skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: Temporal Prediction of Coastal Water Quality Based on Environmental Factors with Machine Learning
The accurate forecast of algal blooms can provide helpful information for water resource management. However, the complex relationship between environmental variables and blooms makes the forecast challenging. In this study, we build a pipeline incorporating four commonly used machine learning models, Support Vector Regression (SVR), Random Forest Regression (RFR), Wavelet Analysis (WA)-Back Propagation Neural Network (BPNN) and WA-Long Short-Term Memory (LSTM), to predict chlorophyll-a in coastal waters. Two areas with distinct environmental features, the Neuse River Estuary, NC, USA—where machine learning models are applied for short-term algal bloom forecast at single stations for the first time—and the Scripps Pier, CA, USA, are selected. Applying the pipeline, we can easily switch from the NRE forecast to the Scripps Pier forecast with minimum model tuning. The pipeline successfully predicts the occurrence of algal blooms in both regions, with more robustness using WA-LSTM and WA-BPNN than SVR and RFR. The pipeline allows us to find the best results by trying different numbers of neuron hidden layers. The pipeline is easily adaptable to other coastal areas. Experience with the two study regions demonstrated that enrichment of the dataset by including dominant physical processes is necessary to improve chlorophyll prediction when applying it to other aquatic systems.  more » « less
Award ID(s):
2230046
PAR ID:
10482913
Author(s) / Creator(s):
; ; ; ; ;
Publisher / Repository:
MDPI
Date Published:
Journal Name:
Journal of Marine Science and Engineering
Volume:
11
Issue:
8
ISSN:
2077-1312
Page Range / eLocation ID:
1608
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; (, Proceedings of Tackling Climate Change with Machine Learning Workshop at International Conference on Machine Learning (ICML 2021))
    Accurately forecasting Arctic sea ice from sub- seasonal to seasonal scales has been a major scientific effort with fundamental challenges at play. In addition to physics-based earth system models, researchers have been applying multiple statistical and machine learning models for sea ice forecast- ing. Looking at the potential of data-driven sea ice forecasting, we propose an attention-based Long Short Term Memory (LSTM) ensemble method to predict monthly sea ice extent up to 1 month ahead. Using daily and monthly satellite retrieved sea ice data from NSIDC and atmospheric and oceanic variables from ERA5 reanalysis product for 39 years, we show that our multi-temporal ensemble method outperforms several baseline and recently proposed deep learning models. This will substantially improve our ability in predicting future Arctic sea ice changes, which is fundamental for forecasting transporting routes, resource development, coastal erosion, threats to Arctic coastal communities and wildlife. 
    more » « less
  2. ; ; ; ; ; (, Emerging Science Journal)
    Solar energy is a widely accessible, clean, and sustainable energy source. Solar power harvesting in order to generate electricity on smart grids is essential in light of the present global energy crisis. However, the highly variable nature of solar radiation poses unique challenges for accurately predicting solar photovoltaic (PV) power generation. Factors such as cloud cover, atmospheric conditions, and seasonal variations significantly impact the amount of solar energy available for conversion into electricity. Therefore, it is essential to precisely estimate the output of solar power in order to assess the potential of smart grids. This paper presents a study that utilizes various machine learning models to predict solar photovoltaic (PV) power generation in Lubbock, Texas. Mean Squared Error (MSE) and R² metrics are utilized to demonstrate the performance of each model. The results show that the Random Forest Regression (RFR) and Long Short-Term Memory (LSTM) models outperformed the other models, with a MSE of 2.06% and 2.23% and R² values of 0.977 and 0.975, respectively. In addition, RFR and LSTM demonstrate their capability to capture the intricate patterns and complex relationships inherent in solar power generation data. The developed machine learning models can aid solar PV investors in streamlining their processes and improving their planning for the production of solar energy. Doi: 10.28991/ESJ-2023-07-04-02 Full Text: PDF 
    more » « less
  3. ; ; ; ; ; ; ; (, International Conference on Research in Air Transportation 2020, 9th International Conference)
    Phase of flight (POF) prediction estimates the future state of aircraft along planned trajectories, allowing the prediction of potential conflicts as well as optimization of air space, controlled by the Federal Aviation Administration. In this paper, we present a study conducted to develop three different POF forecasting machine learning models and a statistical regression model using four-dimensional GPS and RADAR Track data from 57 flights provided by an En Route Computer System. The investigated machine learning models include Long Short-Term Memory Recurrent Neural Network (LSTM-RNN), Support Vector Machine (SVM), and Neural Ordinary Differential Equations (NODE). These were developed to forecast the horizontal and vertical POF of the current aircraft for the next time step. The results in this study indicate that LSTM-RNN models are more suitable for POF prediction than SVM and statistical regression models, with NODE being a promising model for future trajectory prediction research. 
    more » « less
  4. ; ; (, International Journal of Climatology)
    Abstract This study illustrates the considerable improvement in accuracy achievable for long‐lead forecasts (18 months) of the Ocean Niño Index (ONI) through the utilization of a long short‐term memory (LSTM) machine learning algorithm. The research assesses the predictive potential of eight predictors from both tropical and extratropical regions constructed based on sea surface temperature, outgoing longwave radiation, sea surface height and zonal and meridional wind anomalies. In comparison to linear regression model forecasts, the LSTM model outperforms them for both the tropical and extratropical predictor sets. Among all the predictors, the western North Pacific (WNP) index demonstrates the highest prediction skill in ONI forecasts, followed by the North Tropical Atlantic (NTA) index and then the sea surface height index. While other predictors help the LSTM model to forecast either the phase variation of the amplitude variation of the observed ONI, the extratropical WNP predictor enables the LSTM model to forecast both variations. This superiority can be attributed to the involvement of SST anomalies in the WNP region in both tropical and extratropical El Niño–Southern Oscillation (ENSO) dynamics, allowing for the utilization of predictive potential from both components of ENSO dynamics. The study also concludes that the extratropical ENSO dynamics provide a robust source of predictability for long‐lead ENSO forecasts, which can be effectively harnessed using the LSTM model. 
    more » « less
  5. ; ; ; ; ; ; ; ; ; ; et al (, Nature)
    Abstract Phytoplankton blooms in coastal oceans can be beneficial to coastal fisheries production and ecosystem function, but can also cause major environmental problems 1,2 —yet detailed characterizations of bloom incidence and distribution are not available worldwide. Here we map daily marine coastal algal blooms between 2003 and 2020 using global satellite observations at 1-km spatial resolution. We found that algal blooms occurred in 126 out of the 153 coastal countries examined. Globally, the spatial extent (+13.2%) and frequency (+59.2%) of blooms increased significantly ( P  < 0.05) over the study period, whereas blooms weakened in tropical and subtropical areas of the Northern Hemisphere. We documented the relationship between the bloom trends and ocean circulation, and identified the stimulatory effects of recent increases in sea surface temperature. Our compilation of daily mapped coastal phytoplankton blooms provides the basis for global assessments of bloom risks and benefits, and for the formulation or evaluation of management or policy actions. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email