Loop Current SSH Forecasting: A New Domain Partitioning Approach for a Machine Learning Model
A divide-and-conquer (DAC) machine learning approach was first proposed by Wang et al. to forecast the sea surface height (SSH) of the Loop Current System (LCS) in the Gulf of Mexico. In this DAC approach, the forecast domain was divided into non-overlapping partitions, each of which had their own p...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-08-01
|
| Series: | Forecasting |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2571-9394/3/3/36 |
| _version_ | 1797519271786446848 |
|---|---|
| author | Justin L. Wang Hanqi Zhuang Laurent Chérubin Ali Muhamed Ali Ali Ibrahim |
| author_facet | Justin L. Wang Hanqi Zhuang Laurent Chérubin Ali Muhamed Ali Ali Ibrahim |
| author_sort | Justin L. Wang |
| collection | DOAJ |
| description | A divide-and-conquer (DAC) machine learning approach was first proposed by Wang et al. to forecast the sea surface height (SSH) of the Loop Current System (LCS) in the Gulf of Mexico. In this DAC approach, the forecast domain was divided into non-overlapping partitions, each of which had their own prediction model. The full domain SSH prediction was recovered by interpolating the SSH across each partition boundaries. Although the original DAC model was able to predict the LCS evolution and eddy shedding more than two months and three months in advance, respectively, growing errors at the partition boundaries negatively affected the model forecasting skills. In the study herein, a new partitioning method, which consists of overlapping partitions is presented. The region of interest is divided into 50%-overlapping partitions. At each prediction step, the SSH value at each point is computed from overlapping partitions, which significantly reduces the occurrence of unrealistic SSH features at partition boundaries. This new approach led to a significant improvement of the overall model performance both in terms of features prediction such as the location of the LC eddy SSH contours but also in terms of event prediction, such as the LC ring separation. We observed an approximate 12% decrease in error over a 10-week prediction, and also show that this method can approximate the location and shedding of eddy Cameron better than the original DAC method. |
| first_indexed | 2024-03-10T07:40:33Z |
| format | Article |
| id | doaj.art-53c20d8f55c64457b54403a4099be453 |
| institution | Directory Open Access Journal |
| issn | 2571-9394 |
| language | English |
| last_indexed | 2024-03-10T07:40:33Z |
| publishDate | 2021-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Forecasting |
| spelling | doaj.art-53c20d8f55c64457b54403a4099be4532023-11-22T13:06:32ZengMDPI AGForecasting2571-93942021-08-013357057910.3390/forecast3030036Loop Current SSH Forecasting: A New Domain Partitioning Approach for a Machine Learning ModelJustin L. Wang0Hanqi Zhuang1Laurent Chérubin2Ali Muhamed Ali3Ali Ibrahim4Department of Computer Science, University of Illinois at Urbana-Champaign, Champaign, IL 61820, USACEECS Department, Florida Atlantic University, Boca Raton, FL 33431, USAHarbor Branch Oceanographic Institute, Florida Atlantic University, Boca Raton, FL 33431, USACEECS Department, Florida Atlantic University, Boca Raton, FL 33431, USACEECS Department, Florida Atlantic University, Boca Raton, FL 33431, USAA divide-and-conquer (DAC) machine learning approach was first proposed by Wang et al. to forecast the sea surface height (SSH) of the Loop Current System (LCS) in the Gulf of Mexico. In this DAC approach, the forecast domain was divided into non-overlapping partitions, each of which had their own prediction model. The full domain SSH prediction was recovered by interpolating the SSH across each partition boundaries. Although the original DAC model was able to predict the LCS evolution and eddy shedding more than two months and three months in advance, respectively, growing errors at the partition boundaries negatively affected the model forecasting skills. In the study herein, a new partitioning method, which consists of overlapping partitions is presented. The region of interest is divided into 50%-overlapping partitions. At each prediction step, the SSH value at each point is computed from overlapping partitions, which significantly reduces the occurrence of unrealistic SSH features at partition boundaries. This new approach led to a significant improvement of the overall model performance both in terms of features prediction such as the location of the LC eddy SSH contours but also in terms of event prediction, such as the LC ring separation. We observed an approximate 12% decrease in error over a 10-week prediction, and also show that this method can approximate the location and shedding of eddy Cameron better than the original DAC method.https://www.mdpi.com/2571-9394/3/3/36Gulf of Mexicoloop currentSSHforecastingeddy sheddingmachine learning |
| spellingShingle | Justin L. Wang Hanqi Zhuang Laurent Chérubin Ali Muhamed Ali Ali Ibrahim Loop Current SSH Forecasting: A New Domain Partitioning Approach for a Machine Learning Model Forecasting Gulf of Mexico loop current SSH forecasting eddy shedding machine learning |
| title | Loop Current SSH Forecasting: A New Domain Partitioning Approach for a Machine Learning Model |
| title_full | Loop Current SSH Forecasting: A New Domain Partitioning Approach for a Machine Learning Model |
| title_fullStr | Loop Current SSH Forecasting: A New Domain Partitioning Approach for a Machine Learning Model |
| title_full_unstemmed | Loop Current SSH Forecasting: A New Domain Partitioning Approach for a Machine Learning Model |
| title_short | Loop Current SSH Forecasting: A New Domain Partitioning Approach for a Machine Learning Model |
| title_sort | loop current ssh forecasting a new domain partitioning approach for a machine learning model |
| topic | Gulf of Mexico loop current SSH forecasting eddy shedding machine learning |
| url | https://www.mdpi.com/2571-9394/3/3/36 |
| work_keys_str_mv | AT justinlwang loopcurrentsshforecastinganewdomainpartitioningapproachforamachinelearningmodel AT hanqizhuang loopcurrentsshforecastinganewdomainpartitioningapproachforamachinelearningmodel AT laurentcherubin loopcurrentsshforecastinganewdomainpartitioningapproachforamachinelearningmodel AT alimuhamedali loopcurrentsshforecastinganewdomainpartitioningapproachforamachinelearningmodel AT aliibrahim loopcurrentsshforecastinganewdomainpartitioningapproachforamachinelearningmodel |