Cirrus, Transport, and Mixing in the Tropical Upper Troposphere

The impact of cloud radiative heating on transport time scales from the tropical upper troposphere to the stratosphere is studied in two-dimensional numerical simulations. Clouds are idealized as sources of radiative heating and are stochastically distributed in space and time. A spatial probability function constrains clouds to occur in only part of the domain to depict heterogeneously distributed clouds in the atmosphere. The transport time from the lower to upper boundaries (age of air) is evaluated with trajectories. The spectra of age of air obtained in the simulations are bimodal, with the first mode composed of trajectories that remain in the cloudy part of the domain during their passages from the lower to upper boundaries, and the second mode composed of the remaining trajectories that visit the cloud-free regions. For the first group of trajectories only, the mean age scales inversely with the time-mean radiative heating in cloudy air, and the one-dimensional advection-diffusion equation provides an adequate model for transport. However, the exchange between the cloudy and cloud-free regions renders the mean age over all trajectories (including those that visit the cloud-free region) much longer than the time expected if all air parcels remain in cloudy air. In addition, the overall mean age is not inversely proportional to the time-mean heating rate in cloudy air. Sensitivity calculations further show that the sizes, durations, and amplitudes of the individual clouds are also important to the transport time. The results show that the frequently used decomposition of radiative heating into clear-sky and cloud radiative heating may give incorrect interpretations regarding the time scale of transport into the stratosphere.