Conditional Simulation For Spatial Variability In The Groundwater Contamination Transport

Affected by environmental medium and many factors existing in the groundwater environment, the transport ion of pollutants in the groundwater always had high spatial variability. In the study of quality assessment, risk assessment and contamination transport model of the groundwater, we always needed to get enough character of spatial variability. Restricted by many terms, such as the storage and transportation, only few data could be easily got when monitored, which made the spatial distribution information offered by the mathematics model very useful. The conditional simulation is a special mathematics simulation, which can take both the randomicity and structuredness, and reflect the fluctuations and the discreteness of the variables' spatial information.Based on the theory of geostatistics, by the lower-upper triangle decomposition method from local window approach, we had written the calculation program under Mat lab environment. Four pollutants had been chosen as variables of field data of Lang Fang district, including Pb, Cd, Zn and Cu, in order to do conditional simulation on two dimensional surfaces. Then the realization of three grid density could be got, including high, intermediate and low. This paper also did contrast analysis of the result through statistical characteristics, distribution trends, variograms and spatial variation indices in order to examine whether the method could recur the spatial variation of the transport ion of pollutants in the groundwater.Combined the lower-uper triangle decomposition and the local window approach, this study achieved the sequencing and visualization of the conditional simulation calculation. The result showed as follows: lower-uper triangle decomposition is very simple, the simulation and conditioning can be taken at the same time, which enhanced the efficiency of the calculation and breached the limit of the local window approach which need the number of the grid less than 700. Through the contrast analysis, the conditional simulation could get satisfied result and reflect the fluctuations and the discreteness of the variables' spatial information through controlling the condition such as the parameters of the model, variograms and spatial variation indices, however high the density was.