Prediction Of Harmful Algal Blooms Research Articles

Chlorophyll a in lakes and streams of the United States (2005–2022)

The concentration of chlorophyll a in phytoplankton and periphyton represents the amount of algal biomass. We compiled an 18-year record (2005–2022) of pigment data from water bodies across the United States (US) to support efforts to develop process-based, machine learning, and remote sensing models for prediction of harmful algal blooms (HABs). To our knowledge, this dataset of nearly 84,000 sites and over 1,374,000 pigment measurements is the largest compilation of harmonized discrete, laboratory-extracted chlorophyll data for the US. These data were compiled from the Water Quality Portal (WQP) and previously unpublished U.S. Geological Survey’s National Water Quality Laboratory (NWQL) data. Data were harmonized for reporting units, pigment type, duplicate values, collection depth, site name, negative values, and some extreme values. Across the country, data show great variation by state in sampling frequency, distribution, and methods. Uses for such data include the calibration of models, calibration of field sensors, examination of relationship to nutrients and other drivers, evaluation of temporal trends, and other applications addressing local to national scale concerns.

Predicting coastal harmful algal blooms using integrated data-driven analysis of environmental factors

Coastal harmful algal blooms (HABs) have become one of the challenging environmental problems in the world's thriving coastal cities due to the interference of multiple stressors from human activities and climate change. Past HAB predictions primarily relied on single-source data, overlooked upstream land use, and typically used a single prediction algorithm. To address these limitations, this study aims to develop predictive models to establish the relationship between the HAB indicator – chlorophyll-a (Chl-a) and various environmental stressors, under appropriate lagging predictive scenarios. To achieve this, we first applied the partial autocorrelation function (PACF) to Chl-a to precisely identify two prediction scenarios. We then combined multi-source data and several machine learning algorithms to predict harmful algae, using SHapley Additive exPlanations (SHAP) to extract key features influencing output from the prediction models. Our findings reveal an apparent 1-month autoregressive characteristic in Chl-a, leading us to create two scenarios: 1-month lead prediction and current-month prediction. The Extra Tree Regressor (ETR), with an R2 of 0.92, excelled in 1-month lead predictions, while the Random Forest Regressor (RFR) was most effective for current-month predictions with an R2 of 0.69. Additionally, we identified current month Chl-a, developed land use, total phosphorus, and nitrogen oxides (NOx) as critical features for accurate predictions. Our predictive framework, which can be applied to coastal regions worldwide, provides decision-makers with crucial tools for effectively predicting and mitigating HAB threats in major coastal cities.

Development and Application of an Integrated System for the Detection and Prediction of Harmful Algal Blooms in Korea

Harmful algal blooms (HABs) are types of phytoplankton overgrowth that adversely affect marine ecosystems and aquaculture resources. One such HAB species, Cochlodinium polykrikoides, occurs irregularly and causes significant damage to the aquaculture industry along the coastal regions of Korea. In this study, we developed and implemented an integrated system to detect and predict HAB occurrences in real time. This system comprises four main components: (1) a real-time detection system utilizing acoustic sensing, ocean weather, water temperature, salinity, and chlorophyll, satellite images, genetic analysis, and optics; (2) a prediction model system based on current and tidal, HAB occurrence, and HAB movement and diffusion models; (3) an additional data based on HAB information of sampling data and HAB information of GPS data, and (4) an integrated information system utilizing data storage servers and a visualization platform. We applied and assessed the efficiency of this integrated system in the South Sea of Korea from 2017 to 2019. Particularly, HABs occurred significantly in 2019, and the system demonstrated the feasibility of detection and prediction under field conditions. Implementing a more advanced integrated detection and prediction system in the field is anticipated to minimize the damage caused by irregular HAB occurrences every year.

Development of a collaborative framework for quantitative monitoring and accumulation prediction of harmful algal blooms in nearshore areas of lakes

Harmful algal blooms (HABs) in lakes have been a salient problem of global water environment due to their harmful impacts. Present situation and future tendency of HABs in nearshore areas are more important than those in other areas of lakes, but traditional researches generally ignored this point. This study developed a novel framework aiming to timely and accurately grasp both present conditions and accumulation risks of HABs in nearshore areas of lakes. By designing a quantitative monitoring method, land-based video monitoring devices around lakes were used to obtain present situation of HABs in nearshore areas rapidly, periodically and automatically. On this basis, a coupled simulation model was established to predict accumulation risks of HABs in nearshore areas using five indexes, i.e., shoreline index, chlorophyll-a index, algal bloom index, wind direction index and dynamic index. Experimental results indicated that the accuracy of HAB monitoring in nearshore areas is quite satisfying (85.20%) and the prediction results of HABs accumulation in nearshore areas is reasonable and credible. This study demonstrated a high value of the developed collaborative framework in emergency prevention and control of HABs in lakes.

Ensemble Machine Learning of Gradient Boosting (XGBoost, LightGBM, CatBoost) and Attention-Based CNN-LSTM for Harmful Algal Blooms Forecasting.

Harmful algal blooms (HABs) are a serious threat to ecosystems and human health. The accurate prediction of HABs is crucial for their proactive preparation and management. While mechanism-based numerical modeling, such as the Environmental Fluid Dynamics Code (EFDC), has been widely used in the past, the recent development of machine learning technology with data-based processing capabilities has opened up new possibilities for HABs prediction. In this study, we developed and evaluated two types of machine learning-based models for HABs prediction: Gradient Boosting models (XGBoost, LightGBM, CatBoost) and attention-based CNN-LSTM models. We used Bayesian optimization techniques for hyperparameter tuning, and applied bagging and stacking ensemble techniques to obtain the final prediction results. The final prediction result was derived by applying the optimal hyperparameter and bagging and stacking ensemble techniques, and the applicability of prediction to HABs was evaluated. When predicting HABs with an ensemble technique, it is judged that the overall prediction performance can be improved by complementing the advantages of each model and averaging errors such as overfitting of individual models. Our study highlights the potential of machine learning-based models for HABs prediction and emphasizes the need to incorporate the latest technology into this important field.

Tentative Prediction of Harmful Algal Blooms in Yangtze Estuary

By using spatial autoregressive models, this paper mainly studied the limited factors and spatio-temporal prediction of harmful algal blooms (HAB). The environmental factors of sea surface temperature (SST), dissolved organic matter (DOM) and suspended sediment concentration (SSC) were used to model the chlorophyll a (Cha) distribution and the occurrence of HAB. The results indicated that SST had no significant influence on the growth of phytoplankton in May and June, but higher SST in July and August decreased the frequency of occurrence of HAB. SSC was the main limited factor of HAB, and considerably and positively influenced the algal growth and HAB. HAB in Yangtze estuary was mainly caused by nutrients not DOM. The predicted spatial pattern in July 2010 was more dispersed than the actual clustered pattern. This clustered pattern should be driven by the combined impacts of coastal current, Taiwan Worm current and the northward seasonal.

Sub-monthly prediction of harmful algal blooms based on automated cell imaging

Harmful algal blooms (HABs) are an increasing threat to global fisheries and human health. The mitigation of HABs requires management strategies to successfully forecast the abundance and distribution of harmful algal taxa. In this study, we attempt to characterize the dynamics of 2 phytoplankton genera (Pseudo-nitzschia spp. and Dinophysis spp.) in Narragansett Bay, Rhode Island, using empirical dynamic modeling. We utilize a high-resolution Imaging FlowCytobot dataset to generate a daily-resolution time series of phytoplankton images and then characterize the sub-monthly (1–30 days) timescales of univariate and multivariate prediction skill for each taxon. Our results suggest that univariate predictability is low overall, different for each taxon and does not significantly vary over sub-monthly timescales. For all univariate predictions, models can rely on the inherent autocorrelation within each time series. When we incorporated multivariate data based on quantifiable image features, we found that predictability increased for both taxa and that this increase was apparent on timescales >7 days. Pseudo-nitzschia spp. has distinctive predictive dynamics that occur on timescales of around 16 and 25 days. Similarly, Dinophysis spp. is most predictable on timescales of 25 days. The timescales of prediction for Pseudo-nitzschia spp. and Dinophysis spp. could be tied to environmental drivers such as tidal cycles, water temperature, wind speed, community biomass, salinity, and pH in Narragansett Bay. For most drivers, there were consistent effects between the environmental variables and the phytoplankton taxon. Our analysis displays the potential of utilizing data from automated cell imagers to forecast and monitor harmful algal blooms.

A Review of Machine Learning Models for Harmful Algal Bloom Monitoring in Freshwater Systems

Highlights Machine Learning (ML) models are identified, reviewed, and analyzed for HAB predictions. Data preprocessing is vital for efficient ML model development. ML models for toxin production and monitoring are limited. Abstract. Harmful algal blooms (HABs) are detrimental to livestock, humans, pets, the environment, and the global economy, which calls for a robust approach to their management. While process-based models can inform practitioners about HAB enabling conditions, they have inherent limitations in accurately predicting harmful algal blooms. To address these limitations, Machine Learning (ML) models can potentially leverage large volumes of IoT data to aid in near real-time predictions. ML models have evolved as efficient tools for understanding patterns and relationships between water quality parameters and HAB expansion. This review describes ML models currently used for predicting and forecasting HABs in freshwater ecosystems and presents model structures and their application for predicting algal parameters and related toxins. The review revealed that regression trees, random forest, Artificial Neural Network (ANN), Support Vector Regression (SVR), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) are the most frequently used models for HABs monitoring. This review shows ML models' prowess in identifying significant variables influencing algal growth, HAB drivers, and multistep HAB prediction. Hybrid models also improve the prediction of algal-related parameters through improved optimization techniques and variable selection algorithms. While ML models often focus on algal biomass prediction, few studies apply ML models for toxin monitoring and prediction. This limitation can be associated with a lack of high-frequency toxin datasets for model development, and exploring this domain is encouraged. This review serves as a guide for policymakers and researchers to implement ML models for HAB prediction and reveals the potential of ML models for decision support and early prediction for HAB management. Keywords: Cyanobacteria, Freshwater, Harmful algal blooms, Machine learning, Water quality.

Machine learning-based prediction of harmful algal blooms in water supply reservoirs

Abstract Harmful algal blooms (HABs) pose a potential risk to human and ecosystem health. HAB occurrences are influenced by numerous environmental factors; thus, accurate predictions of HABs and explanations about the predictions are required to implement preventive water quality management. In this study, machine learning (ML) algorithms, i.e., random forest (RF) and extreme gradient boosting (XGB), were employed to predict HABs in eight water supply reservoirs in South Korea. The use of synthetic minority oversampling technique for addressing imbalanced HAB occurrences improved classification performance of the ML algorithms. Although RF and XGB resulted in marginal performance differences, XGB exhibited more stable performance in the presence of data imbalance. Furthermore, a post hoc explanation technique, Shapley additive explanation was employed to estimate relative feature importance. Among the input features, water temperature and concentrations of total nitrogen and total phosphorus appeared important in predicting HAB occurrences. The results suggest that the use of ML algorithms along with explanation methods increase the usefulness of predictive models as a decision-making tool for water quality management.

Integrated explainable deep learning prediction of harmful algal blooms

Harmful algal blooms (HABs) can cause serious problems for aquatic ecosystems and human health, as well as massive social costs. Therefore, continuous monitoring and prevention are required. Water quality management is an important task to minimize such algae, and future occurrences can be accurately predicted through optimal water resource management. In this study, we developed a convolutional neural network model using eight water quality variables and four weather variables to predict the concentration of chlorophyll-a in four major Korean rivers. In addition, Deep SHAP was applied to aid in policy decision-making and identify the influence on variables affecting chlorophyll-a. This integrated prediction model showed a 38.01 % reduction in root mean square error and 36.16 % improvement in R-squared compared to the long short-term memory (LSTM) model. This demonstrated the effectiveness of the proposed integrated prediction approach. Furthermore, despite simultaneously predicting HABs at all monitoring stations and training 394 times faster than LSTM-based models, the proposed method exhibited a significant improvement in efficiency and elucidated variable influences that existing models failed to explain. The proposed integrated prediction model can predict HAB spread, identify variable influences to aid decision-makers, and effectively implement preemptive responses, thus reducing economic losses and preserving aquatic ecosystems.

Fouling control in SWRO desalination during harmful algal blooms: A historical review and future developments

Harmful algal blooms (HABs) are a major barrier to the stable and efficient operation of seawater reverse osmosis (SWRO) desalination plants. Although the intensity varies regionally, the occurrence of HABs impairs plant operation through severe particulate, organic, and biological fouling of both pretreatment and RO systems. As a result, considerable effort has been put into developing and optimizing pre-treatment techniques to mitigate these problems. This study reviews the historical changes and current status of pretreatment applications based on >110 datasets for large SWRO desalination plants. Chronological analysis underscores the increasing trend in the operation of dissolved air flotation and ultrafiltration systems since severe HABs in the Middle East from 2008 to 2009. The impact of HAB characteristics on system performance was also analyzed, and the efficiency of current pretreatment systems for the removal of HABs was evaluated, along with proposals for operational guidelines. Finally, future strategies for i) intelligent monitoring and prediction of HABs, ii) improvement of unit processes for the removal of algal organic matter, and iii) utilization of algal biomass to alleviate environmental impact were systematically delineated for smarter, safer, and greener operation of future SWRO desalination plants in response to the occurrence of HABs.

Improved seasonal prediction of harmful algal blooms in Lake Erie using large-scale climate indices

Harmful Algal Blooms lead to multi-billion-dollar losses in the United States due to shellfish closures, fish mortalities, and reluctance to consume seafood. Therefore, an improved early seasonal prediction of harmful algal blooms severity is important. Conventional methods for harmful algal blooms prediction using nutrient loading as the primary driver have been found to be less accurate during extreme bloom years. Here we show that a machine learning approach using observed nutrient loading, and large-scale climate indices can improve the harmful algal blooms prediction in Lake Erie. Moreover, the seasonal prediction of harmful algal blooms can be completed by early June, before the expected peak in harmful algal bloom activity from July to October. This improved early seasonal prediction can provide timely information to policymakers for adopting proper planning and mitigation strategies such as restrictions in harvesting and help in monitoring toxins in shellfish to keep contaminated products off the market.

Spatial-temporal variability of microphytoplankton assemblages including harmful microalgae in a tropical semi-enclosed strait (Johor Strait, Malaysia)

Harmful algal blooms (HABs) were not new to the tropical semi-enclosed Johor Strait, with incident records that could trace back to the 1980s. HAB monitoring in the area, often, is reactive, focusing only on HAB taxa previously causing problems but neglecting potential emerging HABs. To develop datasets on HABs that can better inform and improve management practices, monitoring should expand to sample whole microphytoplankton communities. In this study, microphytoplankton community structure across the Strait was investigated. Abundances of microphytoplankton and a suite of in situ water parameters of temperatures, salinity, pH, dissolved oxygen levels, macronutrients, and chlorophyll-a contents were collected at ten sites across the Strait at monthly intervals from January 2017 to December 2018. A total of 48 genera (51 taxa) microphytoplankton were identified microscopically. Diatom was the most diverse group (32 genera), followed by dinophyte (15 genera). Bloom-forming species included diatoms Chaetoceros, Coscinodiscus, Eucampia, Pseudo-nitzschia, Rhizosolenia, Skeletonema, Thalassiosira, and dinophytes Blixaea quinquecornis and Scrippsiella. Diatom taxa that exhibit high in situ growth rates were predominant in the low-nutrient marine-influenced environment. Bloom-forming taxa including HAB taxa were found dominant in the environment with high nutrient levels and mesohaline, salinity-stratified conditions. This study provides valuable baseline data that could assist in monitoring and prediction of HABs in the Johor Strait and could be of reference to other similar tropical coastal systems.

Inland harmful algal blooms (HABs) modeling using internet of things (IoT) system and deep learning

Harmful algal blooms (HABs) have been frequently occurred with releasing toxic substances, which typically lead to water quality degradation and health problems for humans and aquatic animals. Hence, accurate quantitative analysis and prediction of HABs should be implemented to detect, monitor, and manage severe algal blooms. However, the traditional monitoring required sufficient expense and labor while numerical models were restricted in terms of their ability to simulate the algae dynamic. To address the challenging issue, this study evaluates the applicability of deep learning to simulate chlorophyll-a (Chl-a) and phycocyanin (PC) with the internet of things (IoT) system. Our research adopted LSTM models for simulating Chl-a and PC. Among LSTM models, the attention LSTM model achieved superior performance by showing 0.84 and 2.35 (μg/L) of the correlation coefficient and root mean square error. Among preprocessing methods, the z-score method was selected as the optimal method to improve model performance. The attention mechanism highlighted the input data from July to October, indicating that this period was the most influential period to model output. Therefore, this study demonstrated that deep learning with IoT system has the potential to detect and quantify cyanobacteria, which can improve the eutrophication management schemes for freshwater reservoirs.

Long-Term Eutrophication and Dynamics of Bloom-Forming Microbial Communities during Summer HAB in Large Arctic Lake

Harmful algal blooms (HABs) in arctic lakes are recent phenomena. In our study, we performed a long-term analysis (1990–2017) of the eutrophication of Lake Imandra, a large subarctic lake, and explored the biodiversity of bloom-forming microorganisms of a 2017 summer HAB. We performed a 16Sr rRNA metabarcoding study of microbial communities, analysed the associations between N, P, C, and chlorophyll concentrations in the lake water, and developed models for the prediction of HABs based on total P concentration. We have demonstrated that blooms in Lake Imandra occur outside of optimal Redfield ratios and have a nonlinear association with P concentrations. We found that recent summer HABs in a lake occur as simultaneous blooms of a diatom Aulacoseira sp. and cyanobacteria Dolichospermum sp. We have studied the temporal dynamics of microbial communities during the bloom and performed an analysis of the publicly available Dolichospermum genomes to outline potential genetic mechanisms beneath simultaneous blooming. We found genetic traits requisite for diatom-diazotroph associations, which may lay beneath the simultaneous blooming of Aulacoseira sp. and Dolichospermum sp. in Lake Imandra. Both groups of organisms have the ability to store nutrients and form a dormant stage. All of these factors will ensure the further development of the HABs in Lake Imandra and the dispersal of these bloom-forming species to neighboring lakes.

LSTM Networks to Improve the Prediction of Harmful Algal Blooms in the West Coast of Sabah.

Harmful algal bloom (HAB) events have alarmed authorities of human health that have caused severe illness and fatalities, death of marine organisms, and massive fish killings. This work aimed to perform the long short-term memory (LSTM) method and convolution neural network (CNN) method to predict the HAB events in the West Coast of Sabah. The results showed that this method could be used to predict satellite time series data in which previous studies only used vector data. This paper also could identify and predict whether there is HAB occurrence in the region. A chlorophyll a concentration (Chl-a; mg/L) variable was used as an HAB indicator, where the data were obtained from MODIS and GEBCO bathymetry. The eight-day dataset interval was from January 2003 to December 2018. The results obtained showed that the LSTM model outperformed the CNN model in terms of accuracy using RMSE and the correlation coefficient r as the statistical criteria.

Prediction of harmful algal blooms in large water bodies using the combined EFDC and LSTM models

Harmful algal blooms (HABs) is a worldwide water environmental problem. HABs usually happens in short time and is difficult to be controlled. Early warning of HABs using data-driven models is prospective in making time for taking precaution against HABs. High-frequency water quality monitoring data are necessary to improve the reliability of the model, but it is expensive. This research used environmental fluid dynamics code (EFDC) to extend one-point data obtained by only one instrument to the whole 249 ha water area instead of multi-instruments monitoring, followed by Long short-term memory (LSTM) to predict the HABs in the whole water body. Correlation analysis and principal component analysis were used to reduce the data dimension and improve model accuracy. Finally, the LSTM model was calibrated to predict chlorophyll-a (Chl-a) for the next 1 to 3 time steps. The Nash-Sutcliffe efficiency coefficient (NSE) and mean absolute percentage error (MAPE) of EFDC-LSTM were 0.797–0.991 and 2.74–13.16%, respectively, suggesting the promising utilization of this model in early warning systems for HABs. EFDC-LSTM achieves high-precision HABs forecasting in a cost-effective manner, providing a reliable way to detect HABs in advance.

Quantifying uncertainty cascading from climate, watershed, and lake models in harmful algal bloom predictions

In response to increased harmful algal blooms (HABs), hypoxia, and nearshore algae growth in Lake Erie, the United States and Canada agreed to phosphorus load reduction targets. While the load targets were guided by an ensemble of models, none of them considered the effects of climate change. Some watershed models developed to guide load reduction strategies have simulated climate effects, but without extending the resulting loads or their uncertainties to HAB projections. In this study, we integrated an ensemble of four climate models, three watershed models, and four HAB models. Nutrient loads and HAB predictions were generated for historical (1985–1999), current (2002–2017), and mid-21st-century (2051–2065) periods. For the current and historical periods, modeled loads and HABs are comparable to observations but exhibit less interannual variability. Our results show that climate impacts on watershed processes are likely to lead to reductions in future loading, assuming land use and watershed management practices are unchanged. This reduction in load should help reduce the magnitude of future HABs, although increases in lake temperature could mitigate that decrease. Using Monte-Carlo analysis to attribute sources of uncertainty from this cascade of models, we show that the uncertainty associated with each model is significant, and that improvements in all three are needed to build confidence in future projections.

Unsteady Multi-Element Time Series Analysis and Prediction Based on Spatial-Temporal Attention and Error Forecast Fusion

Harmful algal blooms (HABs) often cause great harm to fishery production and the safety of human lives. Therefore, the detection and prediction of HABs has become an important issue. Machine learning has been increasingly used to predict HABs at home and abroad. However, few of them can capture the sudden change of Chl-a in advance and handle the long-term dependencies appropriately. In order to address these challenges, the Long Short-Term Memory (LSTM) based spatial-temporal attentions model for Chlorophyll-a (Chl-a) concentration prediction is proposed, a model which can capture the correlation between various factors and Chl-a adaptively and catch dynamic temporal information from previous time intervals for making predictions. The model can also capture the stage of Chl-a when values soar as red tide breaks out in advance. Due to the instability of the current Chl-a concentration prediction model, the model is also applied to make a prediction about the forecast reliability, to have a basic understanding of the range and fluctuation of model errors and provide a reference to describe the range of marine disasters. The data used in the experiment is retrieved from Fujian Marine Forecasts Station from 2009 to 2011 and is combined into 8-dimension data. Results show that the proposed approach performs better than other Chl-a prediction algorithms (such as Attention LSTM and Seq2seq and back propagation). The result of error prediction also reveals that the error forecast method possesses established advantages for red tides prevention and control.

Comparison of Spatial and Temporal Genetic Differentiation in a Harmful Dinoflagellate Species Emphasizes Impact of Local Processes

Population genetic studies provide insights into intraspecific diversity and dispersal patterns of microorganisms such as protists, which help understanding invasions, harmful algal bloom development and occurrence of seafood poisoning. Genetic differentiation across geography has been reported in many microbial species indicating significant dispersal barriers among different habitats. Temporal differentiation has been less studied and its frequency, drivers and magnitude are poorly understood due to a lack of integral studies. The toxic dinoflagellate species Gambierdiscus caribaeus was sampled during two years in the Florida Keys, and repeatedly from 2006 to 2016 at St. Thomas, US Virgin Islands (USVI), including a three-year period with monthly sampling, to compare spatial and temporal genetic differentiation. Samples from the USVI site showed high temporal variability in local population structure, which correlated with changes in salinity and benthic habitat cover. In some cases, temporal variability exceeded spatial differentiation, despite apparent lack of connectivity and dispersal across the Greater Caribbean Region based on the spatial genetic data. Thus, local processes such as selection might have a stronger influence on population structure in microorganisms than geographic distance. The observed high temporal genetic diversity challenges the prediction of harmful algal blooms and toxin concentrations, but illustrates also the evolutionary potential of microalgae to respond to environmental change.