STOCHASTIC SIMULATION OF SPATIAL-DISTRIBUTION OF SOIL-MOISTURE

A stochastic method is presented which uses a soil moisture model (R model) to calculate the spatial distribution of soil moisture in arable land affected by percolating water. The method proceeds from the variability of two model parameters in the area under examination and their correlations. The model parameters have to be distributed normally or they have to originate from a normal distribution. As an example, a comparison was made between the proportions of waterlogged areas in a very small catchmentare aduring spring cultivation determined either by mapping in the field or by simulation. The above method is superior to soil moisture simulation made merely on the basis of mean values of soil parameters. Advantages over another stochastic method are discussed.