Total Suspended Solids (TSS) and chlorophyll-a concentration are two critical parameters to monitor water quality. Since directly collecting samples for laboratory analysis can be expensive, this paper presents a methodology to estimate this information through remote sensing and Machine Learning (ML) techniques. TSS and chlorophyll-a are optically active components, therefore enabling measurement by remote sensing. Two study cases in distinct water bodies are performed, and those cases use different spatial resolution data from Sentinel-2 spectral images and unmanned aerial vehicles together with laboratory analysis data. In consonance with the methodology, supervised ML algorithms are trained to predict the concentration of TSS and chlorophyll-a. The predictions are evaluated separately in both study areas, where both TSS and chlorophyll-a models achieved R-squared values above 0.8.

Additional measures of in situ water quality monitoring in natural environments can be obtained through remote sensing because certain elements in water modify its spectral behavior. One of the indicators of water quality is the presence of algae, and the aim of this study was to propose an alternative method for the quantification of chlorophyll in water by correlating spectral data, infrared images, and limnology data. The object of study was an artificial lake located at Unisinos University, São Leopoldo/RS, Brazil. The area has been mapped with a modified NGB (near infrared (N), green (G) and blue (B)) camera coupled to an unmanned aerial vehicle (UAV). From the orthorectified and georeferenced images, a modified normalized difference vegetation index (NDVImod) image has been generated. Additionally, 20 sampling points have been established on the lake. At these points, in situ spectral analysis with a spectroradiometer has been performed, and water samples have been collected for laboratory determination of chlorophyll concentrations. The correlation resulted in two models. The first model, based on the multivariate analysis of spectral data, and the second model, based on polynomial equations from NDVI, had coefficients of determination (R2) of 0.86 and 0.51, respectively. This study confirmed the applicability of remote sensing for water resource management using UAVs, which can be characterized as a quick and easy methodology.

The concentration of suspended solids in water is one of the quality parameters that can be recovered using remote sensing data. This paper investigates the data obtained using a sensor coupled to an unmanned aerial vehicle (UAV) in order to estimate the concentration of suspended solids in a lake in southern Brazil based on the relation of spectral images and limnological data. The water samples underwent laboratory analysis to determine the concentration of total suspended solids (TSS). The images obtained using the UAV were orthorectified and georeferenced so that the values referring to the near, green, and blue infrared channels were collected at each sampling point to relate with the laboratory data. The prediction of the TSS concentration was performed using regression analysis and artificial neural networks. The obtained results were important for two main reasons. First, although regression methods have been used in remote sensing applications, they may not be adequate to capture the linear and/or non-linear relationships of interest. Second, results show that the integration of UAV in the mapping of water bodies together with the application of neural networks in the data analysis is a promising approach to predict TSS as well as their temporal and spatial variations.

Water quality monitoring through remote sensing with UAVs is best conducted using multispectral sensors; however, these sensors are expensive. We aimed to predict multispectral bands from a low-cost sensor (R, G, B bands) using artificial neural networks (ANN). We studied a lake located on the campus of Unisinos University, Brazil, using a low-cost sensor mounted on a UAV. Simultaneously, we collected water samples during the UAV flight to determine total suspended solids (TSS) and dissolved organic matter (DOM). We correlated the three bands predicted with TSS and DOM. The results show that the ANN validation process predicted the three bands of the multispectral sensor using the three bands of the low-cost sensor with a low average error of 19%. The correlations with TSS and DOM resulted in R² values of greater than 0.60, consistent with literature values.

In view of the complexity of the processes associated to water quality, this study objective is to identify the main factors related to spatial and seasonal variability in water courses of the Lobo Stream River Basin. To this, multivariate statistical techniques were used. Data collection for water quality and streamflow variables were carried out monthly, from May 2018 to April 2019, at 10 monitoring points along basin's tributaries. The results show that, during the dry season, the main causes for water quality decrease are related to erosion process on the river margins, which is intensified by inadequate handling in livestock activities in some monitoring points. In the rainy season, the main causes are related to soil leaching in agricultural areas that increases the nitrogen compounds concentration and reduces water quality. However, in addition to this, it was noted that regardless the environmental conditions, the most impactful factor is the point pollution from the effluent discharge of Itirapina City sewage treatment plant, responsible for nutrient concentration increase, organic contamination, OD reduction, and, consequently, water quality deterioration. With this, the study shows how multivariate statistical analysis enables more relevant evaluation of water quality data variability and supports further studies in the basin.

Rock porosity is one of the most significant parameters in fluid-flow simulation in the context of carbonate reservoirs. The hydrocarbon industry uses porosity to assess the production potential of oil and gas in carbonate environments. Traditional methods to determine porosity are limited to discrete measurements and generally demand extra resources associated with careful analysis of logs, rock sampling and laboratory analysis. This paper investigates an alternative to estimate porosity in carbonate rocks using pointwise hyperspectral data and machine learning. The method is contiguous, does not require rock sampling and was validated in various rock plug samples collected from two distinct carbonate outcrops. The samples were analysed in the laboratory to determine ground-truth values for the effective porosity and reflectance in visible and infrared regions of the spectra. The supervised regression methods applied were able to estimate a robust relationship between the effective porosity of carbonate rocks and spectral behaviour in characteristic spectral features of carbonate, hydroxyl, molecular water and Fe/Mn. The results obtained here suggest the soundness of the indirect approach to estimate porosity with most of the models trained achieving a coefficient of determination above 0.8 and mean absolute deviation of less than 2%.