1: \begin{abstract}

2:   Motivated by 

3:   systems in which droplets grow and shrink in a turbulence-driven

4:   supersaturation field, we investigate the problem of turbulent

5:   condensation in a general manner. Using direct numerical

6:     simulations we show that the turbulent fluctuations of the

7:   supersaturation field offer different conditions for the growth of

8:   droplets which evolve in time due to turbulent transport and

9:   mixing. Based on that, we propose a Lagrangian stochastic model for

10:   condensation and evaporation of small droplets in turbulent

11:   flows. It consists of a set of stochastic integro-differential

12:   equations for the joint evolution of the squared radius and the

13:   supersaturation along the droplet trajectories. The model has two

14:   parameters fixed by the total amount of water and the thermodynamic

15:   properties, as well as the Lagrangian integral timescale of the

16:   turbulent supersaturation.  The model reproduces very well the

17:   droplet size distributions obtained from direct numerical

18:   simulations and their time evolution. A noticeable result is that,

19:   after a stage where the squared radius simply diffuses, the system

20:   converges exponentially fast to a statistical steady state

21:   independent of the initial conditions.  The main mechanism involved

22:   in this convergence is a loss of memory induced by a significant

23:   number of droplets undergoing a complete evaporation before growing

24:   again.  The statistical steady state is characterised by an

25:   exponential tail in the droplet mass distribution. These results

26:   reconcile those of earlier numerical studies, once these various

27:   regimes are considered.

28: \end{abstract}

29: