An Extended Model Evaluation Method using Multiple Assessment Indices (MAIs) under Uncertainty in Rainfall-Runoff Modeling 강우-유출 모델링의 불확실성 고려한 다중 평가지수에 의한 확장형 모형평가 방법원문보기
Conventional methods of model evaluation usually rely only on model performance based on a comparison of simulated variables to corresponding observations. However, this type of model evaluation has been criticized because of its insufficient consideration of the various uncertainty sources involved...
Conventional methods of model evaluation usually rely only on model performance based on a comparison of simulated variables to corresponding observations. However, this type of model evaluation has been criticized because of its insufficient consideration of the various uncertainty sources involved in modeling processes. This study aims to propose an extended model evaluation method using multiple assesment indices (MAIs) that consider not only the model performance but also the model structure and parameter uncertainties in rainfall-runoff modeling. A simple reservoir model (SFM) and distributed kinematic wave models (KWMSS1 and KWMSS2 using topography from 250m, 500m, and 1km digital elevation models) were developed and assessed by three MAIs for model performance, model structural stability, and parameter identifiability. All the models provided acceptable performance in terms of a global response, but the simpler SFM and KWMSS1 could not accurately represent the local behaviors of hydrographs. In addition, SFM and KWMSS1 were structurally unstable; their performance was sensitive to the applied objective functions. On the other hand, the most sophisticated model, KWMSS2, performed well, satisfying both global and local behaviors. KMSS2 also showed good structural stability, reproducing hydrographs regardless of the applied objective functions; however, superior parameter identifiability was not guaranteed. Numerous parameter sets could lead to indistinguishable hydrographs. This result supports that while making a model complex increases its performance accuracy and reduces its structural uncertainty, the model is likely to suffer from parameter uncertainty. The proposed model evaluation process can provide an effective guideline for identifying a reliable hydrologic model.
Conventional methods of model evaluation usually rely only on model performance based on a comparison of simulated variables to corresponding observations. However, this type of model evaluation has been criticized because of its insufficient consideration of the various uncertainty sources involved in modeling processes. This study aims to propose an extended model evaluation method using multiple assesment indices (MAIs) that consider not only the model performance but also the model structure and parameter uncertainties in rainfall-runoff modeling. A simple reservoir model (SFM) and distributed kinematic wave models (KWMSS1 and KWMSS2 using topography from 250m, 500m, and 1km digital elevation models) were developed and assessed by three MAIs for model performance, model structural stability, and parameter identifiability. All the models provided acceptable performance in terms of a global response, but the simpler SFM and KWMSS1 could not accurately represent the local behaviors of hydrographs. In addition, SFM and KWMSS1 were structurally unstable; their performance was sensitive to the applied objective functions. On the other hand, the most sophisticated model, KWMSS2, performed well, satisfying both global and local behaviors. KMSS2 also showed good structural stability, reproducing hydrographs regardless of the applied objective functions; however, superior parameter identifiability was not guaranteed. Numerous parameter sets could lead to indistinguishable hydrographs. This result supports that while making a model complex increases its performance accuracy and reduces its structural uncertainty, the model is likely to suffer from parameter uncertainty. The proposed model evaluation process can provide an effective guideline for identifying a reliable hydrologic model.
* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.
제안 방법
For parameter identifiability assessment, we applied the SCEM-UA to estimate individual posterior parameter distributions, and then investigated the uniqueness of the calibrated parameters from the probability density functions of each model. Here, the highest density values of each distribution were used as the individual indicators of parameter identifiability, and the mean value of each maximum identifiability indicator was used for the MPII.
In this study, three different types of rainfall-runoff models, from a simple lumped model to distributed kinematic wave models, were developed under an object-oriented hydrological modeling system. Moreover, three different spatial resolutions of a digital elevation model (DEM) were used to investigate the scale effect on both model performance and uncertainty assessment in distributed rainfall-runoff modeling.
In this study, three different types of rainfall-runoff models, from a simple lumped model to distributed kinematic wave models, were developed under an object-oriented hydrological modeling system. Moreover, three different spatial resolutions of a digital elevation model (DEM) were used to investigate the scale effect on both model performance and uncertainty assessment in distributed rainfall-runoff modeling.
This paper proposes an extended model evaluation framework under uncertainty in rainfall-runoff modeling for identifying a more reliable model. The new framework follows the basic concepts of uncertainty proposed by Beven (2002) and Wagener and Gupta (2005).
It admits numerous plausible representations providing identically good model performance measures, while newly developed criteria are used to assess other inherent model characteristics related to structural and parameter uncertainties. We prepared seven different rainfall-runoff models ranging from a simple lumped model to sophisticated distributed models and then evaluated the models with respect to model performance, model structural stability, and parameter identifiability. A highly ranked model by these criteria is structurally stable, shows less parameter uncertainty, and ensures accurate prediction results.
데이터처리
The performances of each model structure were assessed using the Nash-Sutcliffe coefficient (NSC) for the two periods; the two measures were then averaged to obtain the MPI.
이론/모형
This paper proposes an extended model evaluation framework under uncertainty in rainfall-runoff modeling for identifying a more reliable model. The new framework follows the basic concepts of uncertainty proposed by Beven (2002) and Wagener and Gupta (2005). It admits numerous plausible representations providing identically good model performance measures, while newly developed criteria are used to assess other inherent model characteristics related to structural and parameter uncertainties.
성능/효과
The overall results of the model evaluation demonstrate that the ideal model structure, which guarantees the best values in terms of the three criteria, was not found in this study. The distributed model, KWMSS2, was much better than the simple models, SFM and KWMSS1, in terms of two evaluative criteria, MPI and MSSI, but KWMSS2 did not ensure the best parameter identifiability. Therefore, additional constraints that are able to reject unreliable parameter set(s) and provide reliable prediction results need to be combined in the proposed modeling framework for further model identification.
후속연구
The distributed model, KWMSS2, was much better than the simple models, SFM and KWMSS1, in terms of two evaluative criteria, MPI and MSSI, but KWMSS2 did not ensure the best parameter identifiability. Therefore, additional constraints that are able to reject unreliable parameter set(s) and provide reliable prediction results need to be combined in the proposed modeling framework for further model identification.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.