Deep learning for sea surface temperature reconstruction under cloud occlusion

Sea Surface Temperature (SST) reconstructions from satellite images affected by cloud gaps have been extensively documented in the past three decades. Here we describe several deep learning models to fill the cloud-occluded areas starting from MODIS Aqua nighttime L3 images in the Italian Seas. To tackle this challenge, after testing different models and methodologies, we employed a type of Convolutional Neural Network model (U-Net) to reconstruct cloud-covered portions of satellite imagery while preserving the integrity of observed values in cloud-free areas. We demonstrate the high precision of U-Net with respect to available products done using OI interpolation algorithms. Our results are promising with respect to some earlier studies while suggesting further investigation for more robust intercomparison. • Enhanced SST Data Reconstruction using deep neural networks. • Comparison and tuning of different neural architectures. • Notable improvements over traditional L4 SST reconstruction. • Robust and reliable solution for Climate Reanalysis. • Showcase the potential of intelligent data-driven technology in oceanography.