Effects of Vegetation and Infiltration Feedbacks on Hydrologic Partitioning and Droughts

This dissertation addresses feedbacks between vegetation dynamics and land surface response to rainfall events, particularly in Mediterranean climates. Specifically, we ask how a saturated hydraulic conductivity value (ks) that is tied to vegetation biomass affects how water is divided into infiltration and runoff under a range of conditions. First, a field campaign in Sardinia was conducted in which a 4 m by 4 m rainfall simulator was constructed and deployed on a number of dates. Measurements of surface runoff from the plot and soil moisture within the plot informed estimates of the effective ks for each experimental run, and a comparison between ks and vegetation height measurements revealed a monotonically increasing relationship between the two. We then fit a logistic equation to this relationship and incorporated it into the calculations of a coupled vegetation dynamics and land surface model. Using the model, which is calibrated for the Sardinia field site, we investigated the effect of the variable ks by comparing the model results of biomass, saturation, and runoff to results using a static ks. We then used the same model to investigate the effects of a variable ks on drought recovery by simulating drought severity through a range of biomass levels relative to a no-drought condition. Our modeling results revealed that the primary result of a variable ks is modification of the quantity and mechanism of surface runoff; specifically, runoff increased over the constant ks case and shifted from saturation excess runoff to infiltration excess runoff. These effects are more pronounced in drier conditions and when rainfall intensities are in a critical region similar to the ks value. We conclude that a dynamic ks value is relevant for prediction of surface runoff and may improve the performance of land surface models.