1 Quick calculation of minimum in-stream flow in drainage basins of Central Alps G. Becciu - A. Bianchi - C. A. Fassò - E. Larcan Politecnico di Milano, Italy - Dpt. of Hydraulic, Environmental and Surveying Engineering. A. Fassò University of Bergamo, Italy – Dpt. of Engineering (partially supported by italian MURST cofin98 grant). Keywords: Environmental Statistics, Regression Diagnostic, Minimum in-stream flow, Rainfall- runoff model ABSTRACT: The unit discharge corresponding to the minimum in-stream flow in a given reach of a stream is expressed as the product of an hydrologic factor f H which includes the hydrologic in- formation on the catchment, and an environmental factor f A , which includes the biologic and envi- ronmental characters to be safeguarded. As for f H , 1/3 of the minimum annual value of the daily unit discharge averaged over at least five years is assumed. When measurements of daily flowrates are not available, f H is to be calculated by means of a rainfall-runoff model, such as the PRMS. To enable a quicker calculation of f H , some multiple linear regression models are proposed between f H and some geomorphologic and hydrologic characters of the catchment (mean elevation h, average slope i of the main river channel upstream of the section and average annual precipitation P). The models are tested over a sample of 42 catchments in Central Alpine region of Lombardia. 1 INTRODUCTION In the course of a research project conducted on behalf of the Lombardia Regional Government by the Department of Hydraulic, Environmental and Surveying Engineering (DIIAR) of the Politec- nico di Milano concerning the evaluation of minimum in-stream flow in Central Alpine catchments of Lombardia, the unit discharge corresponding to the minimum in-stream flow in a given reach of a watercourse was expressed as the product of an environmental factor f A and an hydrologic factor f H . The environmental factor is intended to include the biologic and environmental characters to be safeguarded. According to the results of the research, f A may be assumed variable between 1 and 2, its value depending on the level of naturalness of the fluvial ecosystem, on the water quality (through the Extended Biotic Index), on the morphology of the river bed (pool percentage) and on the administrative and operational constraints related to the possibility of recreational, touristic and fishing uses (e.g. f A = 2 is suggested for catchments included in natural parks). The hydrologic factor f H , which is intended to include the hydrologic information on the catchment, is assumed equal to the fraction 1/n of the minimum annual value of the daily unit dis- charge averaged over at least five years (according to the rules followed e.g. in some Italian, Ger- man and Austrian regions). In the present paper n = 3 is assumed, but obviously the results can be adapted immediately to any other value of n. f H could be calculated directly if measured daily flowrates were available, but this is a very rare opportunity in small mountainous catchments, such as those involved in the said research. Lack of direct flow measurements, the problem can be solved resorting to one of the rainfall-runoff models proposed in the literature, which can be calibrated using the precipitation data as well as the natu- ral flow data for hydrologically similar monitored catchments; after calibration, the model is ap- plied to the catchment under study, for the purpose of simulating its natural flowrates on the basis of precipitation measurements in the catchment itself (if available) and in the surrounding region.