Separating the effects of changes in land cover and climate: a hydro-meteorological analysis of the past 60 yr in Saxony, Germany

Abstract. Understanding and quantifying the impact of changes in climate and land use/land cover on water availability is a prerequisite to adapt water management; yet, it can be difficult to separate the effects of these different impacts. In this paper we illustrate a separation and attribution method based on a Budyko framework. We assume that evapotranspiration (E_T) is limited by the climatic forcing of precipitation (P) and evaporative demand (E₀), but modified by land-surface properties. Impacts of changes in climate (i.e., E₀/P) or land-surface changes on E_T alter the two dimensionless measures describing relative water (E_T/P) and energy partitioning (E_T/E₀), which allows us to separate and quantify these impacts. We use the separation method to quantify the role of environmental factors on E_T using 68 small to medium range river basins covering the greatest part of the German Federal State of Saxony within the period of 1950–2009. The region can be considered as a typical central European landscape with considerable anthropogenic impacts. In the long term, most basins are found to follow the Budyko curve which we interpret as a result of the strong interactions of climate, soils and vegetation. However, two groups of basins deviate. Agriculturally dominated basins at lower altitudes exceed the Budyko curve while a set of high altitude, forested basins fall well below. When visualizing the decadal dynamics on the relative partitioning of water and energy the impacts of climatic and land-surface changes become apparent. After 1960 higher forested basins experienced large land-surface changes which show that the air pollution driven tree damages have led to a decline of annual E_T on the order of 38%. In contrast, lower, agricultural dominated areas show no significant changes during that time. However, since the 1990s effective mitigation measures on industrial pollution have been established and the apparent brightening and regrowth has resulted in a significant increase of E_T across most basins. In conclusion, data on both, the water and the energy balance is necessary to understand how long-term climate and land cover control evapotranspiration and thus water availability. Further, the detected land-surface change impacts are consistent in space and time with independent forest damage data and thus confirm the validity of the separation approach.