This project addresses the coupled eco-hydrologic dynamics of semi-arid Mediterranean watersheds, where the surface drinking-water reservoirs depend almost exclusively on overland flow in the rainy season and where the vegetation growth is water limited. The motivation centers on the observed and projected decreases in winter rainfall, stemming from shifts in large-scale atmospheric circulation patterns. Presently, the ability to predict future hydrologic behavior of these systems is limited by gaps in understanding of how changes in the vegetation cover interact with precipitation dynamics to control the hydrological response. The proposed research will test a set of five focused hypotheses in a Sardinian study basin that is already instrumented and broadly representative of semi-arid Mediterranean watersheds. This project builds on experimental and modeling studies that the PI and Italian collaborator have been conducting for the past several years on Sardinia, the second largest island in the Mediterranean. This project would add overland flow experiments to capture the effect of variable grass cover and rainfall intensity on infiltration and runoff generation. Analysis will be conducted over three overlapping time periods, representing: 1) the intensive field experiment record (~ 5+ years), 2) the satellite record (~ 20+ years), and 3) the streamflow and precipitation record (~ 80 years). Analysis will include possible tradeoffs between infiltration and evapotranspiration. Intellectual Merit: This project is designed to reveal how change and variability in seasonal precipitation patterns excite vegetation dynamics and how these changes in turn impact the hydrological budget of semi-arid watersheds. The tools developed will enable rigorous prediction of the response of eco-hydrological systems to climate forcing. The connection of the results to the large-scale circulation indices is expected to lead to improved predictability at the seasonal time scale. Furthermore, the broad prevalence of Mediterranean climates (e.g. California) simplifies the transferability of the approach and findings. Broader Impacts: The results of the proposed activity will provide direct support for planning water resources needs in the face of systematic changes in Mediterranean precipitation regimes. The undergraduate and Ph.D. students trained on this project will experience broad and meaningful interdisciplinary research experience, including collaborative periods abroad. Preliminary efforts were funded by Duke start-up support to the PI. Duke will provide funding (Pratt Fellowship) for summer support of the undergraduate student researcher.
Significant attention has been paid to how future water resources are likely to respond to changes in climate, but typically these efforts have used empirical analysis to relate past annual precipitation amounts to annual discharge with stationary (non changing) statistics. However, the composition of the landscape and, consequently, its hydrological response are not static. The vegetation cover and its hydrological function depend on time-varying land use and climate controls. Also, different components of the vegetation may be excited and influenced by changes in climate that occur over different time scale. Hence, strictly lumped-empirical analyses over past land cover conditions can be poor predictors for future hydrologic response. In this project we studied and developed an understanding of how changes in rainfall excite changes in vegetation cover, and then how this vegetation in turn controls how much of subsequent rainfall infiltrates into the soil or runs off into the surface water reservoirs. This can be a large impat: If rain falls on land with little vegetation over, the rain can quickly runoff and fill the reservoirs; conversely, if rain falls on lush vegetation the infiltration of that water into the soil is efficient and little water reaches the streams. This was ignored in previous research, but made clear from this NSF project.
