A major focus of GAPP is improving our ability to predict hydrologic quantities such as evaporation, precipitation, and runoff at monthly, seasonal, and interannual timescales, and at spatial scales from continental to that of a local watershed. Persistence of anomalies in land-surface and atmospheric variables over such timescales is recognized as a possible aid to improving this predictability. Feedbacks between the land-surface and the atmosphere, such as local recycling of precipitation, are thought to enhance persistence in both surface and atmospheric variables, and a better understanding of how, when, where, and at what scales these feedback mechanisms operate is necessary for achieving desired predictability goals. In particular, mesoscale atmospheric circulations forced by subgrid-scale (to a GCM) heterogeneity in the underlying landscape may play a potentially significant role in how precipitation recycling operates in a given region. The interactions between heterogeneity in soil moisture, the induced mesoscale circulations, and the corresponding impact of the resulting clouds and precipitation on the surface energy and water balance, introduce the potential for complex, nonlinear feedbacks between the surface and the atmosphere. These feedbacks may provide a mechanism for lengthening the timescale of mesoscale land-atmosphere interaction, thus perhaps significantly influencing large-scale precipitation and soil moisture over time. How this mechanism operates in the real climate system, and how the effects would vary between regions that are climatologically very different, such as the continental U.S. and Amazonia, is not known. The objectives of the project proposed here are: (i) to investigate seasonal-scale persistence of anomalies in soil moisture and related variables, globally, though with a focus on the GAPP U.S. study area and the LBA Amazonian study area, using a sophisticated atmosphere/land/hydrology model; (ii) to uncover persistence mechanisms resulting from specific physical-dynamical processes and feedbacks in the coupled land-atmosphere system; (iii) to investigate the impact of subgrid-scale landscape heterogeneity in the GAPP and LBA study areas on large-scale precipitation recycling, the evolution of the soil moisture distribution over time, and large-scale persistence. This research is expected to address key aspects of the GAPP objectives, related to improving extended-range hydrologic predictability, as identified in the GAPP Science Plan.