scholarly journals Fully distributed rainfall-runoff modeling using spatial-temporal graph neural network

2022 ◽  
Author(s):  
Zhongrun Xiang ◽  
Ibrahim Demir
Keyword(s):  
Time Series ◽  
Deep Learning ◽  
Spatial Information ◽  
Model Performance ◽  
Grid Cell ◽  
Great Promise ◽  
Rainfall Runoff ◽  
Streamflow Forecasting ◽  
Runoff Modeling ◽  
Runoff Model

Recent studies using latest deep learning algorithms such as LSTM (Long Short-Term Memory) have shown great promise in time-series modeling. There are many studies focusing on the watershed-scale rainfall-runoff modeling or streamflow forecasting, often considering a single watershed with limited generalization capabilities. To improve the model performance, several studies explored an integrated approach by decomposing a large watershed into multiple sub-watersheds with semi-distributed structure. In this study, we propose an innovative physics-informed fully-distributed rainfall-runoff model, NRM-Graph (Neural Runoff Model-Graph), using Graph Neural Networks (GNN) to make full use of spatial information including the flow direction and geographic data. Specifically, we applied a time-series model on each grid cell for its runoff production. The output of each grid cell is then aggregated by a GNN as the final runoff at the watershed outlet. The case study shows that our GNN based model successfully represents the spatial information in predictions. NRM-Graph network has shown less over-fitting and a significant improvement on the model performance compared to the baselines with spatial information. Our research further confirms the importance of spatially distributed hydrological information in rainfall-runoff modeling using deep learning, and we encourage researchers to incorporate more domain knowledge in modeling.

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 Conceptual Rainfall–Runoff Model Performance with Different Spatial Rainfall Inputs

2011 ◽  
Vol 12 (5) ◽  
pp. 1100-1112 ◽  
Author(s):  
J. Vaze ◽  
D. A. Post ◽  
F. H. S. Chiew ◽  
J.-M. Perraud ◽  
J. Teng ◽  
...  
Keyword(s):  
Model Simulation ◽  
Model Performance ◽  
Daily Rainfall ◽  
Grid Cell ◽  
Rainfall Series ◽  
Data Availability ◽  
Rainfall Time Series ◽  
Rainfall Runoff ◽  
Rainfall Runoff Model ◽  
Runoff Model

Abstract Different methods have been used to obtain the daily rainfall time series required to drive conceptual rainfall–runoff models, depending on data availability, time constraints, and modeling objectives. This paper investigates the implications of different rainfall inputs on the calibration and simulation of 4 rainfall–runoff models using data from 240 catchments across southeast Australia. The first modeling experiment compares results from using a single lumped daily rainfall series for each catchment obtained from three methods: single rainfall station, Thiessen average, and average of interpolated rainfall surface. The results indicate considerable improvements in the modeled daily runoff and mean annual runoff in the model calibration and model simulation over an independent test period with better spatial representation of rainfall. The second experiment compares modeling using a single lumped daily rainfall series and modeling in all grid cells within a catchment using different rainfall inputs for each grid cell. The results show only marginal improvement in the “distributed” application compared to the single rainfall series, and only in two of the four models for the larger catchments. Where a single lumped catchment-average daily rainfall series is used, care should be taken to obtain a rainfall series that best represents the spatial rainfall distribution across the catchment. However, there is little advantage in driving a conceptual rainfall–runoff model with different rainfall inputs from different parts of the catchment compared to using a single lumped rainfall series, where only estimates of runoff at the catchment outlet is required.

scholarly journals A note on leveraging synergy in multiple meteorological datasets with deep learning for rainfall-runoff modeling

2020 ◽  
Author(s):  
Frederik Kratzert ◽  
Daniel Klotz ◽  
Sepp Hochreiter ◽  
Grey S. Nearing
Keyword(s):  
Deep Learning ◽  
Short Term Memory ◽  
Rainfall Runoff ◽  
Data Set ◽  
Simulation Accuracy ◽  
Meteorological Forcing ◽  
Runoff Modeling ◽  
Long Short Term Memory ◽  
Rainfall Runoff Model ◽  
Runoff Model

Abstract. A deep learning rainfall-runoff model can take multiple meteorological forcing products as inputs and learn to combine them in spatially and temporally dynamic ways. This is demonstrated using Long Short Term Memory networks (LSTMs) trained over basins in the continental US using the CAMELS data set. Using multiple precipitation products (NLDAS, Maurer, DayMet) in a single LSTM significantly improved simulation accuracy relative to using only individual precipitation products. A sensitivity analysis showed that the LSTM learned to utilize different precipitation products in different ways in different basins and for simulating different parts of the hydrograph in individual basins.

scholarly journals Step wise Multi-criteria Performance Evaluation of Rainfall-Runoff Models using WETSPRO

1970 ◽  
Vol 7 (1) ◽  
pp. 18-29 ◽  
Author(s):  
PC Shakti ◽  
NK Shrestha ◽  
P Gurung
Keyword(s):  
Time Series ◽  
Performance Evaluation ◽  
Assessment Tool ◽  
Model Performance ◽  
Rainfall Runoff ◽  
Peak Flows ◽  
Low Flows ◽  
Flow Components ◽  
Rainfall Runoff Model ◽  
Runoff Model

This paper illustrates a methodology to evaluate model’s performance of rainfall runoff model using a tool called WETSPRO (Water Engineering Time Series PROcessing tool). Simulated results of physically based semidistributed model - SWAT (Soil and Water Assessment Tool) for Kliene Nete watershed (581 km2), Belgium are considered in this study. Paper presents a series of sequential time series processing tasks to be performed to evaluate model’s performance thoroughly. The problem of serial dependence and heteroscedasticity is addressed and model performance evaluation on different flow components (peak flows, low flows and volume) and flow volume is carried. Performance evaluation of both flow components on their extremes is also performed. Two most commonly used goodness-fit-statistics (Mean Square Error – MSE and Nash Sutcliff Efficiency − NSE) are used with number of complementary graphical plots for evaluation propose. Results indicated model’s robust performance on peak flows although base flows are slightly underestimated especially for lower return periods. Cumulative flow volumes tend to be overestimated. Based upon the study, some recommendations are summarized to enhance model’s ability to simulate the flows events. Keywords: Rainfall runoff model; SWAT; WETSPRO; Kliene Nete; peak flows; low flows. DOI: http://dx.doi.org/10.3126/jhm.v7i1.5613 JHM 2010; 7(1): 18-29

scholarly journals Unsupervised Multi-Level Feature Extraction for Improvement of Hyperspectral Classification

2021 ◽  
Vol 13 (8) ◽  
pp. 1602
Author(s):  
Qiaoqiao Sun ◽  
Xuefeng Liu ◽  
Salah Bourennane
Keyword(s):  
Feature Extraction ◽  
Deep Learning ◽  
Spatial Information ◽  
Hyperspectral Data ◽  
Great Promise ◽  
Learning Models ◽  
Single Level ◽  
Multiple Networks ◽  
Multi Level ◽  
Hyperspectral Classification

Deep learning models have strong abilities in learning features and they have been successfully applied in hyperspectral images (HSIs). However, the training of most deep learning models requires labeled samples and the collection of labeled samples are labor-consuming in HSI. In addition, single-level features from a single layer are usually considered, which may result in the loss of some important information. Using multiple networks to obtain multi-level features is a solution, but at the cost of longer training time and computational complexity. To solve these problems, a novel unsupervised multi-level feature extraction framework that is based on a three dimensional convolutional autoencoder (3D-CAE) is proposed in this paper. The designed 3D-CAE is stacked by fully 3D convolutional layers and 3D deconvolutional layers, which allows for the spectral-spatial information of targets to be mined simultaneously. Besides, the 3D-CAE can be trained in an unsupervised way without involving labeled samples. Moreover, the multi-level features are directly obtained from the encoded layers with different scales and resolutions, which is more efficient than using multiple networks to get them. The effectiveness of the proposed multi-level features is verified on two hyperspectral data sets. The results demonstrate that the proposed method has great promise in unsupervised feature learning and can help us to further improve the hyperspectral classification when compared with single-level features.

Physics-guided deep learning for rainfall-runoff modeling by considering extreme events and monotonic relationships

2021 ◽  
pp. 127043
Author(s):  
Kang Xie ◽  
Pan Liu ◽  
Jianyun Zhang ◽  
Dongyang Han ◽  
Guoqing Wang ◽  
...  
Keyword(s):  
Deep Learning ◽  
Extreme Events ◽  
Rainfall Runoff ◽  
Runoff Modeling

scholarly journals Multi-time-scale input approaches for hourly-scale rainfall–runoff modeling based on recurrent neural networks

2021 ◽  
Author(s):  
Kei Ishida ◽  
Masato Kiyama ◽  
Ali Ercan ◽  
Motoki Amagasaki ◽  
Tongbi Tu
Keyword(s):  
Time Series ◽  
Time Scale ◽  
Time Series Data ◽  
Series Data ◽  
Computational Time ◽  
Rainfall Runoff ◽  
Runoff Modeling ◽  
Input Time ◽  
Target Data ◽  
Resolution Data

Abstract This study proposes two effective approaches to reduce the required computational time of the training process for time-series modeling through a recurrent neural network (RNN) using multi-time-scale time-series data as input. One approach provides coarse and fine temporal resolutions of the input time-series data to RNN in parallel. The other concatenates the coarse and fine temporal resolutions of the input time-series data over time before considering them as the input to RNN. In both approaches, first, the finer temporal resolution data are utilized to learn the fine temporal scale behavior of the target data. Then, coarser temporal resolution data are expected to capture long-duration dependencies between the input and target variables. The proposed approaches were implemented for hourly rainfall–runoff modeling at a snow-dominated watershed by employing a long short-term memory network, which is a type of RNN. Subsequently, the daily and hourly meteorological data were utilized as the input, and hourly flow discharge was considered as the target data. The results confirm that both of the proposed approaches can reduce the required computational time for the training of RNN significantly. Lastly, one of the proposed approaches improves the estimation accuracy considerably in addition to computational efficiency.

scholarly journals The HYDROMED model and its application to semi-arid Mediterranean catchments with hill reservoirs 2: Rainfall-runoff model applications to three Mediterranean hill reservoirs

2001 ◽  
Vol 5 (4) ◽  
pp. 554-562 ◽  
Author(s):  
R. Ragab ◽  
D. Moidinis ◽  
J. Albergel ◽  
J. Khouri ◽  
A. Drubi ◽  
...  
Keyword(s):  
Goodness Of Fit ◽  
Model Performance ◽  
Limited Information ◽  
Rainfall Runoff ◽  
Significant Events ◽  
Reliable Parameter ◽  
Rainfall Runoff Model ◽  
Parameter Values ◽  
Runoff Model

Abstract. The objective of this work was to assess the performance of the newly developed HYDROMED model. Three catchments with hill reservoirs were selected. They are El-Gouazine and Kamech in Tunisia and Es Sindiany in Syria. The rainfall, the spillway flow and volume of water in the reservoirs were used as input to the model. Events that generated spillway flow were preferred for calibration. The results confirmed that the HYDROMED model is capable of reproducing the runoff volume at all the three sites. In calibrating single events, the model performance was high as measured by the Nash-Sutcliffe criterion for goodness of fit. In some events this value was as high as 98%. In simulation mode, the highest Nash-Sutcliffe criterion value was close to 70% in the El-Gouazine and Kamech catchments and close to 50% in the Es Sindiany catchment. Given the limited information available, especially on the unrecorded releases in the three catchments, the hydrological impact of site geology (e.g. Kamech), the unrecorded operator intervention during the spillway flow (e.g. Es Sindiany) and other unaccounted factors (e.g siltation, evaporation, etc.), these results are by and large very encouraging. However, they could be further improved as and when more information on the unrecorded parameters becomes available. Additionally, the results of this work highlighted the need for long term records with a large number of significant events that are able to generate spillway flow to obtain more consistent and reliable parameter values. It also highlights the need for more accurately recorded releases for irrigation and other uses. As these results are encouraging, more tests on those three and other sites are planned. Keywords: HYDROMED, rainfall-runoff model, Mediterranean, conceptual model

scholarly journals Can a Calibration-Free Dynamic Rainfall‒Runoff Model Predict FDCs in Data-Scarce Regions? Comparing the IDW Model with the Dynamic Budyko Model in South India

Hydrology ◽  
2019 ◽  
Vol 6 (2) ◽  
pp. 32 ◽  
Author(s):  
Nag ◽  
Biswal
Keyword(s):  
Model Performance ◽  
Data Availability ◽  
Ungauged Basins ◽  
Rainfall Runoff ◽  
Discharge Data ◽  
Study Region ◽  
Calibration Free ◽  
Rainfall Runoff Model ◽  
Runoff Model ◽  
Free Dynamic

Construction of flow duration curves (FDCs) is a challenge for hydrologists as most streams and rivers worldwide are ungauged. Regionalization methods are commonly followed to solve the problem of discharge data scarcity by transforming hydrological information from gauged basins to ungauged basins. As a consequence, regionalization-based FDC predictions are not very reliable where discharge data are scarce quantitatively and/or qualitatively. In such a scenario, it is perhaps more meaningful to use a calibration-free rainfall‒runoff model that can exploit easily available meteorological information to predict FDCs in ungauged basins. This hypothesis is tested in this study by comparing a well-known regionalization-based model, the inverse distance weighting (IDW) model, with the recently proposed calibration-free dynamic Budyko model (DB) in a region where discharge observations are not only insufficient quantitatively but also show apparent signs of observational errors. The DB model markedly outperformed the IDW model in the study region. Furthermore, the IDW model’s performance sharply declined when we randomly removed discharge gauging stations to test the model in a variety of data availability scenarios. The analysis here also throws some light on how errors in observational datasets and drainage area influence model performance and thus provides a better picture of the relative strengths of the two models. Overall, the results of this study support the notion that a calibration-free rainfall‒runoff model can be chosen to predict FDCs in discharge data-scarce regions. On a philosophical note, our study highlights the importance of process understanding for the development of meaningful hydrological models.

Exploring signatures-based calibration of hydrological models for prediction in ungauged basins.

2020 ◽  
Author(s):  
Marco Dal Molin ◽  
Dmitri Kavetski ◽  
Mario Schirmer ◽  
Fabrizio Fenicia
Keyword(s):  
Time Series ◽  
Accurate Estimation ◽  
Ungauged Basins ◽  
Rainfall Runoff ◽  
Multiple Sources ◽  
Important Challenge ◽  
Rainfall Runoff Model ◽  
Approximate Bayesian ◽  
Correct Estimation ◽  
Runoff Model

<p>One of the open challenges in catchment hydrology is prediction in ungauged basins (PUB), i.e. being able to predict catchment responses (typically streamflow) when measurements are not available. One of the possible approaches to this problem consists in calibrating a model using catchment response statistics (called signatures) that can be estimated at the ungauged site.<br>An important challenge of any approach to PUB is to produce reliable and precise predictions of catchment response, with an accurate estimation of the uncertainty. In the context of PUB through calibration on regionalized streamflow signatures, there are multiple sources of uncertainty that affect streamflow predictions, which relate to:</p><ul><li>The use streamflow signatures, which, by synthetizing the underlying time series, reduce the information available for model calibration;</li> <li>The regionalization of streamflow signatures, which are not observed, but estimated through some signature regionalization model;</li> <li>The use of a rainfall-runoff model, which carries uncertainties related to input data, parameter values, and model structure.</li> </ul><p>This study proposes an approach that separately accounts for the uncertainty related to the regionalization of the signatures from the other types; the implementation uses Approximate Bayesian Computation (ABC) to infer the parameters of the rainfall-runoff model using stochastic streamflow signatures. <br>The methodology is tested in six sub-catchments of the Thur catchment in Switzerland; results show that the regionalized model produces streamflow time series that are similar to the ones obtained by the classical time-domain calibration, with slightly higher uncertainty but similar fit to the observed data. These results support the proposed approach as a viable method for PUB, with a focus on the correct estimation of the uncertainty.</p>

Sign in / Sign up

Export Citation Format

Share Document