Ocean Satellite Data Fusion for High-Resolution Surface Current Maps

Alisa Kugusheva,Hannah Bull,Evangelos Moschos,Artemis Ioannou,Briac Le Vu,Alexandre Stegner

DOI: https://doi.org/10.3390/rs16071182

IF: 5

2024-03-29

Remote Sensing

Abstract:Real-time reconstruction of ocean surface currents is a challenge due to the complex, non-linear dynamics of the ocean, the small number of in situ measurements, and the spatio-temporal heterogeneity of satellite altimetry observations. To address this challenge, we introduce HIRES-CURRENTS-Net, an operational real-time convolutional neural network (CNN) model for daily ocean current reconstruction. This study focuses on the Mediterranean Sea, a region where operational models have great difficulty predicting surface currents. Notably, our model showcases higher accuracy compared to commonly used alternative methods. HIRES-CURRENTS-Net integrates high-resolution measurements from the infrared or visible spectrum—high resolution Sea Surface Temperature (SST) or chlorophyll (CHL) images—in addition to the low-resolution Sea Surface Height (SSH) maps derived from satellite altimeters. In the first stage, we apply a transfer learning method which uses a high-resolution numerical model to pre-train our CNN model on simulated SSH and SST data with synthetic clouds. The observation of System Simulation Experiments (OSSEs) offers us a sufficient training dataset with reference surface currents at very high resolution, and a model trained on this data can then be applied to real data. In the second stage, to enhance the real-time operational performance of our model over previous methods, we fine-tune the CNN model on real satellite data using a novel pseudo-labeling strategy. We validate HIRES-CURRENTS-Net on real data from drifters and demonstrate that our data-driven approach proves effective for real-time sea surface current reconstruction with potential operational applications such as ship routing.

environmental sciences,imaging science & photographic technology,remote sensing,geosciences, multidisciplinary

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the real - time reconstruction of high - resolution ocean surface current fields, especially for the Mediterranean region. Specifically, the main challenges faced by the authors include: 1. **Complex non - linear ocean dynamics**: Ocean surface currents are affected by complex and non - linear physical processes, making their prediction and reconstruction difficult. 2. **Limited in - situ measurement data**: Direct in - situ measurements (such as buoys, moored stations, etc.) are very sparse and cannot provide sufficient spatio - temporal coverage to construct real - time current field maps. 3. **Spatio - temporal heterogeneity of satellite observations**: Sea surface height (SSH) data observed by satellite altimeters are unevenly distributed in space and time, making it difficult to capture fine - scale structures. To solve these problems, the authors introduced a convolutional neural network (CNN) model named HIRES - CURRENTS - Net, which can combine multiple types of satellite data (such as high - resolution sea surface temperature (SST), chlorophyll concentration images, and low - resolution SSH) to achieve daily high - resolution ocean surface current field reconstruction. Through this method, they hope to improve the prediction accuracy of ocean surface current fields and provide support for practical applications (such as ship route optimization). ### Main innovation points - **Multi - modal data fusion**: Integrates high - resolution SST and chlorophyll images and low - resolution SSH data, and uses these data from different sources to enhance the performance of the model. - **Transfer learning and OSSE training**: Use high - resolution numerical models to pre - train the CNN model and further train it with high - quality reference data generated by system simulation experiments (OSSEs). - **Pseudo - label strategy**: Generate pseudo - labels through high - quality satellite data with a time delay, which are used to fine - tune the model, thereby reducing the domain gap between simulated data and real data. - **Cloud - handling ability**: The model can maintain good performance on SST images containing clouds, which is very important in practical operations because clouds will block part of the observed data. Through these methods, the authors have shown that their model has higher accuracy and robustness in reconstructing ocean surface current fields than existing methods.

Ocean Satellite Data Fusion for High-Resolution Surface Current Maps

A Deep Learning Approach to Spatiotemporal Sea Surface Height Interpolation and Estimation of Deep Currents in Geostrophic Ocean Turbulence

Reconstruction of subsurface ocean state variables using Convolutional Neural Networks with combined satellite and in situ data

Estimating Ocean Surface Currents With Machine Learning

Super-Resolving Ocean Dynamics from Space with Computer Vision Algorithms

Mapping high-resolution surface current by incorporating geostrophic equilibrium with surface quasigeostrophic theory using multi-source satellite observations

High-Resolution Seamless Daily Sea Surface Temperature Based on Satellite Data Fusion and Machine Learning over Kuroshio Extension

GEST: Accurate global ocean surface current reconstruction withmulti-scale dynamics-informed neural network

Towards Daily High-resolution Inundation Observations using Deep Learning and EO

Spatial-temporal Offshore Current Field Forecasting Using Residual-learning Based Purely CNN Methodology with Attention Mechanism

Nearshore Bathymetry from ICESat-2 LiDAR and Sentinel-2 Imagery Datasets Using Physics-Informed CNN

Inversion of Sea Surface Currents From Satellite‐Derived SST‐SSH Synergies With 4DVarNets

CHLNET: A novel hybrid 1D CNN-SVR algorithm for estimating ocean surface chlorophyll-a

A Deep Framework for Eddy Detection and Tracking From Satellite Sea Surface Height Data

A Novel Approach for Estimating Sea Surface Currents From Numerical Models and Satellite Images: Validation and Application

Estimating daily semantic segmentation maps of classified ocean eddies using sea level anomaly data from along-track altimetry

Spatiotemporal Prediction of Monthly Coastal Upwelling Scenario in SST Fields Using Deep Learning Based Models

3D reconstruction of ocean velocity from high-frequency radar and acoustic Doppler current profiler: a model-based assessment study

An information flow-based sea surface height reconstruction through machine learning

Radar Echo Reconstruction in Oceanic Area via Deep Learning of Satellite Data

Deep Learning for Mesoscale Eddy Detection With Feature Fusion of Multisatellite Observations