Contrails contribute substantially to the overall climate impact of aviation. Mitigation strategies include technological measures, such as reducing soot with sustainable fuels, and operational measures, where flights are rerouted vertically to avoid ice-supersaturated regions (ISSRs). Operational avoidance is limited by uncertainty in weather data, contrail lifetime, and resolution of satellite imagery. The aim of this thesis is twofold: to develop a contrail-toflight attribution algorithm to determine contrail avoidance regions and to evaluate it on a controlled dataset with known ground truth. The method enforces local physical consistency constraints based on two hypotheses: nearby contrails share correlated wind errors and originate from flights at similar altitudes due to the limited vertical extent of ISSRs. Overall, the findings support the use of physical-consistency constraints in contrail-to-flight attribution as a useful complement to, and standalone alternative for, geometric matching.