Probing the precipitation life cycle by iterative rain cell tracking

Monitoring the life cycle of convective rain cells requires a Lagrangian viewpoint where the observer moves with the dominant background flow. To adopt such a moving reference frame, we design, validate, and apply a simple rain cell tracking methodwhich we term iterative rain cell tracking (IRT)for spatio-temporal precipitation data. IRT iteratively identifies the formation and dissipation of rain cells and determines the large-scale flow. The iteration is repeated until reaching convergence. As validated using reanalysis wind speeds, repeated iterations lead to substantially increased agreement of the background flow field and an increased number of complete tracks. Our method is thereby able to monitor the growth and intensity profiles of rain cells and is applied to a high-resolution (5 min and 1x1 km(2)) data set of radar-derived rainfall intensities over Germany. We then combine this data set with surface temperature observations and synoptic observations to group tracks according to convective and stratiform conditions. Convective tracks show clear life cycles in intensity, with peaks shifted off-center toward the beginning of the track, whereas stratiform tracks have comparatively featureless intensity profiles. Our results show that the convective life cycle can lead to convection-dominating precipitation extremes at short time scales, while track-mean intensities may vary much less between the two types. The observed features become more pronounced as surface temperature increases, and in the case of convection even exceeded the rates expected from the Clausius-Clapeyron relation.