PNNL

Abstract

Land-atmosphere interactions play an important role in summer rainfall in the central US where mesoscale convective systems (MCSs) contribute to 30-70% of warm season precipitation. Here, using a unique combination of observed MCS and non-MCS rainfall and land surface simulations with numerical tracers to quantify soil moisture sourced from MCS and non-MCS rainfall, we investigate the soil moisture-precipitation feedbacks associated with MCS and non-MCS rainfall events and their hydrologic footprints. Characterized by higher intensity and larger area per storm, earlier-season (April-June) MCS rainfall produces coherent mesoscale spatial heterogeneity in soil moisture important for initiating July afternoon rainfall dominated by non-MCS events. On the other hand, soil moisture sourced from both earlier-season MCS and non-MCS rainfall contribute to lower level atmospheric moistening favorable for upscale growth of MCSs at night. However, soil moisture sourced from MCS rainfall contributes to July MCS rainfall with longer lead time because with higher intensity, MCS rainfall percolates into the deeper soil with longer memory. Therefore, earlier-season MCS rainfall dominates soil moisture-precipitation feedback and potentially provides an important source of predictability for summer rainfall in the major agricultural region of the central US and other continental regions frequented by MCSs.