Decomposing Daily Baltic Sea Thermal-Anomaly Evolution into Rearrangement and Residual Change Using Unbalanced Optimal Transport

Abstract

This repository provides the complete analysis code and supporting datasets for the manuscript “Decomposing Daily Baltic Sea Thermal-Anomaly Evolution into Rearrangement and Residual Change Using Unbalanced Optimal Transport.” It includes the reproducible Python analysis workflow and validation tests, daily Baltic Sea depth-mean volumetric heat-content fields derived from Copernicus Marine data for 1993–2025, daily ERA5 surface net-heat-flux data, and processed NetCDF files containing the data underlying the manuscript figures. The workflow uses unbalanced optimal transport to separate daily thermal-anomaly evolution into spatial rearrangement and residual local change. Rearrangement is an advection-like pattern-displacement diagnostic—not an estimate of ocean currents or heat transport—while the residual represents source- and sink-like change rather than a closed physical heat budget.