Effect of environmental covariable selection in the hydrological modeling using machine learning models to predict daily streamflow.

There are different methods for predicting streamflow, and, recently machine learning has been widely used for this purpose. This technique uses a wide set of covariables in the prediction process that must undergo a selection to increase the precision and stability of the models. Thus, this work aimed to analyze the effect of covariable selection with Recursive Feature Elimination (RFE) and Forward Feature Selection (FFS) in the performance of machine learning models to predict daily streamflow. The study was carried out in the Piranga river basin, located in the State of Minas Gerais, Brazil. The database consisted of an 18-year-old historical series (2000-2017) of streamflow data at the outlet of the basin and the covariables derived from the streamflow of affluent rivers, precipitation, land use and land cover, products from the MODIS sensors, and time. The highly correlated covariables were eliminated and the selection of covariables by the level of importance was carried out by the RFE and FFS methods for the Multivariate Adaptive Regression (EARTH), Multiple Linear Regression (MLR), and Random Forest (RF) models. The data were partitioned into two groups: 75% for training and 25% for validation. The models were run 50 times and had their performance evaluated by the Nash Sutcliffe efficiency coefficient (NSE), Determination coefficient (R2), and Root of Mean Square Error (RMSE). The three models tested showed satisfactory performance with both covariable selection methods, however, all of them proved to be inaccurate for predicting values associated with maximum streamflow events. The use of FFS, in most cases, improved the performance of the models and reduced the number of selected covariables. The use of machine learning to predict daily streamflow proved to be efficient and the use of FFS in the selection of covariables enhanced this efficiency.

Machine learning models for streamflow regionalization in a tropical watershed.

This study aims to assess different machine learning approaches for streamflow regionalization in a tropical watershed, analyzing their advantages and limitations, and to point the benefits of using them for water resources management. The algorithms applied were: Random Forest, Earth and linear model. The response variables were the three types of minimum streamflow (Q7.10, Q95 and Q90), besides the long-term average streamflow (Qmld). The database involved 76 environmental covariates related to morphometry, topography, climate, land use and cover, and surface conditions. The elimination of covariates was performed using two processes: Pearson's correlation analysis and importance analysis by Recursive Feature Elimination (RFE). To validate the models, the following statistical metrics were used: Nash-Sutcliffe coefficient (NSE), percent bias (PBIAS), Willmott's index of agreement (d), coefficient of determination (R2), root mean square error (RMSE), mean absolute error (MAE) and relative error (RE). The linear model was unsatisfactory for all response variables. The results show that nonlinear models performed well, and their covariate of greatest predictive importance was flow equivalent to the precipitated volume, considering the subtraction of an abstraction factor of 750 mm (Peq750). Generally, the Random Forest and Earth models showed similar performances and great ability to predict the minimum streamflow and long-term average streamflow assessed, constituting powerful and promising alternatives for the streamflow regionalization in support to the management and integrated planning of water resources at the level of river basins.

Use of multivariate statistical methods to analyze the monitoring of surface water quality in the Doce River basin, Minas Gerais, Brazil.

The objective of the present study was to evaluate the water quality data in the Minas Gerais portion of the Doce River basin in order to analyze the current monitoring network by identifying the main variables to be maintained in the network, their possible sources of pollution, and the best sampling frequency. Multivariate statistical techniques (factor analysis/principal components analysis, FA/PCA and cluster analysis, CA) complemented by the analysis of violation of the framing classes were used for this purpose. Water quality variables common to 64 monitoring sites were analyzed for the base period from 2010 to 2017. The water quality variables were analyzed considering the different monitoring campaigns: (a) partial campaigns; (b) total campaigns; and (c) monthly campaigns. It was identified from the FA/PCA results, that, when the partial campaign data were analyzed, the variables selected represent the high susceptibility that the basin presents to erosion and the release of domestic effluents in its water bodies. When the data of total campaigns were evaluated, representative variables of the contamination by heavy metals from industrial and mining activities were included. Therefore, the analysis of violation of the framing classes made possible to identify five critical variables: thermotolerant coliforms, dissolved iron, total phosphorus, and total manganese, which reinforced the results obtained in FA/PCA. Based on the results of the analyses, it was recommended to include variables associated with heavy metal contamination in the partial campaigns, prioritizing the dissolved iron and total manganese, as well as total chloride sampling only for the total campaigns. The evaluated data from the monthly campaigns, the CA showed that although the quarterly monitoring frequency is satisfactory, the monthly monitoring is more appropriate for the monitoring of water quality in the Minas Gerais portion of the Doce River basin.