Water Level Data Research Articles

Research on Water Level Changes in Lake Mead

Lake Mead, the largest reservoir in the U.S., assumes a pivotal role in the water supply. However, in recent years, the reservoir's water volume has gradually decreased, and many states have cut back on water supply. Therefore, it is crucial to explore the pattern of water level changes in Lake Mead and thereby find ways to alleviate water shortage. We first discussed the pattern of development of historical water levels in Lake Mead. We classified the annual minimum and maximum water level data of Lake Mead using hierarchical clustering, and determined the group of clusters using the Elbow Method to perform the analysis, obtaining the best result. We then classified the historical water levels into early drought periods, non-drought periods, and recent drought periods based on the clustering results. Two models were also built to predict future Lake Mead water level changes. To facilitate the overall analysis, we drew scatterplots of the maximum and minimum water levels concerning time and found that the data of the first model is more consistent and regular, so we predicted the data using Linear Fitting-based time series analysis. However, the data needed since 2005 in model 2 showed a very obvious segmentation in roughly 2011, as seen from the scatterplot. We built a breakpoint regression model to make the prediction.

A Hydrological Data Prediction Model Based on LSTM with Attention Mechanism

With the rapid development of IoT, big data and artificial intelligence, the research and application of data-driven hydrological models are increasing. However, when conducting time series analysis, many prediction models are often directly based on the following assumptions: hydrologic time series are normal, homogeneous, smooth and non-trending, which are not always all true. To address the related issues, a solution for short-term hydrological forecasting is proposed. Firstly, a feature test is conducted to verify whether the hydrological time series are normal, homogeneous, smooth and non-trending; secondly, a sequence-to-sequence (seq2seq)-based short-term water level prediction model (LSTM-seq2seq) is proposed to improve the accuracy of hydrological prediction. The model uses a long short-term memory neural network (LSTM) as an encoding layer to encode the historical flow sequence into a context vector, and another LSTM as a decoding layer to decode the context vector in order to predict the target runoff, by superimposing on the attention mechanism, aiming at improving the prediction accuracy. Using the experimental data regarding the water level of the Chu River, the model is compared to other models based on the analysis of normality, smoothness, homogeneity and trending of different water level data. The results show that the prediction accuracy of the proposed model is greater than that of the data set without these characteristics for the data set with normality, smoothness, homogeneity and trend. Flow data at Runcheng, Wuzhi, Baima Temple, Longmen Town, Dongwan, Lu’s and Tongguan are used as input data sets to train and evaluate the model. Metrics RMSE and NSE are used to evaluate the prediction accuracy and convergence speed of the model. The results show that the prediction accuracy of LSTM-seq2seq and LSTM-BP models is higher than other models. Furthermore, the convergence process of the LSTM-seq2seq model is the fastest among the compared models.

Investigation of Data-Driven Rating Curve (DDRC) Approach

Flooding is a recurring natural disaster worldwide; developing countries are particularly affected due to poor mitigation and management strategies. Often discharge is used to inform the flood forecast. The discharge is usually inferred from the water level via the rating curve because the latter is relatively easy to measure compared to the former. This research focuses on Cambodia, where data scarcity is prevalent, as in many developing countries. Thus, the rating curve has not been updated, making it difficult to effectively evaluate the performance of the global streamflow services, such as the Global Flood Awareness System (GloFAS) and Streamflow Prediction Tool (SPT), whose longer lead time can benefit the country in taking early action. In this study, we used time series of water level and discharge data to understand the changes in the flood plain to generate a data-derived rating curve for fifteen stations in Cambodia. We deployed several statistical and data-driven techniques to derive a generalized, scalable, and region-agnostic method. We further validated the process by applying it to ten stations in the US and found similar performance. In Cambodia, we obtained an average Kling Gupta Efficiency (KGE) of ∼99% & an average Relative Root Mean Squared Error (RRMSE) of 12% with an average Mean Absolute Error (MAE) of 200 m3/s. In the US, overall KGE was 97%, with an average RRMSE of 17% and an average MAE of 32 m3/s. The results indicated that the distribution of the dataset was key in deriving a good rating curve and that the stations with a low flow stations generally had higher errors than the high flow stations. The time series approach was shown to have more probability in capturing the high-end and low-end events compared to traditional method, where usually fewer data points are used. The study demonstrates that time series of data has valuable information to update the rating curve, especially in a data-scarce country.

Tidal and non-tidal sea level analysis in enclosed and inland basins: The Black, Aegean, Marmara, and Eastern Mediterranean (Levantine) Seas

Long-term water level data with 15-minute temporal resolution from 18 tide gauge stations were used to analyze tidal and non-tidal oscillations in four different seas along the Turkish coasts. High-resolution spectral analysis revealed that the semi-diurnal peaks (M2, N2, S2, K2) were higher than diurnal (O1, P1, S1, K1) in the enclosed basins: the Black, Aegean, and Levantine Seas. In the inland basin (the Sea of Marmara), however, both diurnal and mixed but mainly semi-diurnal tides were observed. Tidal ranges are the smallest in the Sea of Marmara as the basin is not large enough to generate its tide. In addition to this, the two narrow and shallow straits connecting it to the larger water bodies isolate this sea from the tidal forcing of two larger water bodies. The second smaller diurnal and semi-diurnal tidal amplitudes were observed in the Black Sea. Both it is an enclosed water body and meteorologically and hydrologically (river inflow and precipitation) the most active water body among all, the tide is masked by this these water level oscillations. Generally, there was an increasing trend in the average water level from fall to early summer corresponding to the hydrologically active period. The annual solar radiational harmonics (SA), which were in the range of 5 to 10 cm, were observed in the whole considered seas while semiannual (SSA,) harmonics, which were in the ranges smaller than 5 cm, were only observed in the Black and the Levantine Seas. The tide is minimum (∼O(10 cm)) in the Sea of Marmara and maximum in the Levantine Sea ((∼O(40–50 cm)). The contribution of the tidal harmonics to the water level energy is around one-half-fold of non-tidal ones in the easternmost part of the Levantine Sea. The largest water level ranges around 100 to 150 cm, however, observed over short periods (day to a week) due to severe meteorological conditions.

Towards an efficient streamflow forecasting method for event-scales in Ca River basin, Vietnam

Study regionThe Ca River basin is located in the North Central Coast area of Vietnam Study focusThis study aims to develop a deep learning framework that is both effective and straightforward in order to forecast water levels in the Ca River basin in advance of multiple time steps for the event scales. We have thoroughly studied and assessed two deep learning models (DLMs), long-short term memory (LSTM) and gated recurrent unit (GRU), for their capacity to forecast water levels, focusing on various aspects such as the influence of sequence length or the impact of hyperparameter selection. Besides, two data scenarios were established using hydrological data from eight severe floods between 2007 and 2019 to examine the effect of input variables on model performance. Water level data was employed for both the scenarios (S1 and S2), whereas precipitation data was used only in S2. The cross-validation technique was used dynamically to address the issue of limited data. The inputs were reformatted as tensors and were then randomly divided into subsets. This flexible tuning preserved the sequential nature of the hydrological data while enabling the DLMs to be trained efficiently. New hydrological insights for the regionThe findings revealed that both the models exhibited equally excellent performances. The NSE of the LSTM model varies from 0.999–0.971 compared to 0.998–0.974 of the GRU model, corresponding to forecast cases from one to four-time steps ahead. This indicated that the use of multiple-input data types (S2) contrary to only one date type (S1) does not necessarily improve the forecasting performance. LSTM/GRU models with one hidden layer are adequate for delivering high performance while minimizing the data processing time.

One Ranney Well Can Make a Difference: The Impacts of a Radial Collector Well on Groundwater Level and Quality in the Cedar River Alluvial Aquifer

AbstractThe City of Cedar Rapids, Iowa, depends on groundwater from the Cedar River alluvial aquifer for residential and industrial use. In 2020, the city completed an additional radial collector well, or Ranney well, and was concerned that pumping from the well at high rates may lower water level elevations in the aquifer, reduce yields from nearby production wells, and change the quality of produced water. During an operational test of the well's pumps, the U.S. Geological Survey and the city collected water level and water quality data to evaluate the effects of increased pumping rates on the aquifer and nearby production wells. Results indicated that a high rate of pumping from the new well caused sustained declines in water levels near the well and other nearby production wells, and, if maintained, the aquifer water level in the vicinity would continue to decline to levels observed during the 2012 drought. Aquifer specific conductance and temperature were altered and matched trends and values of the river, and river‐to‐well travel time was shortened from 7‐17 days to about 3 days. Results may also provide insights to other municipal water resource managers when considering wellfield design, production expectations, and long‐term management strategies for radial collector well production during drought, low streamflow, and times when high concentrations of nitrate‐N or organic pesticide compounds in the river may limit production options.

Synthesis of the ICESat/ICESat-2 and CryoSat-2 observations to reconstruct time series of lake level

ABSTRACT Synthesis of multi-satellite altimetry facilitates the acquisition of long-term changes in lake level but may induce biases due to inconsistent data sources, and thus remains largely unexplored. This study investigates the integrated application of Ice, Cloud, and land Elevation Satellite (ICESat), ICESat-2, and CryoSat-2 missions to provide consecutive lake level series in 2003–2020. The sample comprises 48 lakes in the world with gauge-based or Hydroweb water level data. The CryoSat-2 data are first adjusted to the ICESat-2 height reference based on the mean values of their monthly water level difference during their overlapping period (2018–2020). Then, the corrected CryoSat-2 data are used to link the ICESat and ICESat-2 data. Results show that in the sample lakes, the deviations between CryoSat-2 and ICESat-2 data vary and range from −0.78 m to 0.13 m. The data quality of the synthesized time series is evaluated by comparing against the validation data, with an average R 2 of 0.84. This study shows that CryoSat-2 has the potential of filling the gap between ICESat and ICESat-2. The three altimeters can be expected to integrate effectively for monitoring lake water level changes in the past two decades.

Trophic status observations for Honghu Lake in China from 2000 to 2021 using Landsat Satellites

Eutrophication of lakes harms aquatic organisms, changes the function of lakes and causes water pollution, which has affected global public health. As an internationally important wetland reserve, Honghu Lake is polluted to varying degrees and the eutrophication trend has accelerated. The objective of this study is to quarterly monitor and assess the trophic status of Honghu Lake from 2000 to 2021 using the trophic level index (TLI). The TLI was retrieved by developing a semi-empirical model based on radial basis function neural network (RBFNN), air temperature and water level data were added to the model as input parameters, forming a control group with the model that did not include air temperature and water level data. Based on 178 Landsat images from 2000 to 2021, the eutrophication status of the Honghu Lake over the last 20 years was successfully determined. The main findings are as follows: 1) The accuracy of TLI retrieval model, with air temperature and water level data included (R2 = 0.723, RMSE = 4.971), is significantly higher than that without air temperature and water level data (R2 = 0.197, RMSE = 10.453); 2) The eutrophication level of the waters in the northwestern part of Honghu Lake was higher than those observed in other lake water; 3) There were significant seasonal fluctuations in the eutrophication of Honghu Lake. The highest TLI values were observed in summer and autumn, while the lowest values were observed in winter and spring; 4) The eutrophication trend of the Honghu Lake has significantly accelerated from 2013 onwards, showing a change in eutrophication status, from light eutrophic to mid-eutrophic. Our results indicate that the TLI retrieval method with air temperature and water level is more predictive. Moreover, the eutrophication status for the entire lake from 2000 to 2021 was determined, which may contribute to the protection of the water environment.

Paddy Field Water Movement Through Soil Profiles Under Different Water Management Practices: A HYDRUS 1D Model Study

Physical measurement of hydrological processes through soil profile is very complicated and timeconsuming. Complex and coupled physical processes like water movement with soil matric potential in puddled paddy field can be simulated using physical process-based model HYDRUS 1D. The model simulation was setup for the multilayered (different soil materials at 0-15 and 15-30 cm depth) paddy fields having continuous flooded irrigation (CFI) practice and water saving Alternate Wetting and Drying (AWD) practice. Measured soil physical properties of three Bangladesh Rice Research Institute (BRRI) regional station farms (Kushtia, Sirajganj, and Rangpur) were used as model input, initial and boundary conditions configuration. The model was calibrated and validated using the water data of a dry season field experiment in Kushtia. The calibrated (RMSE of 0.54 cm, d of 0.94, NSE of 0.89) water level data validated successfully with observed water level data of AWD practiced paddy field (d of 0.95, NSE of 0.92). Soil water content reached the threshold/critical level in AWD practice (-101 cm of water soil matric potential at 15 cm soil depth) earlier in light textured soil (loam or sandy loam) compared to heavy textured soil (clay). The physical properties of the layered soils (i.e., soil particle size distribution and soil water release curve, SWRC) did not affect much on water movement in CFI practice, but it had substantial impact on field water movement under AWD practice. The change in soil water storage followed the general trend for respective soil water holding and releasing capacity, clay soil was heavier and released water slowly than that of loam or sandy loam soils. The positive water flux above 15 cm of soil profile mainly drove the water flow due to evapotranspiration and soil water and pressure distribution along the soil profile while the negative fluxes below 15 cm of soil depth due to infiltration or percolation contributed as a secondary force. A basic understanding of HYDRUS simulated results would lead to realize the total physiohydrological environment in the paddy field. Bangladesh Rice J. 25 (2) : 57-67, 2021

Landslide Displacement Prediction Based on Multivariate LSTM Model.

There are many frequent landslide areas in China, which badly affect local people. Since the 1980s, there have been more than 200 landslides in China with a death toll of 30 or more people at a time, economic losses of more than CNY 10 million or significant social impact. Therefore, the study of landslide displacement prediction is very important. The traditional ARIMA and LSTM models are commonly used for forecasting time series data. In our study, a multivariable LSTM landslide displacement prediction model is proposed based on the traditional LSTM model, which integrates rainfall and reservoir water level data. Taking the Baijiabao landslide in the Three Gorges Reservoir area as an example, the data of displacement, rainfall and reservoir water level of monitoring point ZG323 from November 2006 to December 2012 were selected for this study. Our results show that the displacement prediction results of the multivariable LSTM model are more accurate than those of the ARIMA and the univariate LSTM models, and the mean square, root mean square and mean absolute errors are the smallest, which are 0.64223, 0.8014 and 0.50453 mm, respectively. Therefore, the multivariable LSTM model method has higher accuracy and better application prospects in the displacement prediction of the Baijiabao landslide, which can provide a certain reference for the displacement prediction of the same type of landslide.

A hybrid deep learning model based on feature capture of water level influencing factors and prediction error correction for water level prediction of cascade hydropower stations under multiple time scales

The operating conditions of large cascade hydropower stations are complex. Improving the water level prediction accuracy of large cascade hydropower stations is significant for flood control, shipping, irrigation, etc. A new hybrid model based on a refined deep residual shrinkage network and an optimized gated recurrent unit – long term memory network (GRU-LSTM) model is proposed for water level prediction at different time scales. First, to make the input data in the form of a feature map so the model can capture the weight characteristics of each influencing factor when the water level changes, the water level, flow rate, and hydroelectric power plant output data are constructed as high-dimensional feature inputs. To reduce the prediction error caused by downstream tributary backwater jacking, downstream tributary flow is added to the input data. At the same time, a semisoft threshold function with an adjustable function is inserted into the model to improve feature recognition accuracy, reduce deviation, and eliminate the noise of the original hydrological data. Second, an error weight correction function is used to adjust the error between the predicted water level and the observed water level so that the model can automatically adjust the weights of each influencing factor according to the evaluation results of water level prediction. In this process, the parameters of LSTM are optimized using the Archimedes optimization algorithm, and the output results of the LSTM and GRU networks are weighted to obtain a more accurate water level prediction. The proposed model achieves good accuracy and efficiency in downstream water level prediction of the Xiangjiaba Hydropower Station, outperforms existing hydrodynamic and artificial intelligence methods, and has robust scalability in flood forecasting and urban rainwater forecasting.

Combined Well Multi-Parameter Logs and Low-Flow Purging Data for Soil Permeability Assessment and Related Effects on Groundwater Sampling

Cost-effective remediation is increasingly dependent on high-resolution site characterization (HRSC), which is supposed to be necessary prior to interventions. This paper aims to evaluate the use of low-flow purging and sampling water level data in estimating the horizontal hydraulic conductivity of soils. In a new quali-quantitative view, this procedure can provide much more information and knowledge about the site, reducing time and costs. In case of high heterogeneity along the well screen, the whole procedure, as well as the estimation method, could be less effective and rigorous, with related issues in the purging time. The result showed significant permeability weighted sampling, which could provide different results as the pump position changes along the well screen. The proposed study confirms this phenomenon with field data, demonstrating that the use of multiparameter well logs might be helpful in detecting the behaviour of low-permeability layers and their effects on purging and sampling. A lower correlation between low-flow permeability estimations and LeFranc test results was associated with high heterogeneity along the screen, with a longer purging time. In wells P43, MW08 and MW36, due to the presence of clay layers, results obtained differ for almost one order of magnitude and the purging time increases (by more than 16 min). However, with some precautions prior to the field work, the low-flow purging and sampling procedure could become more representative in a shorter time and provide important hydrogeological parameters such as hydraulic conductivity with many tests and high-resolution related results.

Hydrologic‐based modelling of burn depth potentials in degraded peat soils

AbstractPeatland drainage may degrade system resilience to high intensity, soil‐consuming fires. Peat soil fires are unique in that they can smoulder vertically through the soil column, with a multitude of consequences including large carbon emissions, altered hydrology, and dramatic shifts in vegetation communities. In this work we developed and verified a new method to model peat burn depths with readily available water level and peat hydraulic property data at the Great Dismal Swamp National Wildlife Refuge (VA and NC, USA). To model peat burn depths across 11 sites in the Great Dismal Swamp National Wildlife Refuge we combined water table time series data and soil moisture release curves, developed at multiple depths, with moisture‐to‐ignition thresholds. A subset of the results from this empirical modelling approach of peat burn depth severity were compared against those made using a mechanistic model of soil moisture, HYDRUS 1‐D. By comparing modelled burn depth potentials between these two approaches for a range of peats, we confirmed that our simpler, water table‐based approach had similar performance to HYDRUS 1‐D in drained and degraded peats, like those found in the Great Dismal Swamp National Wildlife Refuge. A comparative analysis of modelled burn depths across our study site found that water table position and peat water holding capacity were the key governing controls on burn depth potential. Our findings suggest that drainage weakens both short‐ and long‐term controls on peat burn depths by reducing soil moisture and by decreasing peat water holding capacity. This new approach offers land managers with an additional tool for assessing risk while offering insight into the drivers of peatland wildfire severity.

Numerical modeling of debris transport due to tsunami flow in a coastal urban area

Many numerical models have been developed to estimate the behavior of debris generated by tsunamis. However, such numerical models have been validated through comparison with experimental data under simple conditions such as simple topography. Therefore, model applicability to actual topographic conditions such as ports and urban areas remains largely unknown. In this study, the modeling of contact with the seafloor and collision with buildings in an existing debris transport model (STOC-DM) is improved to enhance actual practical applicability. Moreover, debris numerical simulations are conducted using the results of hydrodynamic numerical simulations with different water level data obtained under the same wave generation conditions as used in the experiments, and the numerical results are compared with results obtained from debris experiments conducted in a previous study using a coastal urban model. The model improvements contribute to reproducing the statistical characteristics of debris position and the final arrival position; however, the modeled results on debris velocity cannot reproduce the experimental results. Therefore, sensitivity analysis is also conducted by applying disturbances not only to the fluid but also to the parameters of the debris model, e.g., drag coefficient, inertia coefficient, and friction coefficient. The combination results considering both disturbances improve the distribution of the debris velocity that cannot be reproduced using fluid disturbance alone. • Numerical modeling of contact with the seafloor and collision with the buildings was improved to enhance actual practical applicability. • The model improvements contributed to improving the statistical characteristics of debris position and the final arrival position. • However, the modeled results on debris velocity could not reproduce the experimental results. • Sensitivity analysis was conducted by applying disturbances not only to the fluid but also to the parameters of the debris model, e.g., drag coefficient, inertia coefficient, and friction coefficient. • The combination results considering both disturbances improved the distribution of the debris velocity.

Study of climate change patterns of the Kashkadarya River basin based on GIS technologies

The article deals with the problem of studying changes in some climatic indicators (temperature, precipitation) in the Kashkadarya River basin. For this purpose, the analysis of land surface temperature and area of snow and ice cover on Landsat 8 images was used. Comparison of the obtained results with the results of analysis of archival data of Uzhydromet and water level data of gauging stations showed that the growth rate of average annual temperatures in Kashkadarya province is 0.21oC, and the rate of reduction of snow-ice cover area is 1% per year. in general, which has decreased over the past 15 years, by 75%. Such trends may lead to a significant reduction in the river flow of the Kashkadarya River in the coming years.

Novel Methods for Imputing Missing Values in Water Level Monitoring Data

Hydrological data are collected automatically from remote water level monitoring stations and then transmitted to the national water management centre via telemetry system. However, the data received at the centre can be incomplete or anomalous due to some issues with the instruments such as power and sensor failures. Usually, the detected anomalies or missing data are just simply eliminated from the data, which could lead to inaccurate analysis or even false alarms. Therefore, it is very helpful to identify missing values and correct them as accurate as possible. In this paper, we introduced a new approach - Full Subsequence Matching (FSM), for imputing missing values in telemetry water level data. The FSM firstly identifies a sequence of missing values and replaces them with some constant values to create a dummy complete sequence. Then, searching for the most similar subsequence from the historical data. Finally, the identified subsequence will be adapted to fit the missing part based on their similarity. The imputation accuracy of the FSM was evaluated with telemetry water level data and compared to some well-established methods - Interpolation, k-NN, MissForest, and also a leading deep learning method - the Long Short-Term Memory (LSTM) technique. Experimental results show that the FSM technique can produce more precise imputations, particularly for those with strong periodic patterns.

Assimilation of SAR-derived flood extent observations for improving fluvial flood forecast – A proof-of-concept

As the severity and occurrence of flood events tend to intensify with climate change, the need for flood forecasting capability increases. In this regard, the Flood Detection, Alert and rapid Mapping (FloodDAM) project, funded by Space for Climate Observatory initiatives, was set out to develop pre-operational tools dedicated to enabling quick responses in flood-prone areas, and to improve the reactivity of decision support systems. This work focuses on the assimilation of 2D flood extent data (expressed in terms of wet surface ratios) and in-situ water level data to improve the representation of the flood plain dynamics with a Telemac-2D model and an Ensemble Kalman Filter (EnKF). The EnKF control vector was composed friction coefficients and corrective parameter to the input forcing. It is then augmented with the water level state averaged over several floodplain zones. This work was conducted in the context of Observing System Simulation Experiments (OSSE) based on a real flood event occurred in January-February 2021 on the Garonne Marmandaise catchment. This allows to validate the observation operator associated to the wet surface ratio observations as well as the dual state-parameter sequential correction implemented in this work. The merits of assimilating SAR-derived flood plain data complementary to in-situ water level observations are shown in the control parameter and observation spaces with 1D and 2D assessment metrics. It was also shown that the correction of the hydraulic state significantly improved the flood dynamics, especially during the recession. This proof-of-concept study paves the way towards near-real-time flood forecast, making the most of remote sensing-derived flood observations.

Enterococcus exceedances related to environmental variability at New Jersey ocean beaches.

Microbial pollution at ocean beaches is a global public health problem that can be exacerbated by excessive rainfall, particularly at beaches adjacent to urban areas. Rain is acknowledged as a predictive factor of Enterococcus levels at NJ beaches, but to date no study has explicitly examined the link. Here, five beaches (156 observations) in Monmouth County, NJ, with storm drain outflows present were sampled for Enterococcus and water quality during dry and wet periods. Hypotheses included (1) beaches differ in Enterococcus levels, (2) Enterococcus is present year-round, and (3) Enterococcus exceedances could be modeled based on environmental parameters. Beaches showed significantly different median Enterococcus levels, with site SEA2 (Neptune Blvd. in Deal, NJ) lower than others and site SEA4 (South Bath Ave. in Long Branch, NJ) higher than the other sites. Elevated Enterococcus levels were detected at water temperatures from 6.5 to 22.2°C. Multiple linear regression models identified rainfall (+), water temperature (+), and water level (-) as related to Enterococcus concentrations levels at these beaches. For the purpose of simulating the efficacy of different monitoring strategies, a hindcast model of Enterococcus abundance based on historic rainfall, water temperature, and water level data was produced. Results indicated that once-per-week sampling detected ~14% (e.g., 1/7) exceedance events, while sampling during summer alone detected ~ 50% of annual exceedance events. Models of Enterococcus exceedance based on readily available environmental time series have the potential to supplement and improve Enterococcus monitoring at NJ beaches.

Using machine learning architecture to optimize and model the treatment process for saline water level analysis

AbstractWater is a vital resource that makes it possible for human life forms to exist. The need for freshwater consumption has significantly increased in recent years. Seawater treatment facilities are less dependable and efficient. Deep learning systems have the potential to increase the efficiency as well as the accuracy of salt particle analysis in saltwater, which will benefit water treatment plant performance. This research proposed a novel method for optimization and modelling of the treatment process for saline water based on water level data analysis using machine learning (ML) techniques. Here, the optimization and modelling are carried out using molecular separation-based reverse osmosis Bayesian optimization. Then the modelled water saline particle analysis has been carried out using back propagation with Kernelized support swarm machine. Experimental analysis is carried out based on water salinity data in terms of accuracy, precision, recall, and specificity, computational cost, and Kappa coefficient. The proposed technique attained an accuracy of 92%, precision of 83%, recall of 78%, specificity of 81%, computational cost of 59%, and Kappa coefficient of 78%.

APPLICATION OF MIKE FLOOD MODEL FOR INUNDATION SIMULATION IN YEN BAI CITY

Flood is a natural disaster that can cause loss of life and damage to property. The objective of this study is to create a flood map of Yen Bai city, where flood occurs regularly during heavy rain and Thao River's water level rises. The study used Mike Flood model based on the connection of 1D and 2D hydrodynamic models. The model parameters were calibrated and validated againts observed water level data of two major flood events in August 2008 and July 2009 which were measured by Bao Ha and Yen Bai hydrological stations, in addition with historical flood survey data in 2008. Flood map was developed corresponding to the alarm levels. The results show that the most affected areas by flood are 4 communes of Au Lau, Hop Minh, Tuy Loc, and Hong Ha wards.