Field Observations Of Salinity Research Articles

Novel Salinity Modeling Using Deep Learning for the Sacramento–San Joaquin Delta of California

Water resources management in estuarine environments for water supply and environmental protection typically requires estimates of salinity for various flow and operational conditions. This study develops and applies two novel deep learning (DL) models, a residual long short-term memory (Res-LSTM) network, and a residual gated recurrent unit (Res-GRU) model, in estimating the spatial and temporal variations of salinity. Four other machine learning (ML) models, previously developed and reported, consisting of multi-layer perceptron (MLP), residual network (ResNet), LSTM, and GRU are utilized as the baseline models to benchmark the performance of the two novel models. All six models are applied at 23 study locations in the Sacramento–San Joaquin Delta (Delta), the hub of California’s water supply system. Model input features include observed or calculated tidal stage (water level), flow and salinity at model upstream boundaries, salinity control gate operations, crop consumptive use, and pumping for the period of 2001–2019. Meanwhile, field observations of salinity at the study locations during the same period are also utilized for the development of the predictive use of the models. Results indicate that the proposed DL models generally outperform the baseline models in simulating and predicting salinity on both daily and hourly scales at the study locations. The absolute bias is generally less than 5%. The correlation coefficients and Nash–Sutcliffe efficiency values are close to 1. Particularly, Res-LSTM has slightly superior performance over Res-GRU. Moreover, the study investigates the overfitting issues of both the DL and baseline models. The investigation indicates that overfitting is not notable. Finally, the study compares the performance of Res-LSTM against that of an operational process-based salinity model. It is shown Res-LSTM outperforms the process-based model consistently across all study locations. Overall, the study demonstrates the feasibility of DL-based models in supplementing the existing operational models in providing accurate and real-time estimates of salinity to inform water management decision making.

Hydrodynamic Modeling of Hurricane Dennis Impact on Estuarine Salinity Variation in Apalachicola Bay

ABSTRACT Huang, W.; Hagen, S., and Bacopoulos, P., 2014. Hydrodynamic modeling of Hurricane Dennis impact on estuarine salinity variation in Apalachicola Bay. Hurricane Dennis made landfall in Florida on 10 July 2005 and caused a storm surge of 2.4 m in Apalachicola. Field observations of salinity, winds, and river inflows are available at a few stations in the bay during the hurricane event. Presented in this paper is a numerical modeling study that investigates the effects of Hurricane Dennis on estuarine mixing and transport. A previously calibrated three-dimensional (3D) estuarine hydrodynamic model was further validated by the field observations of salinity and water levels during the hurricane event. A large-scale storm-surge model was used to provide storm-surge hydrographs at the five estuarine boundaries for the 3D estuarine model. This model integration proved successful with the results indicating that the model predictions of storm-surge hydrographs and salinity match well with observations. M...

Estimation of flushing time in a monsoonal estuary using observational and numerical approaches

Estimation of flushing time (T F) in an estuary is important for water quality analysis, and it is one of the major transport time scales used in estuaries to quantify the hydrodynamic processes and for water resources management strategies. Thus, the main objective of the present study was to estimate the T F in the Mandovi estuary (a monsoonal estuary) on the west coast of India. In the study, field observations of salinity (FOS) and a numerical model are used to describe the T F during the south-west monsoon (heavy river discharge), post-monsoon (moderate river discharge) and pre-monsoon (negligible river discharge). T F was calculated for 12 (months) river discharge scenarios (the calculation was done under monthly mean flow conditions). Among the 12 scenarios, four scenarios were within an extremely low flow range: 0.8–2.2 m3s−1; four were in a moderate flow range: 4.3–67 m3s−1; and the remaining four were in a high flow range: 119.0–506.6 m3s−1. The T F calculated from FOS and numerical model showed good matches during the periods of heavy river discharge and moderate river discharge. The results from FOS (numerical model) showed that the T F was 1.12 (1.11) days for a river flow of 506.6 m3s−1 (during the south-west monsoon), and the T F reached 251.28 (592.38) days under extreme low flow of 0.8 m3s−1 (during the pre-monsoon). Regression equations fitted by power and exponential functions were derived from FOS and numerical model simulations to correlate T F with monthly mean river discharges. The power regression equation derived from FOS (numerical model) showed good statistical fit with data (r = −0.997 (−1.0)) for any given river discharge compared to the exponential regression equation (r = −0.81 (−0.80)). Hence, the power regression equation can be used for the rapid evaluation of the impact of drought (during the pre-monsoon) and flood (during the SW monsoon) scenarios in the estuary.