scholarly journals Monthly Rainfall-Runoff Modeling at Watershed Scale: A Comparative Study of Data-Driven and Theory-Driven Approaches

Water ◽  
2018 ◽  
Vol 10 (9) ◽  
pp. 1116 ◽  
Author(s):  
Won Chang ◽  
Xi Chen
Keyword(s):  
Comparative Study ◽  
Surface Runoff ◽  
Aridity Index ◽  
Hydrologic Model ◽  
Data Driven ◽  
Vegetation Coverage ◽  
Potential Evaporation ◽  
Rainfall Runoff ◽  
Runoff Modeling ◽  
Input Variables

Data-driven machine learning approaches have been rapidly developed in the past 10 to 20 years and applied to various problems in the field of hydrology. To investigate the capability of data-driven approaches in rainfall-runoff modeling in comparison to theory-driven models, we conducted a comparative study of simulated monthly surface runoff at 203 watersheds across the contiguous USA using a conceptual model, the proportionality hydrologic model, and a data-driven Gaussian process regression model. With the same input variables of precipitation and mean monthly aridity index, the two models showed similar performance. We then introduced two more input variables in the data-driven model: potential evaporation and the normalized difference vegetation index (NDVI), which were selected based on hydrologic knowledge. The modified data-driven model performed much better than either the conceptual or original data-driven model. A sensitivity analysis was conducted on all three models tested in this study, which showed that surface runoff responded positively to increased precipitation. However, a confounding effect on surface runoff sensitivity was found among mean monthly aridity index, potential evaporation, and NDVI. This confounding was caused by complex interconnections among energy supply, vegetation coverage, and climate seasonality of the watershed system. We also conducted an uncertainty analysis on the two data-driven models, which showed that both models had reasonable predictability within the 95% confidence interval. With the additional two input variables, the modified data-driven model had lower prediction uncertainty and higher prediction accuracy.

Effect of input variables on rainfall-runoff modeling using a deep learning method

2021 ◽  
Author(s):  
Kazuki yokoo ◽  
Kei ishida ◽  
Takeyoshi nagasato ◽  
Ali Ercan
Keyword(s):  
Deep Learning ◽  
Air Temperature ◽  
Meteorological Variables ◽  
Learning Method ◽  
Rainfall Runoff ◽  
Model Accuracy ◽  
Runoff Modeling ◽  
Input Variables ◽  
Rainfall Runoff Model ◽  
Runoff Model

<p>In recent years, deep learning has been applied to various issues in natural science, including hydrology. These application results show its high applicability. There are some studies that performed rainfall-runoff modeling by means of a deep learning method, LSTM (Long Short-Term Memory). LSTM is a kind of RNN (Recurrent Neural Networks) that is suitable for modeling time series data with long-term dependence. These studies showed the capability of LSTM for rainfall-runoff modeling. However, there are few studies that investigate the effects of input variables on the estimation accuracy. Therefore, this study, investigated the effects of the selection of input variables on the accuracy of a rainfall-runoff model by means of LSTM. As the study watershed, this study selected a snow-dominated watershed, the Ishikari River basin, which is in the Hokkaido region of Japan. The flow discharge was obtained at a gauging station near the outlet of the river as the target data. For the input data to the model, Meteorological variables were obtained from an atmospheric reanalysis dataset, ERA5, in addition to the gridded precipitation dataset. The selected meteorological variables were air temperature, evaporation, longwave radiation, shortwave radiation, and mean sea level pressure. Then, the rainfall-runoff model was trained with several combinations of the input variables. After the training, the model accuracy was compared among the combinations. The use of meteorological variables in addition to precipitation and air temperature as input improved the model accuracy. In some cases, however, the model accuracy was worsened by using more variables as input. The results indicate the importance to select adequate variables as input for rainfall-runoff modeling by LSTM.</p>

scholarly journals Rainfall-runoff modeling using Doppler weather radar data for Adyar watershed, India

MAUSAM ◽  
2021 ◽  
Vol 65 (1) ◽  
pp. 49-56
Author(s):  
S.JOSEPHINE VANAJA ◽  
B.V. MUDGAL ◽  
S.B. THAMPI
Keyword(s):  
Weather Radar ◽  
Radar Data ◽  
Rain Gauge ◽  
Hydrologic Model ◽  
Rainfall Data ◽  
Doppler Weather Radar ◽  
Rainfall Runoff ◽  
Cyclonic Storm ◽  
Rain Gauges ◽  
Runoff Modeling

Precipitation is a significant input for hydrologic models; so, it needs to be quantified precisely. The measurement with rain gauges gives the rainfall at a particular location, whereas the radar obtains instantaneous snapshots of electromagnetic backscatter from rain volumes that are then converted into rainfall via algorithms. It has been proved that the radar measurement of areal rainfall can outperform rain gauge network measurements, especially in remote areas where rain gauges are sparse, and remotely sensed satellite rainfall data are too inaccurate. The research focuses on a technique to improve rainfall-runoff modeling based on radar derived rainfall data for Adyar watershed, Chennai, India. A hydrologic model called ‘Hydrologic Engineering Center-Hydrologic Modeling System (HEC-HMS)’ is used for simulating rainfall-runoff processes. CARTOSAT 30 m DEM is used for watershed delineation using HEC-GeoHMS. The Adyar watershed is within 100 km radius circle from the Doppler Weather Radar station, hence it has been chosen as the study area. The cyclonic storm Jal event from 4-8 November, 2010 period is selected for the study. The data for this period are collected from the Statistical Department, and the Cyclone Detection Radar Centre, Chennai, India. The results show that the runoff is over predicted using calibrated Doppler radar data in comparison with the point rainfall from rain gauge stations.

scholarly journals A comparative study of artificial intelligence and conventional techniques for rainfall-runoff modeling

2017 ◽  
Vol 10 (2) ◽  
pp. 441-449 ◽  
Author(s):  
VIJAY KUMAR SINGH ◽  
BHASKAR PRATAP SINGH ◽  
ASHISH KUMAR ◽  
VIVEKANAND VIVEKANAND
Keyword(s):  
Artificial Intelligence ◽  
Comparative Study ◽  
Rainfall Runoff ◽  
Runoff Modeling

Rainfall-runoff modeling using conceptual, data driven, and wavelet based computing approach

2013 ◽  
Vol 493 ◽  
pp. 57-67 ◽  
Author(s):  
P.C. Nayak ◽  
B. Venkatesh ◽  
B. Krishna ◽  
Sharad K. Jain
Keyword(s):  
Data Driven ◽  
Rainfall Runoff ◽  
Runoff Modeling ◽  
Conceptual Data ◽  
Computing Approach

Hydrodynamic rainfall-runoff modeling using different approaches for runoff creation

2020 ◽  
Author(s):  
Karl Broich ◽  
Thomas Pflugbeil ◽  
Johannes Mitterer ◽  
Markus Disse
Keyword(s):  
Water Depth ◽  
Surface Runoff ◽  
Flash Flood ◽  
Flash Floods ◽  
Rainfall Runoff ◽  
Runoff Generation ◽  
Hydrodynamic Models ◽  
Additional Advantage ◽  
Runoff Modeling ◽  
Scs Cn

<p>After extreme flash floods events 2016 in Bavaria, the cooperation project HiOS (reference map for surface runoff and flash floods) was started aiming at the detailed analysis of risk generated by flash floods using GIS methods as well as hydrological and hydrodynamic models. Part of the risk analysis is done using hydrodynamic rainfall-runoff modeling (HDRRM). HDRRM gets more and more popular since hydrodynamic models are able to accept rainfall as input. But most of the known hydrodynamic models have no integrated precipitation modules and therefore cannot be used uniquely for rainfall-runoff modeling. In this study, TELEMAC-2D is used for HDRRM because it already contains the SCS-CN-method and offers the possibility to implement new precipitation modules due to its open source license. An additional advantage of TELEMAC-2D is the good scaling on HPC cluster systems.</p><p>In this study, two different approaches for runoff creation will be compared. (1) The well-proven SCS-CN method calculates effective rain. Due to its simple structure, the process of runoff generation is completely decoupled from runoff concentration. Consequently, SCS-CN cannot account for re-infiltration due to surface runoff. (2) However, the Green-Ampt infiltration (GAI) is coupled to surface runoff as long as the water depth is non-zero. GAI is implemented recently and thus will be described in more detail. Both approaches are first tested using a simple model setup. The general model performance of the enhanced hydrodynamic rainfall-runoff modeling (EHDRRM) is verified using the case study Simbach/Triftern in Bavaria. This extreme flash flood event from 1<sup>st</sup> June 2016 hit the townships Simbach am Inn and Triftern. It is well documented and all necessary data is available in good quality. The main setup for the catchment area of 47 km² resp. 90 km² is built on a 1x1 m DEM, land use data, hydrological soil group data and 5 min-RADOLAN precipitation data. The calculated catchment outflow can be verified by measured data at the gauging stations in Simbach am Inn resp. Triftern. All comparisons include as reference results for precipitation without losses by infiltration.</p><p>The hydrodynamic precipitation runoff modeling HDRRM has proven to be a useful method for calculating flow paths, depths and velocities with a high spatial resolution during flash flood events. If the process of runoff generation is performed by the hydrodynamic model EHDRRM then the quality of results is improved significantly while keeping the modeling procedure simple. Concerning infiltration, EHDRRM allows for a physically correct representation taking the actual local water depth into consideration.</p>

Sign in / Sign up

Export Citation Format

Share Document