Improving the hydrological role of vegetation over Mediterranean mountain areas: the study case of the dehesa ecosystem in Cardeña-Montoro Natural Park (Spain)

Dehesas are a characteristic Mediterranean mountain ecosystem, result of a co-evolution of autochthonous ecosystems and human settlement in a sustainable balance, with high relevance from both the environmental (biodiversity) and socioeconomic (livestock farming, including Iberian pork food industry) point of views. They are characterized by a complex vegetation cover structure composed of isolated trees, mainly holm oak, cork oak and oak, Mediterranean shrubs, and pastures. This complex vegetation organization is the main determinant of the water and energy balance partitioning and therefore of the hydrological states of the ecosystem. The methodologies traditionally used to assess the hydrological role of vegetation usually are not able to correctly represent this complexity. On the one hand, distributed physically based hydrological models usually lack a detailed vegetation representation at subgrid scale. On the other hand, the spatiotemporal resolutions of traditionally used satellite remote sensing information are not enough to capture these vegetation dynamics.

This work proposes to improve the vegetation representation over this complex ecosystem combining remote sensing information at different spatiotemporal scales. A dynamical spectral mixture analysis using Sentinel-2 data is developed in the Cardeña-Montoro Natural Park (southern Spain). Terrestrial photography taken hourly at two spatial scales, detail (∼m) and plot (∼100m) at an experimental site located within the area, are used as ground truth. Detail scale images allowed to monitor vegetation changes throughout the year (e.g. tree blooming, pasture wither) and are used to select different combinations of endmembers to include in the analysis. Images at the plot scale are used to validate the obtained vegetation dynamical product. The new product, with errors smaller than 5% during the validation process, not only constitutes a better representation of vegetation, but is also the first step in a data assimilation scheme for better hydrological processes representation over these areas.

This work has been funded by project SIERRA Seguimiento hIdrológico de la vEgetación en montaña mediteRránea mediante fusión de sensores Remotos en Andalucía, with the economic collaboration of the European Funding for Rural Development (FEDER) and the Office for Economy, Knowledge, Enterprises and University of the Andalusian Regional Government.