to spatial rainfall patterns: an evaluation using observed data. J. Hydrol. 159: 305-308.

OgdenFL, Sharif HO, Senarath SUS, Smith JA, BaeckML, Richardson JR (2000) Hydrologic analysis of the Fort Collins, Colorado, flash flood of 1997. J.Hydrol. 228: 82-100.

O'Loughlin G, HuberW, ChocatB (1996) Rainfall-runoff processes and modelling. J. Hydraulic Res. 34(6): 733-751.

Pessoa ML, Bras RL, Williams ER (1993) Use of weather radar for flood forecasting in the Sieve River basin: A sensitivity analysis. J.Appl. Meteorol. 32(3): 462-475.

Quirmbach M (2003) Nutzung von Wetterradardaten fur Niederschlags- und Abflussvorhersagen in urbanen Einzugsgebieten. PhD Thesis, Ruhr-Universitat Bochum, Bochum, Germany, p. 178.

Quirmbach M, Schultz G, Frehmann T (1999) Use of weather radar for combined control of an urban drainage system and a sewage treatment plant. Impacts of Urban Growth on Surface Water and Groundwater Quality, Proceedings of IUGG 99 Symposium HS5, Birmingham, July 1999, IAHS Publication No. 259, pp. 245-250.

Refsgaard JC, Storm B (1996) Construction, calibration and validation of hydrological models. In: MB Abbot, JC Refsgaard (Eds.), Distributed Hydrological Modelling. Kluwer Academic, Dordrecht, Niederlande, pp. 41 -54.

Schilling W, Fuchs L (1984) A Quantitative Assessment of Uncertainties in StormwaterModeling. Urban storm drainage. Göteborg, Schweden, pp. 625-633.

Schulla J (1997) Hydrologische Modellierung von Flußeinzugsgebieten zur Abschätzung der Folgen von Klimaanderungen, Züricher Geographische Hefte, 65. ETH Zurich, Zürich, Schweiz.

Sempere Torres D, PorraJM, Creutin JD (1994) A general formulation for raindrop size distribution. J. Appl. Meteorol. 33: 1494-1502.

Seo DJ, Breidenbach JP, Johnson ER (1999) Real-time estimation of mean field bias in radar rainfall data. J. Hydrol. 223: 131-147.

Smith JA, Krajewski WF (1993) A modelling study of rainfall-rate-reflectivity relationships. Water Resour. Res. 29(8): 2505-2514.

Sun X, Mein RG, Keenan TD, Elliott JF (2000) Flood estimation using radar and raingauge data. J. Hydrol. 239: 4-18.

SyedKH, Goodrich DC, Myers DE, Sorooshian S (2003) Spatial characteristics of thunderstorm rainfall fields and their relation to runoff. J. Hydrol. 271: 1-21.

Terblanche DE, Pegram GGS, Mittermaier MP (2001) The development of weather radar as a research and operational tool for hydrology in South Africa. J. Hydrol. 241: 3-25.

Uchida T, Kosugi K, Mizuyama T (2002) Effects of pipe flow and bedrock groundwater on runoff generation in a steep headwater catchment in Ashiu, central Japan. Water Resour. Res., 38(7): 24-38.

Uhlenbrook S (1999) Untersuchung und Modellierung der Abflussbildung in einem mesoskaligen Einzugsgebiet, Freiburger Schriften zur Hydrologie, 10. Universität Freiburg, Freiburg, Germany.

Uhlenbrook S, Didszun J, Leibundgut Ch (2003c) Runoff generation processes on hillslopes and their susceptibility to global change. In: UM Huber, MA Reasoner, B Bugmann (Eds.), 2003: Global Change and Mountain Regions: A State of Knowledge Overview. Advances in Global Change Research. Kluwer Academic Publishers, Dordrecht, Netherlands.

Uhlenbrook S, Frey M, Leibundgut Ch, Maloszewski P (2002) Residence time based hydrograph separations in a meso-scale mountainous basin at event and seasonal time scales. Water Resour. Res. 38(6): 1 -14.

Uhlenbrook S, Leibundgut Ch (2002) Process-oriented catchment modelling and multiple-response validation. Hydrol. Process. 16: 423-440.

Uhlenbrook S, McDonnell JJ, Leibundgut Ch (2003a) Runoff Generation and Implications for River Basin Modelling, Special Issue of the Journal Hydrological Processes, 17, John Wiley & Sons, Chichester, UK, p. 2.

Uhlenbrook S, Roser S, Tilch N (2003b) Hydrological process representation at the meso-scale: The potential of a distributed, conceptual catchment model. Accepted by J. Hydrol. 291: 278-296.

Uhlenbrook S, SeibertJ, Leibundgut C, Rodhe A (1999) Prediction uncertainty of conceptual rainfall-runoff models caused by problems to identify model parameters and structure. Hydrol. Sci. J. 44(5): 279-299.

UijlenhoetR, StrickerJNM (1999) A consistent rainfall parameterization based on the exponential raindrop size distribution. J. Hydrol. 218(3-4): 101-127.

Winchell M, Gupta HV, Sorooshian S (1998) On the simulation of infiltration and saturation-excess runoff using radar-based rainfall estimates: Effects of algorithm uncertainty and pixel aggregation. Water Resour. Res. 34(10): 2655-2670.

Woods R, Grayson R, Western A, Duncan M, Wilson D, Young R, Ibbitt R, Henderson R, McMahon T (2000) Experimental design and initial results from the Mahurangi River Variability Experiment: MARVEX. In: V Lakshmi, JD Albertson, J Schaake (Eds.), Land Surface Hydrology, Meteorology and Climate: Observations and Modeling. Water Science and Application, Vol. 3, American Geophysical Union, Washington, DC, pp. 201 -213.

Woods R, Sivapalan M (1999) A synthesis of space-time variability in storm response: rainfall, runoff generation and routing. Water Resour. Res. 35(8): 2469-2485.

0 0

Post a comment