Application of the Triple Diagram Method in forecasting lake water level, on the example of Lake Charzykowskie

Abstract

The work focused on forecasting changes in lake water level. The study employed the Triple Diagram Method (TDM) using geostatistical tools. TDM estimates the value by information from an earlier two periods of observation, refers as lags. The best results were obtained for data with an average a 1-week lag. At the significance level of 1σ, a the forecast error of ±2 cm was obtained. Using separate data for warm and cold months did not improve the efficiency of TDM. At the same time, analysis of observations from warm and cold months explained trends visible in the distribution of year-round data. The methodology, built on case study and proposed evaluation criteria, may function as a universal solution. The proposed methodology can be used to effectively manage water-level fluctuations both in postglacial lakes and in any case of water-level fluctuation.