Formatted Title
The Development of Decision Unit Multiple Increment Sampling (DUMIS) Guidelines for Site Characterization in China
Background/Objectives
The first two decades of the new millennium have seen dramatic changes in the skylines and improved living conditions of China's major cities. Brownfield redevelopment provides a valuable source of land, often in good locations. The urge for land development and the requirements of relevant regulations and laws have led to the development of a series of technical guidelines for environmental site investigation, monitoring and remediation confirmation, etc.
Similar to guidelines in many other countries, existing Chinese guidelines for the characterization of contaminated soils at industrial sites focused on past guidance for the collection of discrete samples. Geostatistical tools are then used to interpolate discrete soil data to derive soil contamination distribution maps. As cost-effective as this may seem, the combination of discrete data and geostatistical interpolation often results in incorrect predictions of average concentrations or contamination boundaries.
Approach/Activities
The main problem with discrete sampling is poor spatial representativeness. Numerous studies have shown that heterogeneity illustrated by the distribution of contaminants in soil can be in the range of a few centimetres or meters. In contrast, the resolution of traditional site characterization is typically 20 meters or more. Therefore, discrete sampling can never provide sufficient site characterization to support truly reliable, cost-effective remediation and rapid site closure. This heterogeneity and field sampling error has remained largely unobserved in traditional discrete sample investigations due to the lack of appropriate field replicates.
Furthermore, from a risk assessor's point of view, it is the average concentration for a defined Exposure Area that matters, not the concentration at single points, because by default receptors are randomly exposed to contaminated soil in a given exposure area. In the worst case, discrete data can miss a truly contaminated site (or part of a site), only to find it contaminated at a later stage. Although the drawbacks of discrete sampling plus geostatistical interpolation have been known for some time, this approach is still used in China and many other countries.
Pierre Gy's pioneering work on the Theory of Particulate Sampling (TOS) laid the foundation for the development of multiple increment sampling (MIS) methods. Similar terms include incremental sampling methodology, ISM, etc. USEPA, USACE, ITRC, and several state environmental agencies have published policy or technical guidance on the use of TOS for sampling heterogeneous solids such as contaminated soils. The technical document published by the Hawaii Department of Health is perhaps the most comprehensive and up-to-date. With the assistance of Dr Roger Brewer of the Hawaii DOH, the Institute of Soil Science, Chinese Academy of Sciences (ISSCAS) has conducted field research at dozens of agricultural and industrial sites in China. Based on the experience gained in both the US and China, ISSCAS successfully proposed and drafted Chinese national guidelines for DUMIS sampling, which has been in effect since 1 October 2023.
Results/Lessons Learned
Although it is not currently mandatory, the Chinese DUMIS guidelines provide an alternative approach to site characterization with the aim of achieving sample representativeness, reproducible data and reliable decisions.