Preliminary results regarding the simulation of a streamflow strongly influenced by anthropogenic use in an alpine context: the case of the Grosina valley (northern Italy).

The study investigates the interactions between surface water and meteo-climate processes in an Alpine valley (Grosina, northern Italy) characterized by anthropogenic modifications affecting the hydrologic cycle. Grosina valley, an accessory valley of Valtellina on the border between Italy and Switzerland, features a central-alpine climatic type. The valley is composed of two main branches - Eita (62 km²) and Sacco (71 km²). Along the Eita stream, close to its confluence with the Sacco stream, a dam was built in 1960 for hydro-power exploitation and regulation purposes. After the confluence, the river takes the name of Roasco. Anthropogenic modifications of the natural water system include two diversion channels in the main branches that connect them to the dam and a third diversion tunnel that brings a high volume of water into the dam lake from a hydroelectric plant located outside the watershed.

The study general aim is two-fold: i) setting up a prototypal operational hydrologic model (forecast period of about one week) for water use management and ii) applying a hydrologic model for estimating impacts of climatic changes on water resources and the hydrologic cycle in the medium/long-term (decadal and multi-decadal analyses). The first step of this project is common to the two aims and involves the definition of the conceptual model and the implementation-calibration of a hydrologic model in such a challenging environment, representative of the multiple and concurrent uses of water resources in mountain areas.

The modeling of the Grosina valley catchment has been carried out exploiting the potentialities of the GEOframe system, an open-source, semi-distributed hydrologic model. It is a component-based model since it is developed starting from the creation of single modules (components) that describe the principal physical processes of the hydrologic cycle. After identifying Hydrological Response Units (HRUs) and their connections through a geomorphological analysis, contributions and losses to the system were considered by exploiting the components of meteorological data interpolation (from 22 stations), radiation calculation, partitioning between solid and liquid precipitation, and evapotranspiration. Then, the calibration of the model was performed by comparing the simulated flow to discharge data recorder at the diversion points, the dam, and a hydrometer placed at the end of the valley (hourly timestep). This phase proved to be very complex and demanding since only the measure of the derived flow, namely the flow captured for hydropower purposes, was available. Therefore, at the diversion points, it was chosen to estimate the natural flow as the sum of the derived flow and the minimum environmental flow (MEF), focusing the match between observation and simulation on the baseflow behavior rather than the discharge peaks. The calibration phase led to a good correspondence between simulated and observed flow with Nash-Sutcliffe Efficiency values greater than 0.6 at all points investigated.