Abstract
Before the soil contamination is investigated on the ground, there are soft information about soil contamination in many case. Soft information will suggest the possibility of soil contamination but it is vague such as ″This area may be high concentration but there is no surveyed data.″ So there are few methods in order to search optimal arrangement of sampling location using soft information. In this study, optimal arrangement of sampling location was decided by geostatistics method using informative prior distribution in order to apply soft information. The result showed that using soft information is more suitable than using non information in order to decide optimal arrangement of sampling location. Geostatistics method can also apply the previous survey of the general condition of soil contamination. One stage sampling plan is that all sampling location has decided in one time. Multiple stage sampling plan is that some sampling location has been decided and gives information to next arrangement of sampling location. When the number of total sampling points is equal, multiple stage sampling plan is more suitable than one stage sampling plan.
