Abstract. The regulating role of glaciers in catchment run-off is of fundamental importance in sustaining people living in low-lying areas. The reduction in glacierized areas under the effect of climate change disrupts the distribution and amount of run-off, threatening water supply, agriculture and hydropower. The prediction of these changes requires models that integrate hydrological, nivological and glaciological processes. In this work we propose a local model that combines the nivological and glaciological scales. The model describes the formation and evolution of the snowpack and the firn below it, under the influence of temperature, wind speed and precipitation. The model has been implemented in two versions: (1) a multi-layer one that considers separately each firn layer and (2) a single-layer one that models firn and underlying glacier ice as a single layer. The model was applied at the site of Colle Gnifetti (Monte Rosa massif, 4400–4550 ma.s.l.). We obtained an average reduction in annual snow accumulation due to wind erosion of 2×103 kgm-2yr-1 to be compared with a mean annual precipitation of about 2.7×103 kgm-2yr-1. The conserved accumulation is made up mainly of snow deposited between April and September, when temperatures above the melting point are also observed. End-of-year snow density, instead, increased an average of 65 kg m−3 when the contribution of wind to snow compaction was added. Observations show a high spatial and interannual variability in the characteristics of snow and firn at the site and a correlation of net balance with radiation and the number of melt layers. The computation of snowmelt in the model as a sole function of air temperature may therefore be one of the reasons for the observed mismatch between model and observations.