Uncovering the influence of hydrological and climate variables in chlorophyll-A concentration in tropical reservoirs with machine learning

Carvalho, Taís Maria Nunes; Neto, Iran Eduardo Lima; Filho, Francisco de Assis de Souza

doi:10.1007/s11356-022-21168-z

Cited by 17 publications

(15 citation statements)

References 64 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The approach presented in this study achieves this analysis using a highly heterogeneous collection of in-situ observations, investigating the performance of different ML models to determine Chla levels in 149 dams in the Brazilian state of Ceará. Although most dams were predominantly eutrophic throughout the years, different human activities and pollution sources have contributed to eutrophication processes lead to Chla spatiotemporal fluctuations and algal blooms [17,20,144,145].…”

Section: Discussion Of the Resultsmentioning

confidence: 99%

Chlorophyll-a Estimation in 149 Tropical Semi-Arid Reservoirs Using Remote Sensing and Machine Learning

Oliveira Santos,

Guimarães,

Lima Neto

et al. 2024

Preprint

View full text Add to dashboard Cite

Eutrophication, a global concern, impacts water quality, ecosystems, and human health. It’s crucial to monitor algal blooms in freshwater reservoirs, as they indicate the trophic condition of a waterbody through Chlorophyll-a (Chla) concentration. Traditional monitoring methods, however, are expen-sive and time-consuming. Addressing this hindrance, we developed models using remotely sensed data from the Sentinel-2 satellite for large-scale coverage, including its bands and spectral indexes, to estimate the Chla concentration on 149 freshwater reservoirs in Ceará, Brazil. Several machine learning models were trained and tested, including k-nearest neighbours, random forests, extreme gradient boosting, the least absolute shrinkage, group method of data handling (GMDH), and sup-port vector machine models. A stepwise approach determined the best subset of input parameters. Using a 70/30 split for the training and testing datasets, the best-performing model was the GMDH, achieving an R2 of 0.91, MAPE of 102.34%, and RMSE of 20.38 g/L, which are values consistent with the ones found in the literature. Nevertheless, the predicted Chla concentration values were most sensitive to the red, green, and near infra-red bands.

show abstract

Section: Discussion Of the Resultsmentioning

confidence: 99%

Chlorophyll-a Estimation in 149 Tropical Semi-Arid Reservoirs Using Remote Sensing and Machine Learning

Oliveira Santos,

Guimarães,

Lima Neto

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…The approach presented in this study achieved such an analysis by using a highly heterogeneous collection of in situ observations and investigating the performance of different ML models in determining the Chla levels in 149 reservoirs in the Brazilian state of Ceará. Although most reservoirs are predominantly eutrophic throughout the years, various human activities and pollution sources have contributed to the eutrophication processes, leading to Chla spatiotemporal fluctuations and algal blooms [30,34,128,129].…”

Section: Discussionmentioning

confidence: 99%

Chlorophyll-a Estimation in 149 Tropical Semi-Arid Reservoirs Using Remote Sensing Data and Six Machine Learning Methods

Oliveira Santos,

Guimarães,

Neto

et al. 2024

Remote Sensing

Self Cite

View full text Add to dashboard Cite

It is crucial to monitor algal blooms in freshwater reservoirs through an examination of chlorophyll-a (Chla) concentrations, as they indicate the trophic condition of these waterbodies. Traditional monitoring methods, however, are expensive and time-consuming. Addressing this hindrance, we conducted a comprehensive investigation using several machine learning models for Chla modeling. To this end, we used in situ collected water sample data and remote sensing data from the Sentinel-2 satellite, including spectral bands and indices, for large-scale coverage. This approach allowed us to conduct a comprehensive analysis and characterization of the Chla concentrations across 149 freshwater reservoirs in Ceará, a semi-arid region of Brazil. The implemented machine learning models included k-nearest neighbors, random forest, extreme gradient boosting, the least absolute shrinkage, and the group method of data handling (GMDH); in particular, the GMDH approach has not been previously explored in this context. The forward stepwise approach was used to determine the best subset of input parameters. Using a 70/30 split for the training and testing datasets, the best-performing model was the GMDH model, achieving an R2 of 0.91, an MAPE of 102.34%, and an RMSE of 20.4 μg/L, which were values consistent with the ones found in the literature. Nevertheless, the predicted Chla concentration values were most sensitive to the red, green, and near-infrared bands.

show abstract

“…The multilayer perceptron (MLP) stands as a variant of the feedforward neural network architecture, characterized by the inclusion of multiple hidden layers, each replete with a multitude of interconnected neurons [27,28]. Leveraging its capacity for profound nonlinear modeling, the MLP emerges as a potent tool for capturing intricate associations entwining ocean Chl-a concentration and a myriad of input features.…”

Section: Methods For Predicting Chl-a Multilayer Perceptronmentioning

confidence: 99%

“…At every iteration, the nascent decision tree is engineered with the specific objective of rectifying the residuals stemming from the antecedent model iteration. Nevertheless, it is imperative to exercise judicious caution when navigating the domain of hyperparameter tuning during the model training process, owing to GBDT's pronounced sensitivity to hyperparameters [28,29].…”

Section: Gradient Boosted Decision Treesmentioning

confidence: 99%

“…The primary objective is to assess the efficacy of several widely employed machine learning algorithms in the task of reconstructing Chl-a concentration patterns specific to this region. The machine learning algorithms employed in this study include Multilayer Perceptron (MLP) [4,27,28], Random Forest (RF) [5,16,25,29,30], Gradient Boosted Decision Tree (GBDT) [28,29], and K-Nearest Neighbor (KNN) [14,18,23,27,31]. Our model differs from previous studies in that it utilizes easily accessible data on ocean dynamical processes as explanatory variables.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Improvement in Spatiotemporal Chl-a Data in the South China Sea Using the Random-Forest-Based Geo-Imputation Method and Ocean Dynamics Data

Li,

Shao,

Zhang

et al. 2023

JMSE

View full text Add to dashboard Cite

The accurate estimation of the spatial and temporal distribution of chlorophyll-a (Chl-a) concentrations in the South China Sea (SCS) is crucial for understanding marine ecosystem dynamics and water quality assessment. However, the challenge of missing values in satellite-derived Chl-a data has hindered obtaining complete spatiotemporal information. Traditional methods for deriving Chl-a are based on the modeling of measured sensor data and in situ measurements. Spatiotemporal imputation of Chl-a is difficult due to the inaccessibility of the measured Chl-a. In this study, we introduce an innovative approach that incorporates an ocean dynamics dataset and utilizes the random forest algorithm for predicting the Chl-a concentration in the SCS. The method combines the spatiotemporal feature pattern of Chl-a and the main influencing factors, and it introduces ocean dynamics data, which has a high correlation with the spatiotemporal distribution of Chl-a, as the input data through feature engineering. Also, we compared Random Forest (RF) with other Machine Learning (ML) methods. The results show that (1) ocean dynamics datasets can provide important data support for Chl-a imputation by capturing the impact of dynamical processes on ecological roles in the South China Sea. (2) The RF method is the superior imputation method for the reconstruction of Chl-a in the South China Sea, with better model performance and smaller errors. This study provides valuable insight for researchers and practitioners in choosing suitable machine learning methods for the imputation of the Chl-a concentration in the SCS, facilitating a better understanding of the region’s marine ecosystems and supporting effective environmental management.

show abstract

Uncovering the influence of hydrological and climate variables in chlorophyll-A concentration in tropical reservoirs with machine learning

Cited by 17 publications

References 64 publications

Chlorophyll-a Estimation in 149 Tropical Semi-Arid Reservoirs Using Remote Sensing and Machine Learning

Chlorophyll-a Estimation in 149 Tropical Semi-Arid Reservoirs Using Remote Sensing and Machine Learning

Chlorophyll-a Estimation in 149 Tropical Semi-Arid Reservoirs Using Remote Sensing Data and Six Machine Learning Methods

Improvement in Spatiotemporal Chl-a Data in the South China Sea Using the Random-Forest-Based Geo-Imputation Method and Ocean Dynamics Data

Contact Info

Product

Resources

About