化学
浊度
多元统计
蓄水池
环境化学
吸收(声学)
吸收光谱法
谱线
每个符号的零件数
分析化学(期刊)
浊积岩
多元分析
水质
水污染
色谱法
地表水
作者
Ilza Santos,Patrícia Nunes,Maria Clara Baldow,Emanuelle Gianini,Daniel Merhy,Fábio Roland,Maria Jose Bell,Virgílio Anjos
标识
DOI:10.1016/j.saa.2026.128221
摘要
This study evaluates ultraviolet–visible (UV–Vis) spectroscopy as a quantitative approach for turbidity detection in clear freshwater systems, where conventional methods often show limited sensitivity and reduced accuracy at low particle concentrations, particularly in reservoirs used for drinking water supply, where maintaining low turbidity is essential. Water samples were collected in 26 surface and bottom sampling sites, totaling 52 samples, along a clear water drinking water supply reservoir. The samples were analyzed using UV–Vis spectrophotometry (200–900 nm). Principal Component Analysis (PCA) was applied to identify spectral patterns, and Partial Least Squares (PLS) regression was used to develop predictive models for turbidity based on absorbance spectra. The first two PCA components explained 91% of the total variance, and clearly distinguished surface and bottom samples, indicating consistent spectral differences associated with vertical gradients. The PLS model showed a strong correlation (0.71 R 2 ) between predicted and measured turbidity, demonstrating reliable performance under low turbidity conditions. These results indicate that UV–Vis spectroscopy enhances sensitivity in clear waters and enables the detection of subtle variations in suspended particles, overcoming limitations of conventional methods. Given the contrasting results reported in the literature regarding the application of spectroscopy to water samples, this study contributes to improving the use of this technique. Also, by integrating spectral analysis with multivariate calibration, this study refines turbidity assessment in freshwater systems and advances the validation of in situ measurements, providing a non-invasive approach for water quality monitoring and sediment dynamics evaluation.
科研通智能强力驱动
Strongly Powered by AbleSci AI