Clouds in the tropical western Pacific are dominated by reflective deep convective cores with gradually thinning trailing anvil clouds, which play a crucial role in determining tropical cloud radiative effects. The microphysical controls of the thinning process and its changes when subjected to a future warmer climate are still very uncertain.We use the high resolution version of the Exascale Earth System Model (E3SM) to study the thinning process in present day and future climate. We apply Lagrangian forward trajectories starting at peaks of convective activity of the detected mesoscale convective systems (MCS) and track detrained ice crystals to better understand the processes controlling the transition of a thick detrained anvil cloud into a thin cirrus cloud. The trajectories are computed offline from the 1-hourly model-calculated velocity fields and ice crystal sedimentation velocities. The modeled MCS in present day climate have a comparable time evolution to those observed by Himawari geostationary satellite data, with an average lifetime of about 10-15 hours, which accounts only for the optically thick part of the cold cloud shield. However, E3SM fails to simulate the strongest storms, which is reflected by an underestimation of albedo and overestimation of outgoing longwave radiation. The analysis of ice sources (detrainment, vapor deposition, new nucleation events) and sinks (sublimation, aggregation, sedimentation) along trajectories highlights the crucial role of the balance between depositional growth and precipitation formation for the maintenance of aging anvil clouds. Interestingly, deposition of ice on detrained ice crystals contributes to the majority of the upper tropospheric ice mass. On the other hand, about 80% of the initial cloud mass is removed in the form of precipitation within the first 10 hours of the detrainment event, changing its radiative effect from net negative to net positive. The future climate simulation shows an increase in storm frequency and intensity, an increase in ice water content and albedo in convective cores and thick anvils. The trajectory calculations reveal a 15% decrease in cloud lifetime due to climate change, suggesting a decrease in thin high clouds due to a higher precipitation efficiency and a shift to more negative net cloud radiative effects.