Jump to main content or area navigation.

Contact Us

Water: Recovery Potential

Displaying Results

Recovery Potential Screening generates scores as a basis for evaluation, but numerical comparisons through ranking are just one of the techniques that can be used to evaluate and communicate screening results.  Sometimes, results are more effective when formatted in ways that help people understand how conditions in specific places relate to broader, general patterns. For most people, numerical data in tables alone do not communicate this well, even if numbers were the basis for a decision or strategy. This section of the website describes three alternative techniques used for supporting strategic discussions and decisions and communicating recovery potential findings effectively: rank-ordering, mapping and plotting graphs.

Rank-ordering recovery potential scores. The simplest of comparison methods, rank-ordering organizes screened waters from highest to lowest recovery potential based on their scores. This is an easy and transparent method to identify a subset of more restorable waters, whether by selecting a specific number or a percentage of highly-ranked waters in consideration of available restoration resources and capacity.  The appeal of rank-ordering is its clarity and simplicity - some waters or watersheds simply score higher than others - which may help users move past overly complex discussions without decision criteria.

The flexibility in rank-ordering is in which scores to use. The RPI score is one option for a single overall score used to rank-order all the waters based on all the indicators measured. In addition, each of the three summary indices are easy to rank-order to compare scores on a purely ecological, stressor, or social basis. It is also possible to examine the rank-ordering of a single indicator, if sufficiently important.  It is crucial to remember that, if the stressor index is used for rank-ordering, the lowest scores are rank-ordered as the highest recovery potential.  For the ecological and social indices and the RPI score, the highest scores are associated with the highest recovery potential.

Although rank-ordering is useful in distinguishing major differences between very high and very low watershed scores, most data likely to be used in a screening probably does not support assuming that very small scoring differences are significant.  For example, the 237th-ranked watershed probably is not clearly less restorable than the 236th-ranked watershed.  Users are advised not to overemphasize the significance of very small rank-order differences.  One option for further organizing rank-ordered lists in a more generalized ranking is to group them by quantiles, which can be equal-size or separated by natural breaks in the range of values.  See also ways to use rank-ordering (PDF) (4 pp, 727K, About PDF) in Recovery Potential Screening.

Mapping recovery potential assessment results. Much if not most Recovery Potential Screening data is generated through analyses performed with Geographic Information Systems (GIS) that generate maps and related attribute information tables. Many of your indicator values may have been calculated using GIS and compiled in a database file, with these values all linked as attributes to each water body or watershed ID. This linkage to specific geographic features makes it possible to display all your indicators, singly or in combination, on a map basis.

Maps are highly effective at communication of environmental conditions and relative differences between geographic areas. One limitation, however, is the difficulty of displaying two or more parameters that each range in value. As Recovery Potential Screening produces summary scores along three key themes affecting restorability (ecological capacity, stressor exposure and social context), the ability to map all three for simultaneous viewing would be challenging. On the other hand, maps based on a single summary score (e.g., the ecological, stressor or social index, the RPI score, or a single indicator of special interest) are generally clear and effective.  For several examples of mapping techniques applied to analyzing and communicating about recovery potential see also ways to use mapping (PDF) (4 pp, 653K, About PDF) in Recovery Potential Screening.

Displaying results in 3D "bubble plots". This comparison method was adapted specifically for Recovery Potential Screening as a way to visualize the relative influence of ecological, stressor and social context factors on restorability at the same time.  A 3D plot displays screening results for individual waters (each one a 'bubble' on the graph), which are plotted relative to X (stressor index) and Y (ecological index) axes, with bubble size varying with social context index score. For example, this type of display might enable a user to identify those waters that have high ecological and low stressor summary indices by where they fall on the plot, and then further sharpen their focus on those that have larger dots implying higher social scores.

Bubble plots are ideally suited for displaying the ecological, stressor, and social context indices in a way that observers get a sense of the three scores at once. Unlike rank-ordering's simplicity, bubble plotting acknowledges that comparisons are often complex and offers a systematic way to observe and consider the relative influence of three major driving factors on restorability simultaneously. Additional information display concerning a fourth 'dimension' (such as healthy vs. impaired) is possible through use of multi-color bubble plots. See ways to use bubble plotting (PDF) (6 pp, 619K, About PDF) in Recovery Potential Screening for a discussion of how this technique can aid comparison and targeting decisions.

Below is a downloadable tool for creating bubble plots from recovery potential screening data. The tool runs data from a recovery potential data spreadsheet through a 3D plotting script written in the open-source R statistical program Link to EPA's External Link Disclaimer, which is used widely in a variety of scientific activities. Bubble plot instructions with examples are also available.

3D Bubble Plot Script (1K, zipped R script)
3D Bubble Plot Instructions and Examples (PDF) (8 pp, 594K, About PDF)


Jump to main content.