A METHODOLOGY TO EVALUATE THE RELIABILITY OF THE BELGIAN SOIL MAP FORPREDICTING THE ACTUAL WATER-TABLE CHARACTERISTICS

Mast data on Belgian soils were gathered during the National Soil Surv ey (1947-1971), which resulted in the Belgian soil map (1:20,000). Due to the impact of man on the soil some of this information may be outd ated. This study presents a methodology for evaluating soil maps with respect to actual water table characteristics. Ninety dip-wells were l ocated on permanent grassland of two contrasting areas. The phreatic w ater table depth of these dip-wells was recorded fortnightly during a period of 2 to 9 years (1984-1992). A dip-well specific, regressive mo del was developed to describe these water table depth fluctuations and to detect changes in water management. This was found to be the case in 10 dip-wells, making them unsuitable for further analyses. Of the r emaining 80 dip-wells the mean highest water table depth and mean lowe st water table depth were calculated as characterising parameters of t he water table depth fluctuations. In the Polder Area we found several indications of changes in the fluctuations of the water table depth d ue to human interventions during recent decades. Recorded depth to the gIey mottles was found to be no longer related to the actual water ta ble depth characteristics (R(2) = 0.01 ). Accordingly, if information on the actual water table depth fluctuations is needed, these pedogene tic features are no longer workable. The Plateau of Izenberge was hard ly influenced by water management practices. Therefore, the depth to t he gley mottles could still be related to the mean highest water table depth (R(2) = 0.68). Accordingly, it remains possible to quantify the mean highest water table depth on the basis of traditional soil surve y observations like those used during the National Soil Survey. As a c onsequence, the soil map data of this area can still be used to predic t the actual water table characteristics.