Gauged and ungauged catchments: strategies for producing the relevant hydrographs with sufficient accuracy WaterEurope - Workgroup 05 Yesenia Pereyra Mario López Muñoz Anis Boukra Leqi Tian Arthur Henriet Gyembo Tenzin Edwar Forero Date: 15/02/2019 Supervisor: Mr.Gourbesville
Outline 1. Introduction 2. What is a model? 3. How can we select the proper model for this case? 4. Calibration of Models applied on Gauged Catchment 5. Calibration of Models applied on Ungauged Catchment 6. Comparison of the results 7. Recommendations 8. References Outline - Introduction - Model - Classification - Selection - Results - Comparison - Recommendations - References 2
1. Introduction Main objective We must design two new bridges, which will We are required to create a hydrograph to be affected by rivers within a gauged and an know the catchments behaviour ungauged catchment. Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 3
1. Introduction Main objective Var Catchment: Location of the new bridges Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 4
1. Introduction Key concepts: Gauged? Ungauged? Gauged Discharge measurements data about the catchment catchment are available Ungauged No data of discharge on the catchment is available catchment Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 5
1. Introduction Why do we need discharge measurements data? To calibrate ➔ the model accurately to adjust it as possible to the real conditions. How can we obtain accurate hydrographs for ungauged catchments? Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 6
2. Through a Model…. What is a model? A mathematical model is ➔ needed to simulate the catchment behaviour using mathematical equations, logical statements, initial and boundary conditions and expressing relationships between input and output. Representation of input, system, and output for a mathematical model Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 7
3. How can we select the proper model for this case? Data available There is no model defined ➔ for each case, it depends of a wide numbers of variables Computing Hydrological Model power problem We will analyse what kind of ➔ model is recommended to create a suitable hydrograph Level in ungauged catchments accuracy Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 8
3. How can we select the proper model for this case? Models selected to analysis Vésubie Var catchment Ungauged Gauged ● MIKE SHE model ● MIKE SHE model Hydrological Event studied ● HEC-HMS ● HEC-HMS ANN (Artificial Neural Network) ● ● ANN (Artificial Neural Network) 4th 17:00 to 6th 00:00 NAM ● November 94 ● NAM Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 9
3. How can we select the proper model for this case? Hydrological Models Deterministic Stochastic ANN Lumped Distributed Semi-distributed SHE HEC-HMS NAM Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 10
3. How can we select the proper model for this case? Main characteristics of models selected: MIKE SHE ● Hydrological model : Spatially distributed catchment parameters and rainfall to calculate the discharge ● Rectangular mesh and finite difference method for the calculation ● Overland flow thanks to diffusive wave approximation of Saint Venant equations ● 2D model with a 75 m resolution DEM Parameters used : ● Rainfall (Thiessen) ● DEM 75m ● Strickler number ● Net Rainfall Fraction SHE Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 11
3. How can we select the proper model for this case? Main characteristics of models selected: HEC-HMS ● Lumped Model - spatial variations and characteristics & processes are averaged Fast simulation & easy to understand ● ● Requires less parameters Warning: simplified assumptions about some parameters ● Parameters used : ● Gage Weight - Thiessen ● SCS Curve Number with poor soil conditions ● SCS UH - Transformed Method ● Lag Time HEC-HMS ● Constant monthlyflow at Lower Var Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 12
3. How can we select the proper model for this case? Artificial neural networks as approximators of stochastic processes. Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 13
3. How can we select the proper model for this case? Main characteristics of models selected: NAM NAM (Nedbør-Afstrømnings-Model) ● precipitation-runoff-model ● continuously accounting for the water content in four different and mutually interrelated storages. ● Each storage represents different physical elements of the catchment. Basic inputs ● Catchment area （ DEM 75m ） Precipitation Evapotranspiration Extend components ● Temperature NAM is a set of linked mathematical statements describing, in a simplified quantitative form, the behavior of the land phase of the hydrological cycle. Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 14
4. Calibration of Models applied on Gauged Catchment Results obtained with MIKE SHE on the Var Catchment For the studied event (04/11/1994 18:00 - 06/11/1994 00:00), we calibrated the Var catchment changing some parameters (Strickler, Net Rainfall Fraction). Lack of accuracy from cross-section or rainfall data Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 15
5. Calibration of Models applied on Ungauged Catchment Results obtained with MIKE SHE on the Vésubie We assume that Vesubie and Lower Var subcatchments presents the same physical parameters. For the studied event (04/11/1994 18:00 - 06/11/1994 00:00), and in order to have more accurate results, we applied a scale effect (0.14) on the Vésubie, using the Var values to obtain the following Hydrograph. Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 16
4. Calibration of Models applied on Gauged Catchment Results obtained with ANN on the Var Catchment Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 17
4. Calibration of Models applied on Gauged Catchment Results obtained with ANN on the Vésubie Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 18
6. Comparison of the results Results obtained on the Vésubie (ungauged) Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 19
6. Comparison of the results Results obtained on the Lower Var (gauged) Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 20
7. Recommendations MIKE SHE HEC-HMS ANN NAM Flexibility on the watershed Easy Advantages High detail configuration implementation Quick when is adopted Quick Long simulation Non hydrological Low detail Disadvantages Accuracy Simplified model parameters - Less Applying it is not depending on the accuracy feasible data quality For this project, to determine the behaviour of the rivers discharges, we will work with MIKE SHE, a deterministic model. Notwithstanding, it could demand more computational power, but we will produce more accurate results. It’s the more solid model, which take into account several hydrological and physical parameters. Outline - Introduction - Model - Selection - Results - Comparison - Recommendations - References 21
Recommend
More recommend
