Conceptual runoff models have become standard tools for operational hydrological forecasting in Scandinavia. These models are normally based on observations from the national climatological networks, but in mountainous areas the stations are few and sometimes not representative. Due to the great economic importance of good hydrological forecasts for the hydro-power industry attempts have been made to improve the model simulations by support from field observations of the snowpack. The snowpack has been mapped by several methods; airborne gamma-spectrometry, airborne georadars, satellites and by conventional snow courses. The studies cover more than ten years of work in Sweden. The conclusion is that field observations of the snow cover have a potential for improvement of the forecasts of inflow to the reservoirs in the mountainous part of the country, where the climatological data coverages is poor. This is pronounced during years with unusual snow distribution. The potential for model improvement is smaller in the climatologically more homogeneous forested lowlands, where the climatological network is denser. The costs of introduction of airborne observations into the modelling procedure are high and can only be justified in areas of great hydropower potential.