Persistent contrails simulated in 0D models may experience premature evaporation compared to equivalent simulations in 2D models

Observing contrail cirrus from satellites is challenging because of their low optical depth and large coverage. There is hence widespread reliance on contrail models to consider the periods in which contrails are too optically thin to be observable remotely. The lack of observational data for aged contrails means that contrail model validation relies on intercomparison, which is largely still missing in the literature. To address this, we compare CoCiP, the most used contrail model, to APCEMM, a 2D contrail model, under parametrized meteorological conditions. We show that APCEMM contrails persist for much longer than the equivalent CoCiP contrails, that the lifetime optical depth (a proxy for climate impact) in APCEMM is higher than that in CoCiP, and that the sensitivity of the lifetime optical depth to the relative humidity of the ice supersaturated layer is opposite between the models. These observations are explained by considering the contrail evolution in each model: CoCiP only simulates the fallstreak (the period in which the precipitation plume of the contrail has not reached the subsaturated layer), whereas APCEMM simulations exhibit behavior beyond this. Since post-fallstreak behavior has been seen in large eddy simulations and accounts for ~90 % of the APCEMM lifetime optical depth, our findings have significant implications for both global climate predictions and optimized contrail avoidance. Consequently, we call for the development of contrail models using new methodologies and for the increased collection of in situ observational data for aged contrails to robustly validate late lifetime behavior of contrail models.