Semi-distributed Model Research Articles

Hydrologically informed machine learning for rainfall–runoff modelling: towards distributed modelling

Abstract. Despite showing great success of applications in many commercial fields, machine learning and data science models generally show limited success in many scientific fields, including hydrology (Karpatne et al., 2017). The approach is often criticized for its lack of interpretability and physical consistency. This has led to the emergence of new modelling paradigms, such as theory-guided data science (TGDS) and physics-informed machine learning. The motivation behind such approaches is to improve the physical meaningfulness of machine learning models by blending existing scientific knowledge with learning algorithms. Following the same principles in our prior work (Chadalawada et al., 2020), a new model induction framework was founded on genetic programming (GP), namely the Machine Learning Rainfall–Runoff Model Induction (ML-RR-MI) toolkit. ML-RR-MI is capable of developing fully fledged lumped conceptual rainfall–runoff models for a watershed of interest using the building blocks of two flexible rainfall–runoff modelling frameworks. In this study, we extend ML-RR-MI towards inducing semi-distributed rainfall–runoff models. The meaningfulness and reliability of hydrological inferences gained from lumped models may tend to deteriorate within large catchments where the spatial heterogeneity of forcing variables and watershed properties is significant. This was the motivation behind developing our machine learning approach for distributed rainfall–runoff modelling titled Machine Induction Knowledge Augmented – System Hydrologique Asiatique (MIKA-SHA). MIKA-SHA captures spatial variabilities and automatically induces rainfall–runoff models for the catchment of interest without any explicit user selections. Currently, MIKA-SHA learns models utilizing the model building components of two flexible modelling frameworks. However, the proposed framework can be coupled with any internally coherent collection of building blocks. MIKA-SHA's model induction capabilities have been tested on the Rappahannock River basin near Fredericksburg, Virginia, USA. MIKA-SHA builds and tests many model configurations using the model building components of the two flexible modelling frameworks and quantitatively identifies the optimal model for the watershed of concern. In this study, MIKA-SHA is utilized to identify two optimal models (one from each flexible modelling framework) to capture the runoff dynamics of the Rappahannock River basin. Both optimal models achieve high-efficiency values in hydrograph predictions (both at catchment and subcatchment outlets) and good visual matches with the observed runoff response of the catchment. Furthermore, the resulting model architectures are compatible with previously reported research findings and fieldwork insights of the watershed and are readily interpretable by hydrologists. MIKA-SHA-induced semi-distributed model performances were compared against existing lumped model performances for the same basin. MIKA-SHA-induced optimal models outperform the lumped models used in this study in terms of efficiency values while benefitting hydrologists with more meaningful hydrological inferences about the runoff dynamics of the Rappahannock River basin.

A framework for automated and spatially-distributed modeling with the Agricultural Policy Environmental eXtender (APEX) model

Agricultural Best Management Practices (BMPs) are popular approaches to reduce nonpoint source (NPS) pollutant losses. Hydrologic models that can simulate impacts of BMPs at the field-scale can help guide the selection of BMPs. Furthermore, high-performance computing techniques have significant potential for scaling spatial simulations and reducing model runtimes. In this study, a parallel modeling framework for the Agricultural Policy Environmental eXtender (APEX) model was developed for large-scale, high-resolution, spatially-distributed model simulations. It provides a tool for conducting BMP evaluations at field-scale with a distributed architecture and automatic model setup of APEX. Sample results demonstrated the capability of the framework for distributed and semi-distributed modeling and illustrated the performance of parallelization. This framework can help provide guidance for decision makers on agricultural BMPs with large-scale water quality assessments and NPS nutrient loading reductions.

Time‐domain signal and noise modelling and analysis of millimetre‐wave and THz high electron mobility transistors

This study presents the simultaneous signal and noise modelling and analysis of mm-wave/THz high electron mobility transistors (HEMTs) in the linear regime as stochastic active multi-conductor transmission lines (SAMTLs). For such devices, stochastic inhomogeneous telegrapher equations of SAMTLs are extracted assuming the domination of the quasi-transverse electromagnetic mode. Analytical solutions for these equations do not exist. So a numerical analysis of such equations is introduced here at millimetre-wave and terahertz frequencies. These equations are analysed with a numerical method in the time domain, namely the finite-integration technique. A semi-distributed model in the advanced design system commercial software is used to validate the numerical results. Also, the results of noise figure and scattering parameters (S-parameters) for the metal–semiconductor field-effect transistor transistor, which has the same SAMTLs model as the HEMT transistor, are compared with the measurement reports at microwave frequencies.

Sensitivity and uncertainty analysis for streamflow prediction based on multiple optimization algorithms in Yalong River Basin of southwestern China

Semi-distributed model based on spatial information is effective for simulating hydrologic cycle. However, it cannot completely simulate all relevant natural processes. Uncertainty analysis is necessary for achieving high accuracy of hydrological modelling. In this study, optimization algorithms and weighted network analysis techniques were adopted to explore hydrological parameter combination characteristics as well as hydrological simulating uncertainties in Yalong River Basin (YLRB) of southwestern China based on Sequential Uncertainty Fitting version 2 (SUFI-2), Generalized Likelihood Uncertainty Estimation (GLUE) and Particle Swarm Optimization (PSO). The results indicated that: a) groundwater recession factor, effective hydraulic conductivity in channel and saturated hydraulic conductivity could significantly affect streamflow in the studying basin, b) complex parameter combination responded diversely under aforementioned three optimization algorithms. Comparatively, GLUE brought out higher autocorrelation parameter network and c) SUFI-2 and PSO performed better in terms of effective uncertainty analysis and fitting values than those of GLUE. This work could provide references and insights for sensitive parameter modification and prediction uncertainty reduction of streamflow simulation, furthermore contributing to an optimal water resource management.

Flood hazard mapping for data-scarce and ungauged coastal river basins using advanced hydrodynamic models, high temporal-spatial resolution remote sensing precipitation data, and satellite imageries

This paper presents an integrated approach to simulate flooding and inundation for small- and medium-sized coastal river basins where measured data are not available or scarce. By coupling the rainfall–runoff model, the one-dimensional and two-dimensional models, and the integration of these with global tide model, satellite precipitation products, and synthetic aperture radar imageries, a comprehensive flood modeling system for Tra Bong river basin selected as a case study was set up and operated. Particularly, in this study, the lumped conceptual model was transformed into the semi-distributed model to increase the parameter sets of donor basins for applying the physical similarity approach. The temporal downscaling technique was applied to disaggregate daily rainfall data using satellite-based precipitation products. To select an appropriate satellite-derived rainfall product, two high temporal-spatial resolution products (0.1 × 0.1 degrees and 1 h) including GSMaP_GNRT6 and CMORPH_CRT were examined at 1-day and 1-h resolutions by comparing with ground-measured rainfall. The CMORPH_CRT product showed better performance in terms of statistical errors such as Correlation Coefficient, Probability of Detection, False Alarm Ratio, and Critical Success Index. Land cover/land use, flood extent, and flood depths derived from Sentinel-1A imageries and a digital elevation model were employed to determine the surface roughness and validate the flood modeling. The results obtained from the modeling system were found to be in good agreement with collected data in terms of NSE (0.3–0.8), RMSE (0.19–0.94), RPE (− 213 to 0.7%), F1 (0.55), and F2 (0.37). Subsequently, various scenarios of flood frequency with 10-, 20-, 50-, and 100-year return periods under the probability analysis of extreme values were developed to create the flood hazard maps for the study area. The flood hazards were then investigated based on the flood intensity classification of depth, duration, and velocity. These hazard maps are significantly important for flood hazard assessments or flood risk assessments. This study demonstrated that applying advanced hydrodynamic models on computing flood inundation and flood hazard analysis in data-scarce and ungauged coastal river basins is completely feasible. This study provides an approach that can be used also for other ungauged river basins to better understand flooding and inundation through flood hazard mapping.

Karst modelling challenge 1: Results of hydrological modelling

The complexity of karst groundwater flow modelling is reflected by the amount of simulation approaches. The goal of the Karst Modelling Challenge (KMC) is comparing different approaches on one single system using the same data set. Thirteen teams with different computational models for simulating discharge variations at karst springs have applied their respective models on one single data set coming from the Milandre Karst Hydrogeological System (MKHS). The approaches include neural networks, reservoir models, semi-distributed models and fully distributed groundwater models. Four and a half years of hourly or daily meteorological input and hourly discharge data were provided for model calibration. The validation comprised forecasting one year of discharge, without the observed discharge data. The model performance was evaluated using the volume conservation, Nash-Sutcliffe efficiency (NSE) and the Kling-Gupta efficiency (KGE) applied on the total discharge and individual flow components. As a result, the comparison of model performances is a challenging task due to the differences in the model architecture but also required time steps: some of the models require aggregated daily steps while others could be run using hourly data, which provided some interesting differences depending on how the data was transformed. The use of instantaneous data (e.g. value at noon) produces less bias that averaging hourly data over one day. The transformation of hourly into daily data produces a decrease of Nash and KGE of 0.05 to 0.08 (i.e. from 1 to ~0.93). The resulting simulations (forecasted values for year 2016) produced KGEs ranging between 0.83 and 0.37 (0.83 to −0.24 for NSE). Although the simulations matched the monitored flows reasonably well, most models struggled to simulate baseflow conditions accurately. In general, the models that performed the best for this exercise were the global ones (Gardenia and Varkarst), with a limited number of parameters, which can be calibrated using automatic calibration procedures. The neural network models also showed a fair potential, with one providing reasonable results despite the relatively short dataset available for warming-up (4.5 years). Semi-and fully distributed models also suggested that with some more effort they could perform well. The accuracy of model predictions does not seem to increase by using models with more than 9–12 calibration parameters. An evaluation of the relative errors between the forecasted and the observed values revealed that for most models, 50% of the forecasted values contained more than 50% of difference against the observed discharge rate, with 25% having a difference larger than 100%. A significant part of the poorly forecasted values corresponded to base-flow which was surprising given that as base-flow is generally much easier to predict than peak flow. Hence, this shows that modelling approaches and criteria for the calibration are too oriented towards peak-flow sections of the hydrographs, and that improvements could be gained by more focus on the base-flow.

Multi-Spatial Resolution Rainfall-Runoff Modelling—A Case Study of Sabari River Basin, India

One of the challenges in rainfall-runoff modeling is the identification of an appropriate model spatial resolution that allows streamflow estimation at customized locations of the river basin. In lumped modeling, spatial resolution is not an issue as spatial variability is not accounted for, whereas in distributed modeling grid or cell resolution can be related to spatial resolution but its application is limited because of its large data requirements. Streamflow estimation at the data-poor customized locations is not possible in lumped modeling, whereas it is challenging in distributed modeling. In this context, semi-distributed modeling offers a solution including model resolution and estimation of streamflow at customized locations of a river basins with less data requirements. In this study, the Hydrologic Engineering Center-Hydrologic Modeling System (HEC-HMS) model is employed in semi-distribution mode on river basins of six different spatial resolutions. The model was calibrated and validated for fifteen and three selected flood events, respectively, of three types, i.e., single peak (SP), double peak (DP)- and multiple peaks (MP) at six different spatial resolution of the Sabari River Basin (SRB), a sub-basin of the Godavari basin, India. Calibrated parameters were analyzed to understand hydrologic parameter variability in the context of spatial resolution and flood event aspects. Streamflow hydrographs were developed, and various verification metrics and model scores were calculated for reference- and calibration- scenarios. During the calibration phase, the median of correlation coefficient and NSE for all 15 events of all six configurations was 0.90 and 0.69, respectively. The estimated streamflow hydrographs from six configurations suggest the model’s ability to simulate the processes efficiently. Parameters obtained from the calibration phase were used to generate an ensemble of streamflow at multiple locations including basin outlet as part of the validation. The estimated ensemble of streamflows appeared to be realistic, and both single-valued and ensemble verification metrics indicated the model’s good performance. The results suggested better performance of lumped modeling followed by the semi-distributed modeling with a finer spatial resolution. Thus, the study demonstrates a method that can be applied for real-time streamflow forecast at interior locations of a basin, which are not necessarily data rich.

The Performance of Physically Based and Conceptual Hydrologic Models: A Case Study for Makkah Watershed, Saudi Arabia

Population growth and land use modification in urban areas require the use of accurate tools for rainfall-runoff modeling, especially where the topography is complex. The recent improvement in the quality and resolution of remotely sensed precipitation satisfies a major need for such tools. A physically-based, fully distributed hydrologic model and a conceptual semi-distributed model, forced by satellite rainfall estimates, were used to simulate flooding events in a very arid, rapidly urbanizing watershed in Saudi Arabia. Observed peak discharge for two flood events was used to compare hydrographs simulated by the two models, one for calibration and one for validation. To further explore the effect of watershed heterogeneity, the hydrographs produced by three implementations of the conceptual were compared against each other and against the output of the physically-based model. The results showed the ability of the distributed models to capture the effect of the complex topography and variability of land use and soils of the watershed. In general, the GSSHA model required less calibration and performed better than HEC-HMS. This study confirms that the semi-distributed HEC-HMS model cannot be used without calibration, while the GSSHA model can be the best option in the case of a lack of data. Although the two models showed good agreement at the calibration point, there were significant differences in the runoff, discharge, and infiltration values at interior points of the watershed.

Improving Runoff Simulation and Forecasting with Segmenting Delay of Baseflow from Fast Surface Flow in Montane High-Vegetation-Covered Catchments

Due to the complicated terrain conditions in montane catchments, runoff formation is fast and complicated, making accurate simulation and forecasting a significant hydrological challenge. In this study, the spatiotemporal variable source mixed runoff generation module (SVSMRG) was integrated with the long short-term memory (LSTM) method, to develop a semi-distributed model (SVSMRG)-based surface flow and baseflow segmentation (SVSMRG-SBS). Herein, the baseflow was treated as a black box and forecasted using LSTM, while the surface flow was simulated using the SVSMRG module based on hydrological response units (HRUs) constructed using eco-geomorphological units. In the case study, four typical montane catchments with different climatic conditions and high vegetation coverage, located in the topographically varying mountains of the eastern Tibetan Plateau, were selected for runoff and flood process simulations using the proposed SVSMRG-SBS model. The results showed that this model had good performance in hourly runoff and flood process simulations for montane catchments. Regarding runoff simulations, the Nash–Sutcliffe efficiency coefficient (NSE) and correlation coefficient (R2) reached 0.8241 and 0.9097, respectively. Meanwhile, for the flood simulations, the NSE ranged from 0.5923 to 0.7467, and R2 ranged from 0.6669 to 0.8092. For the 1-, 3-, and 5-h baseflow forecasting with the LSTM method, it was found that model performances declined when simulating the runoff processes, wherein the NSE and R2 between the measured and modeled runoff decreased from 0.8216 to 0.8087 and from 0.9095 to 0.8871, respectively. Similar results were found in the flood simulations, the NSE and R2 values declined from 0.7414–0.5885 to 0.7429–0.5716 and from 0.8042–0.6547 to 0.7936–0.6067, respectively. This means that this new model achieved perfect performance in montane catchment runoff and flood simulation and forecasting with 1-, 3-, 5-h steps. Therefore, as it considers vegetation regulation, the SVSMRG-SBS model is expected to improve runoff and flood simulation accuracy in montane high-vegetation-covered catchments.

Optimization of Hydrologic Response Units (HRUs) Using Gridded Meteorological Data and Spatially Varying Parameters

Although complex hydrological models with detailed physics are becoming more common, lumped and semi-distributed models are still used for many applications and offer some advantages, such as reduced computational cost. Most of these semi-distributed models use the concept of the hydrological response unit or HRU. In the original conception, HRUs are defined as homogeneous structured elements with similar climate, land use, soil and/or pedotransfer properties, and hence a homogeneous hydrological response under equivalent meteorological forcing. This work presents a quantitative methodology, called hereafter the principal component analysis and hierarchical cluster analysis or PCA/HCPC method, to construct HRUs using gridded meteorological data and hydrological parameters. The PCA/HCPC method is tested using the water evaluation and planning system (WEAP) model for the Alicahue River Basin, a small and semi-arid catchment of the Andes, in Central Chile. The results show that with four HRUs, it is possible to reduce the relative within variance of the catchment up to about 10%, an indicator of the homogeneity of the HRUs. The evaluation of the simulations shows a good agreement with streamflow observations in the outlet of the catchment with an Nash–Sutcliffe efficiency (NSE) value of 0.79 and also shows the presence of small hydrological extreme areas that generally are neglected due to their relative size.

Caracterización Hídrica de la Cuenca del Arroyo Vaquerías (Córdoba, Argentina) Mediante el Modelado Hidrológico e Información Satelital

Se presenta una caracterización hidrológica, mediante el modelado computacional, de la cuenca del arroyo Vaquerías. Esta cuenca, carente de cualquier sistema de monitoreo, representa la principal fuente del recurso hídrico de la Reserva Natural Vaquerías, perteneciente a la Universidad Nacional de Córdoba. Como fuente de información hidroclimática se recurrió, para la precipitación, al producto CHIRPS (Climate Hazards group InfraRed Precipitation with Station data), y para la evapotranspiración, al producto MOD16A2. Los datos fueron procesados en la plataforma Google Earth Engine. El modelado se realizó a través del modelo semidistribuido TOPMODEL, que en forma simplificada considera la contribución al escurrimiento de los flujos superficial y subterráneo. A pesar de las limitaciones de la información de entrada como del modelo implementado, y en el contexto de la ausencia de información de campo, fue posible caracterizar hidrológicamente la cuenca en estudio.

Soil Moisture Simulation in Selected Austrian Catchements With Use of the Tuw Conceptual Semi-Distributed Rainfall Runoff Model

Nowadays, hydrological rainfall-runoff models are routinely used for modelling hydrological balances, the generation of floods, or droughts. For decades numerous rainfallrunoff models have been tested with many calibration procedures and different model structures. However, for rainfall-runoff, it is challenging to find a particular model structure and parameter set that can correctly describe the complicated flow formation processes in diverse physiographic conditions. Improvements to existing rainfall-runoff modelling concepts and data assimilation are therefore continuously being tested. In recent years, remote sensing has played an increasing role in the surveying of hydrological phenomena. Remote sensing of soil moisture data can be very helpful because soil moisture measurements in field conditions are not always straightforward. The quality of remotely sensed data is rising; nowadays, we can routinely start using data with proper spatial and temporal resolutions. In this paper, we have focused on an evaluation of the parametrisation of the soil moisture submodel of the TUW rainfall-runoff model by remotely sensed soil moisture data. We calibrated the TUW model for three selected catchments in Austria with flat hypsometric characteristics using discharges as a criterion. For the calibration, we used both the lumped and semi-distributed model versions of the model and compared the quality of the soil moisture of both versions. Both the lumped and semi-distributed versions performed well in the discharge simulation. In the case of the soil moisture simulation, we achieved slightly better results with the semi-distributed version of the model. Difficulties with accessing the data from the remote sensing are discussed since remote sensing sensors still have problems when clouds and snow cover the catchments.

Comparison of the variances of a lumped and semi-distributed model parameters

Comparison of the variances of a lumped and semi-distributed model parameters

An evaluation of semidistributed-pipe-network and distributed-finite-difference models to simulate karst systems

Several different approaches have been developed to model the specific characteristics of karst aquifers, taking account of their inherent complex spatial and temporal heterogeneities. This paper sets out the development of a semidistributed modelling approach for applications in an Irish karst context using urban drainage software. The models have proven to be very useful for different studies, with examples given for the ecohydrology of ephemeral karst lakes, extreme groundwater-flood alleviation, karst network investigation, submarine groundwater discharge, and quantification of different recharge and flow components. The limitations of the approach are also highlighted, in particular not being able to simulate diffuse infiltration and flow paths explicitly across the groundwater catchment. Hence, a more distributed, finite-difference modelling approach using MODFLOW Unstructured Grid (USG) with the newly developed Connected Linear Network (CLN) process is then compared against the semidistributed approach on the same karst catchment. Whilst it has proven difficult to achieve the same levels of model performance in simulating the spring flows in the distributed model compared to the semidistributed model, the ability to interrogate the flow paths at any point on the three-dimensional aquifer is demonstrated, which can give new insights into flows (and potential contaminant transport) through such complex systems. The influence of the proximity of highly transmissive conduits on the flow dynamics through the much-lower transmissive matrix cells in which the network is embedded has been particularly investigated.

Evaluating the Effects of Watershed Subdivision on Hydrological Simulation by Swat Model in an Arctic Watershed

The hydrological model SWAT is a state-of-the-art tool for environmental and water resources management. Like other semi-distributed models, the whole river basin in the SWAT model is delineated into smaller sub-basins prior to conducting the simulation. Watershed delineation is an importance step since it could potentially influence the modelling results. The present study aimed to conducte an investigation of the effects of watershed delineation schemes on hydrological simulation in an Arctic watershed Målselv, north of Norway. Four delineation schemes were set up with different threshold drainage area (TDA) from fine to coarse including 100 ha, 2,000 ha, 5,000 ha and 10,000 ha. The model was run on monthly time step from 1979-2012. The results showed higher variation of average annual precipitation and runoff especially in the upstream sections of the watershed by the fine TDA schemes compared to the coarse ones. The average monthly precipitation and runoff slightly increased from the finest TDA scheme to the coarsest scheme. All TDA schemes reproduced the observed tendency of the average monthly and annual streamflow although the peak flow was over and under estimated at different hydro-gauging stations. The higher value of estimated streamflow was found at the coarsest scheme.

Is hillslope-based catchment decomposition approach superior to hydrologic response unit (HRU) for stream-aquifer interaction modelling: Inference from two process-based coupled models

To quantify the surface water-groundwater (SW-GW) exchange flux in a river basin, the inclusion of SW-GW interaction (SGI) process in the next generation earth system models is of growing concern among the hydrologic community that requires a trade-off among model complexity linking streamflow and groundwater flow domains, spatial heterogeneity, and computational demand. Although semi-distributed SGI models seem to be a plausible solution, the question is ‘what could be the effective way to decompose a river basin to assess the sub-surface flux’? Therefore, this study intends to search for a better alternative approach for catchment decomposition between the hillslope- and hydrologic response unit (HRU)-based SGI models; the inherent relationship between the decomposition level and model efficiency; and the appropriate spatial scale of decomposition for hillslopes and HRUs. The typical hydrological models chosen for this study are: the hillslope-based coupled Variable Parameter McCarthy-Muskingum–enhanced hillslope storage Boussinesq (VPMM-ehsB), and the HRU-based coupled Soil and Water Assessment Tool–MODular three-dimensional groundwater FLOW (SWAT-MODFLOW). These semi-distributed models are field-tested to model the streamflow accounting for lateral contributions along the river reach, baseflow, depth to water table, and the SGI flux. The results reveal that the VPMM-ehsB model with hillslope-grid based decomposition outperformed the SWAT-MODFLOW with the HRU-grid based hybrid decomposition in explaining the variance of the entire annual and low (non-monsoon) streamflow time series by 12% and 25%, respectively. The VPMM-ehsB approach has a faster (36%) simulation time than that of the SWAT-MODFLOW showing the potential of the hillslope decomposition approach in the VPMM-ehsB model over the HRU-based approach in the SWAT-MODFLOW for large-scale application in future earth system models.

A review of the adaptability of hydrological models for drought forecasting

Abstract. Drought intensity and frequency are increasing in recent years in multiple regions across the world due to global climate change and consequently drought forecasting research has received more and more attention. Previous studies on drought forecasting mostly focus on meteorological drought based on precipitation and temperature. However, the trend of predicting agriculture and hydrological drought, which consider soil moisture and runoff, have developed rapidly in recent years. Hydrological drought forecasting is based on the hydrological models and the model structure plays a role to improve predictions. This study scrutinized more than 50 hydrological models, including lumped models, semi-distributed models, distributed models, surface water and groundwater coupled models, to explore the adaptability of hydrological models in drought simulation and forecasting. The advantages and disadvantages of typical models, such as DTVGM, GWAVA, and HEC-HMS models were analyzed to provide valuable reference for drought forecasting model development. Future work aims at improving the hydrological models to simulate the drought processes and make better prediction.

Sensitivity Analysis of Rainfall-Runoff Model in Malino Sub-Watershed

Overtime, science has been developing and trying to use modelling system in the management of hydrology. In this paper, we present a sensitivity analysis for the HEC-HMS rainfall-runoff model. Sensitivity analysis is used to determine the parameter which most influences the result of the model. The main objective of this research is to investigate the sensitivity of parameter in model and to analyze the influence of the graph types used for unit hydrograph. In addition, we also compare the result of lumped and semi-distributed models based on stream order to analyse a common watershed issue. This study will make a model of rainfall-runoff using a sub-catchment distribution method based on stream order. ArcGIS and HEC-HMS were used as softwares to process data. The data needed in this study come from secondary data sourced from various sources. The data that will be processed first is DEM (Digital Elevation Model). DEM used in this study was sourced from the Geospatial Information Agency or commonly referred to as DEMNAS (The National DEM). ArcGIS will process geospatial data in the form of DEM which data results can be used directly to make model in HEC-HMS. Then, the results of the rainfall-runoff model will be calibrated using observed unit hydrograph. From the result of sensitivity analysis, it was concluded that impervious is the most sensitive parameter. Moreover, this paper also can be used as guidelines for use in rainfall-runoff modelling on HEC-HMS.

Assessing Hydrological Vulnerability to Future Droughts in a Mediterranean Watershed: Combined Indices-Based and Distributed Modeling Approaches

Understanding the spatiotemporal distribution of future droughts is essential for effective water resource management, especially in the Mediterranean region where water resources are expected to be scarcer in the future. In this study, we combined meteorological and hydrological drought indices with the Soil and Water Assessment Tool (SWAT) model to predict future dry years during two periods (2035–2050and 2085–2100) in a typical Mediterranean watershed in Northern Morocco, namely, Bouregreg watershed. The developed methodology was then used to evaluate drought impact on annual water yields and to identify the most vulnerable sub-basins within the study watershed. Two emission scenarios (RCP4.5 and RCP8.5) of a downscaled global circulation model were used to force the calibrated SWAT model. Results indicated that Bouregreg watershed will experience several dry years with higher frequency especially at the end of current century. Significant decreases of annual water yields were simulated during dry years, ranging from −45.6% to −76.7% under RCP4.5, and from −66.7% to −95.6% under RCP8.5, compared to baseline. Overall, hydrologic systems in sub-basins under the ocean or high-altitude influence appear to be more resilient to drought. The combination of drought indices and the semi-distributed model offer a comprehensive tool to understand potential future droughts in Bouregreg watershed.

Distributed long-term hourly streamflow predictions using deep learning – A case study for State of Iowa

Accurate streamflow forecasting has always been a challenge. Although there were many novel approaches using deep learning models, accuracy of these models is often limited to a short lead time. This study proposes a new deep recurrent neural network approach, Neural Runoff Model (NRM), which has been applied on 125 USGS streamflow gages in the State of Iowa for predicting the next 120 h. We use a semi-distributed model structure with observation and forecast data from the model output of upstream stations as additional input for downstream gages. The proposed model outperforms the streamflow persistence, ridge regression and random forest regression on majority of the gages. Our model has shown strong predictive power and can be used for long-term streamflow predictions. This study also shows that the semi-distributed structure in NRM can improve the streamflow predictions by integrating water level data from upstream stream gauges.