Forecasting Water Levels for Lake Chad

A model for predicting the low levels in Lake Chad and a statistical model for describing the probability of the low‐level variations at various periods ahead are presented. The first model will assist in the planning of the agricultural operations one season in advance for schemes that rely on the lake water. The second provides the low levels on a probabilistic basis and can be used to measure the risk involved in selecting various abstraction levels of water from the lake during the proposed life span of the projects, and this is also a particularly useful aid in the economic and planning studies for future developments. The probabilistic model is based on the Markov chain process applied to two statistical models of the main input variable (the runoff from the River Chari); the first assuming the annual discharges to be random and the second incorporating significant cyclical components, although determined from inadequate data, and a large stochastic component.