Drylands are the areas where yearly potential evapotranspiration is much larger than rainfall so water resources are scare and under threat due to: i) direct human impact (e.g. overuse of water resources); ii) land cover change (e.g. agricultural activity, deforestation/aforestation); and iii) climate change (reduction of rainfall/increase of temperature). The complexity of these impacts, particularly when acting jointly, requires adequate assessment techniques and solutions. Two types of research approaches will be presented at three different scales: i) catchment scale (large scale); ii) plot scale (medium scale) and iii) tree scale (small scale). These approaches, will be presented in three steps: 1) data acquisition; 2) data integration; 3) modelling.

At the catchment scale, various case studies addressing problems of impacts of human activity, land cover change and climate change on water resources, will be discussed. In these studies, different methods of data acquisition will be presented involving: remote sensing, hydrogeophysics, automated monitoring and traditional field sampling. For data integration with numerical distributed models, geospatial data analysis platforms will be evaluated regarding their ability to be interfaced with Integrated Hydrological Models (IHMs). Finally, it will be shown that the IHMs provide optimal solutions for sustainable water resources management.

The research at the plot and tree scale focuses on interactions of trees with water and soil, addressing mainly the ongoing world-debate whether planting or cutting trees in drylands, increases or decreases water resources. In contrast to traditional understanding, lots of recent research, including our research, shows that afforestation of a land, leads to decline of water table and decrease of available water resources. These conclusions are based on whole-tree and plot scale water balance studies, carried out using: i) field sampling; ii) field monitoring of tree sap flow, groundwater level, soil moisture and micrometeorological variables; iii) high resolution remote sensing; and iv) modelling.