Water: Recovery Potential
Step 3: Measure the Indicators
Screening Methodology Tutorial
When the necessary datasets are available and the preferred indicators have been selected, indicator measurement can begin. The analysis first involves compiling a database file with numerous fields of information about each water body or watershed you are screening. This raw data table or tables will be structured with individual waters or watersheds as rows and each of the indicators as columns. Depending on the sources of your indicators, compiling the table may involve a mix of GIS attribute tables, spreadsheet information from monitoring, and other datasets.
Note that even when the screening focuses on water bodies it is necessary to have data on a watershed basis as well. Many recovery-related characteristics are measured on the watershed or on a defined corridor (e.g., characteristics within a 100-meter wide stream buffer).
Develop baseline data attributes. Several attributes of each water body or watershed should be compiled as basic reference information, some of which may also be used in calculating indicator values. For example, watershed area is basic reference information but it is also used in several indicators that involve the percent area of specific land cover types within the watershed.
A useful set of baseline attributes would include:
- water body or watershed ID
- common name
- subgroup, if any (e.g., ecoregion, water body type)
- watershed total area
- watershed lake/reservoir area and land area
- watershed total stream length
- individual stream length, lake area, or corridor area (if individual lakes, rivers or streams are being screened instead of watersheds)
Compile raw data table of indicator values. Using the measurement processes developed for each indicator in step 2, you may now perform the GIS analyses and import existing data to your raw data table from other sources. Some indicators may need to be calculated through a multi-step process. In such cases, data management may be easier if interim values from the calculation process are kept separate from the data table containing only the completed indicator values. It is also helpful to organize the indicators in the raw data table into the three indicator classes (ecological, stressor, and social context) after all have been initially calculated.
Directionally align indicator values. Indicator scores in each of the ecological, stressor, and social context classes need to be directionally consistent within-class, with respect to their effect on recovery potential. All ecological indicator scores should be aligned so that higher values imply better recovery potential, and social context scores should also be aligned so that higher is better. Stressor indicator scores, as most users would expect, are aligned to have a higher score associated with lower recovery potential. Correct and consistent directionality within each indicator class is critical to the quality of your screening results. Verify that higher is better for ecological and social indicators, and lower is better for stressor indicators.
Sometimes it can be necessary to reverse the order of the numerical raw scores of an indicator (for example, an ecological indicator like % highly erodible soils, whose higher values are associated with lower restorability) to align it with the other ecological indicators. If directionality wasn't verified in Step 2's candidate indicator evaluation, it should be rechecked and resolved at this point.
Perform a QA/QC check. Quality control procedures should be part of the indicator analysis and data table development steps. Many sources of human error are minimized by using the scoring spreadsheet (MS Excel xls, 2.65MB) described in step 4, but quality control checks should also address the raw data table, the data transfer from data table to the spreadsheet, and the outputs of the spreadsheet itself. One commonly used approach involves spot-checking sample watersheds by manually checking raw values and repeating or verifying the indicator calculations. It is especially important for your QA/QC process to detect indicators that are not directionally aligned, and data transfer errors in which an entire indicator's values may be incorrect due to faulty calculation, miscopying or mislabeling. These two types of error can skew the results substantially but are relatively easy to find through QC before they do their damage.
Your evaluation procedures should also examine each indicator's set of values in comparison to reference sites of known quality, including healthy as well as impaired waters or watersheds. For each indicator's measured set of values, observe whether the indicator performed as expected with regard to these sites. For example, did a high percentage of healthy reference sites score in the top quartile for a specific ecological indicator? If healthy reference site scores were low, the indicator might have been incorrectly scored. If scattered throughout the value range without apparent pattern, the indicator may be weakly related or unrelated to recovery potential and perhaps should be dropped.
After your raw data table is compiled and checked, you can archive a dated copy for safekeeping and then prepare to use a working copy of the same file for performing your first screening run in step 4.