Groundwater Estimation from Major Physical Hydrology Components Using Artificial Neural Networks and Deep Learning

Abstract

Precise estimation of physical hydrology components including groundwater levels (GWLs) is a challenging task, especially in relatively non-contiguous watersheds. This study estimates GWLs with deep learning and artificial neural networks (ANNs), namely a multilayer perceptron (MLP), long short term memory (LSTM), and a convolutional neural network (CNN) with four different input variable combinations for two watersheds (Baltic River and Long Creek) in Prince Edward Island, Canada. Variables including stream level, stream flow, precipitation, relative humidity, mean temperature, evapotranspiration, heat degree days, dew point temperature, and evapotranspiration for the 2011-2017 period were used as input variables. Using a hit and trial approach and various hyperparameters, all ANNs were trained from scratched (2011)(2012)(2013)(2014)(2015) and validated (2016)(2017). The stream level was the major contributor to GWL fluctuation for the Baltic River and Long Creek watersheds (R 2 = 50.8 and 49.1%, respectively). The MLP performed better in validation for Baltic River and Long Creek watersheds (RMSE = 0.471 and 1.15, respectively). Increased number of variables from 1 to 4 improved the RMSE for the Baltic River watershed by 11% and for the Long Creek watershed by 1.6%. The deep learning techniques introduced in this study to estimate GWL fluctuations are convenient and accurate as compared to collection of periodic dips based on the groundwater monitoring wells for groundwater inventory control and management.Water 2020, 12, 5 2 of 18 procedure that requires thorough knowledge of physical hydrological parameters, big data, hydrological models, model inputs, and the geometry of watersheds [3]. Aspects of hydrogeology-i.e., geological factors affecting the distribution and movement of groundwater underneath the soil surface-need to be properly understood when modeling GWLs and manipulating the modeling results. Watershed scale fluctuations in GWLs occur over a period of several decades, and the resulting cumulative effects on streamflow depletion may not be fully realized for years [4]. Resultantly, depending upon the distance of the pumping station from the stream and the geologic characteristics of the aquifer, the groundwater system may take decades to recover from streamflow depletion caused by intermittent pumping. Components of the surface-and the sub-surface physical hydrology of a watershed-i.e., streamflow and groundwater flow, respectively-are interconnected, making the stream-aquifer interaction one of the key processes governing the groundwater flow pattern in a watershed [1]. Groundwater fluctuations affect streamflow and vice versa, as the pumping wells capture groundwater that would otherwise discharge to connected streams, rivers, and other surface-water bodies [4]. Francis [5] reported that, in typical watersheds of Prince Edward Island, the base flow represents almost 80% of the streamflow in the late summer and fall months. Stream length in these island watersheds ranges from less than 1 km to 2...

Computation of Evapotranspiration with Artificial Intelligence for Precision Water Resource Management

“…However, because of data normalization of the datasets, the RNNs used in the study were successfully converged without over and under fitting with approximately equal training and testing accuracies. Similar results were found in a study by Afzaal et al, [34] as no major effect of dropout was observed with normalized data. Therefore, all the RNNs used in this study were trained without introducing dropout in LSTM layers.…”

“…The root means square error (RMSE) and R 2 were used to evaluate the model effectiveness as these two evaluation parameters have been used in various studies to evaluate the neural networks predictive power [22,33,34]. Values of R 2 closer to 1 represented the models with higher predictive power.…”

“…One of the objectives of this study was to decrease the number of variables to a possible extent without compromising the overall accuracy. A study by Afzaal et al [34] concluded that there was no major improvement in deep learning models predictive accuracy by increasing the number of variables from 2 to 4.…”

Precision Irrigation Strategies for Sustainable Water Budgeting of Potato Crop in Prince Edward Island

“…This has raised concerns about groundwater sustainability, its responsible use, and saltwater intrusion in the system. A recent study by Afzaal et al [25] pointed out the lower groundwater levels than expected in the summer season as a consequence of pumping. This situation has created challenges for water resources managers to meet the supplemental irrigation demands of crops for their sustainable production.…”

Hydrologically Informed Machine Learning for Rainfall-Runoff Modelling: Towards Distributed Modelling

“…The state of art DL capabilities have not yet been tested in hydrological modelling and there are only a few DL applications so far (Shen et al, 2018). Successful DL applications in hydrology include rainfall-runoff modelling (Hu et al, 2018;Fan et al, 2020;Xiang et al, 2020), soil moisture modelling (Xiaodong et al, 2016), precipitation forecasting (Kumar et al, 2019), groundwater estimation (Afzaal et al, 2019) and uncertainty estimation (Gude et al, 2020).…”