Department of Health Seal

TGM for the Implementation of the Hawai'i State Contingency Plan
Section 21.4
ANTICIPATING AND ADDRESSING DATA GAPS

21.4 Anticipating and Addressing Data Gaps

The risk assessor should characterize and address data gaps during the scoping phase of the ERA, as part of the DQO process (see TGM Section 3). A data gap can be generally categorized as resulting from one of two sources: natural variability or incomplete knowledge. A direct evaluation of these types of data gaps can strengthen the DQO process and guide the risk assessor toward a more robust sampling design and a more defensible risk assessment.

The risk assessor should first distinguish between data gaps that result from incomplete knowledge and data gaps that result from inherent variability in the ecosystem. This categorization is based on general knowledge of environmental processes at the site, the CSM, the COPECs, and available data (Table 21-9).


Table 21-9. Data Gap Analysis
For data gaps that result from natural variability in the ecosystem, answer the questions below:
  • Could this data gap be filled by additional study? (If you answer yes, make sure you have correctly identified the data gap as resulting from natural variability rather than lack of information).
  • What is the source of variability for the parameter in question? Daily or seasonal fluctuations, genetic variations (including gender), age, size, and other features may introduce variability. Note that natural variability encompasses differences within the same individual over time (lifetime, seasonal, or daily); among individuals within a population (based on gender, size, or other factors); and among populations.
  • Are existing data adequate to describe the variability statistically using probabilistic models and other quantitative techniques?
    • If yes, describe the methods used to develop probabilistic values and clearly explain any residual uncertainty associated with the values used in the ERA.
    • If no, choose one of the following:
      • Use the most conservative (i.e. most protective) value from the available range and provide rationale for why that value is or is not representative of conditions at the site.
      • Conduct additional study (sampling) to provide the necessary data covering the range of variability.
For data gaps that result from incomplete knowledge about a particular site, chemical, or receptor, answer the questions below:
  • Could this data gap be filled by additional study?
  • What is the range of possible values for the parameter in question?
  • Work through two hypothetical scenarios using the maximum value and the mean value for this parameter, respectively.
  • Consider the two results: Are the results of the two hypothetical scenarios different enough to substantially change remedial decisions at the site?
    • If no, then don’t waste time or money refining this value. (Use the maximum as a default value.)
    • If yes, estimate the value (or order-of-magnitude) at which a different decision would be triggered and design a study to develop a realistic value. The study could be desk-based, in which you search the existing literature and develop a rationale for extrapolating from another study, or for amassing a large set of relevant data to provide a reasonable context for your site. If the value is critical to a decision that will lead to a very expensive or controversial remediation, then you may find it is justifiable to conduct a site-specific study.