In the absence of hydrological information, a model for calculating daily levels of periglacial lakes is proposed. The model is based on water balance relationships for different altitude intervals and takes into account the type of the catchment surface. The choice of the modelling object is due to the availability of field glaciological, meteorological, actinometric and hydrological data measured directly in the catchment area located on the northern macroslope of the South-Chuya Ridge during two ablation seasons. The input data required for modelling were conditionally divided into three blocks: meteorological data without precipitation, precipitation and actinometric data; and the network of state meteorological stations, reanalysis, reference and scientific literature, as well as field data were used as the main (open) sources of initial data. The paper presents the results of modelling water levels obtained with various combinations of sources of initial information. Comparison of the calculated and observed levels showed the possibility of using any sources of information, with the best convergence revealed in calculations based on field data. In mountainous conditions, it is preferable to use data from global weather reanalysis models as an alternative source of information on precipitation due to better interpolation of meteorological station data and consideration of the altitude-exposure features of mountainous areas. In the absence of measured actinometric data, the use of regional values for calculating the melting layer gives a satisfactory result. It was revealed that for unexplored high-mountain basins, the optimal combination of information sources for modeling daily water levels is: meteorological data (excluding precipitation) - from a network meteorological station, precipitation - from reanalysis, actinometric characteristics - from literary data.