The Reformulated Impervious Cover Model
Is Impervious Cover Still Important?
As many of you know, I first introduced the impervious cover model (ICM) back in 1994 as a way of describing the strong negative relationship between subwatershed impervious cover and various indicators stream health. Since then, I have come to realize that virtually everyone hates the ICM. Over the years, stream ecologists have informed me that it is a gross simplification of a variable world, smart growthers complain that the ICM rewards low density sprawl and river advocates contend that it sacrifices urban streams.
Planners worry that just about any level of habitable density they create will condemn streams to death, engineers argue that the ICM omits the wonderful impact of stormwater practices they design, and water quality regulators are offended that it suggests that some highly urban waters can never meet water quality standards. You can also be certain that the ICM is equally loathed by developers, new urbanists, economists and paving contractors.
Despite the fact that it doesn’t get much love, the ICM still hangs in there. The basic predictions of the ICM have been confirmed by a recent review of nearly 60 peer-reviewed stream research studies released in the last five years. These findings are contained in a paper to be published in the Journal of Hydrologic Engineering in 2009 by Lisa Fraley-McNeal, Karen Cappiella and myself, Tom Schueler. A summary of the findings were presented at the Symposium on Urbanization and Stream Ecology held this summer.Final 2014 Baywide Stormwater Retreat Agenda (.pdf)
Basically, two thirds of all the stream monitoring studies confirmed or reinforced the basic ICM relationship, including all ten studies conducted in the Chesapeake Bay watershed. The new studies did point out some new wrinkles and caveats about the IC/stream quality relationship. As a result, we have slightly reformulated the ICM model to reflect this new research.Final 2014 Baywide Stormwater Retreat Agenda (.pdf)
The reformulated ICM includes three important changes to the original model. First, the Impervious Cover (IC)/stream quality relationship is no longer expressed as a straight line, but rather as a “cone” that is widest at lower levels of IC and progressively narrows at higher IC. The cone represents the observed variability in the response of stream indicators to urban disturbance and also the typical range in expected improvement that could be attributed to subwatershed treatment. In addition, the use of a cone rather than a line is consistent with the findings that exact, sharply defined IC thresholds are rare, and that most regions show a generally continuous but variable gradient of stream degradation as IC increases.
Second, the cone width is greatest for IC values less than 10%, which reflects the wide variability in stream indicator scores observed for this range of streams. This modification prevents the misperception that streams with low subwatershed IC will automatically possess good or excellent quality. As noted earlier, the expected quality of streams in this range of IC is generally influenced more by other watershed metrics such as forest cover, road density, riparian continuity, and cropping practices.
Third, the reformulated ICM now expresses the transition between stream quality classifications as a band rather than a fixed line (e.g., 5 to 10% IC for the transition from sensitive to impacted, 20 to 25% IC for the transition from impacted to non-supporting, and 60 to 70% IC for the transition from non-supporting to urban drainage). The band reflects the variability in the relationship between stream hydrologic, physical, chemical, and biological responses and the qualitative endpoints that determine stream quality classifications. It also suggests a watershed manager’s choice for a specific threshold value to discriminate among stream categories should be based on actual monitoring data for their ecoregion, the stream indicators of greatest concern and the predominant predevelopment regional land cover (e.g., crops or forest).
So, the science debate is largely over. Gradients definitely occur in stream quality as a result of urbanization. More research is needed to determine the degree to which these gradients can be modified by watershed protection and restoration practices.
The management debate, however, is just beginning. The question of the day is how does our society responds to the basic ICM relationship? To date, communities have experimented with a wide range of planning, zoning, engineering, regulatory and economic tools to defeat the ICM as shown in the table below. Regrettably, none of these tools individually seems to work well enough. A recent Technical Bulletin issued by the CSN reviews the effectiveness of these tools and argues for a more sophisticated system of urban stream classification, management and permitting.
|Range of Responses to Mitigate the ICM|
|Planning and Zoning Tools||Engineering Tools|
|Regulatory Tools||Economic Tools|