Wavelet and statistical analyses of river water quality parameters: a case study in the Lower Minnesota River

Abstract

Statistical and wavelet analyses are useful tools for analyzing river water quality parameters. In this study, they were employed to study parameters including biochemical oxygen demand (BOD), dissolved oxygen (DO), nitrate (NO3), ammonium (NH4), phosphate (PO4), total phosphorus (TP), total Kjeldahl nitrogen (TKN), chlorophyll a (CHLA), total suspended solids (TSS) and water temperature (TEMP) monitored at five hydrologic stations on the Lower Minnesota River, USA. Strong positive correlations were observed between CHLA-BOD, TP-TKN, TP-TSS and TKN-TSS, with strong negative correlation between DO-TEMP. Daubechies wavelet at level 5 has been calculated for some key water quality parameters as it gives the finer scale approximation and decomposition of each water parameter. The results show that TEMP and DO have relative quasi-periodicity of about one year, while the quasi-periodicity of NH4 and PO4 are weaker than for TEMP and DO. Correlations between some parameters based on wavelet decomposition results are consistent. The fluctuation range characteristics of some parameters were also analyzed through wavelet decomposition.

Water Quality Index of Suceava River in Suceava City Metropolitan Area

“…The wavelet transform is a mathematical technique useful in signal detection, including signals of water quality parameters, such as the annual or seasonal cycles [40,41]. The wavelets are functions used to represent data according to some repetitive patterns and a finite number of scales/resolutions and offers a perspective of both the entire dataset and its details at the same time.…”

Study of the effectiveness of wavelet genetic programming model for water quality analysis in the Uttar Pradesh region

“…Based on evaluation metrics namely pMAPE, MAPE, EVS, R the model DWT-PCA-LSTM was found to outperform. Data was collected from ve hydrologic stations on the Lower Minnesota River, which is the main tributary of the Mississippi River, USA to study whether the water quality parameters can be analyzed using statistical and wavelet analysis (Gao et al, 2018).…”

Connected Sensors, Innovative Sensor Deployment, and Intelligent Data Analysis for Online Water Quality Monitoring

“…The classic analysis choice in time series modelling remains whether to model in the time or frequency domain. In the time domain, classical time series models of autoregressive or moving average (ARIMA models) have been used but as the temporal resolution of monitoring has increased there has been more and more research using the frequency domain, where wavelets and other transforms have been used [174,175]. Further developments in the modelling of environmental time series has come from the application of functional data analysis (FDA) methods [176].…”