Added value of convection-permitting regional models using precipitation event identification and tracking

 

Author

V. Rao Kotamarthi — Argonne National Laboratory

Category

General topics – Clouds

Description

As the spatial resolution of regional scale meteorological simulations becomes increasingly refined, explicit simulation of convection as opposed parameterized convection is becoming feasible. However, to assess the fidelity of these simulations requires newer and more complex metrics and statistical tools. Here we present results from a systematic sensitivity study for summertime over the CONUS, using WRF model. To analyze the results we use a novel rainstorm identification and tracking algorithm that can allocate rainfall to individual events. The results of the analysis show that in these model simulations, the model wet bias is driven by larger than observed spatial extent of the precipitating events in the model. We also find that the model biases have a strong diurnal cycle and this results from the higher spatial growth of long-lived events during the course of the day in the model compared to observations. We show that convection-permitting simulations produce lower model bias in precipitation during an event than from using a parameterized convection model.