Dataset for "A Continuous Implicit Neural Representation Framework with Gradient Regularization for Sea Surface Height Reconstruction From Satellite Altimetry"

This repository archives the raw datasets used in the study: “A Continuous Implicit Neural Representation Framework with Gradient Regularization for Sea Surface Height Reconstruction From Satellite Altimetry.” The datasets analysed in this study are provided by the Ocean Data Challenges initiative maintained by AVISO. This record provides a persistent copy of the original data files used in the numerical experiments, in accordance with the GMD data policy. 1. Gulf Stream region (GF) experiment — 2020 Domain: [65°W–55°W, 33°N–43°N] This experiment uses simulated sea surface height (SSH) fields from the NATL60 nature run within the Ocean Data Challenges framework. dc_obsContains along-track sampling data derived from NATL60 corresponding to multiple altimeter configurations:Topex–Poseidon (interleaved), nadir SWOT, KaRIn SWOT, Jason-1, Geosat-2, and Envisat. dc_refContains the reference SSH field from the NATL60 nature run:NATL60-CJM165_GULFSTREAM_*.1h_SSH.ncThis file serves as the ground truth for reconstruction and evaluation. 2. Western Mediterranean Sea (MEDIT) experiment — 2023 Domain: [1°E–20°E, 30°N–45°N] dc_obsContains along-track data from five nadir altimeters (alg, h2b, j3, s3a, s3b) and the corresponding gridded MDT map.These data are used as inputs for SSH reconstruction. dc_evalContains along-track data from the c2n nadir altimeter.This dataset is used exclusively for evaluation and must not be used in the reconstruction stage, following the Ocean Data Challenges protocol. All files are provided in their original NetCDF format as distributed by the Ocean Data Challenges initiative. For full dataset documentation and licensing information, users should refer to the official AVISO/Ocean Data Challenges resources.