Water Quality Data in Environmental Litigation;
When a case involves environmental chemistry (water, sediments, soils, rocks) or animals (population sizes, distributions, habitat relations) getting the statistical analyses of available data correct — and communicating them to the finders of fact clearly and effectively — are critical.
This is particularly true when a client is penalized by a regulatory agency or sued by an environmental NGO for violating the Clean Water Act by exceeding the arbitrary maximum concentration limit for the chemical. Quite often that the assessment was based on incorrect handling of non-detected values. Correct analyses of the data might show that the high concentrations are inherent natural variability or are artifacts of mis-handling of the very low values.
In a CERCLA (‘Superfund’) case, correctly analyzed water quality data could reduce client liability and, if the chemical of interest is tightly bound to a substrate and not biologically available they might not be libel for any damages.
When an analytical chemistry laboratory reports that an environmental chemical constituent was “not detected” or “below method detection limit” it means the value it unknown; it could be zero or anything up to the minimum reporting level. Therefore, any data analysis that sets these non-detected values to zero, the reporting level, or any arbitrary value between them (half the difference between zero and the reporting level is common and used by the EPA) is wrong. There are statistical models that correctly incorporate non-detected values robustly and correctly, even when these values are 80% of the data set.
Three important aspects of regulatory science for the Clean Water Act depend on technically sound and legally defensible statistical analyses of environmental chemistry that should be clearly explained to the finders of fact in your case:
1) concentrations of individual elements (e.g., toxic metals) or molecules (e.g., organic pesticides) are rarely found in natural systems;
2) how chemical constituents affect specific beneficial uses at specified locations need to be detailed to support the allegations of adverse impacts; and
3) the inherent variability of natural ecosystems needs to be quantified to separate anthropogenic impacts.
Most harmful chemicals in water bodies are found in very low concentrations for several reasons. Among these reasons are chemical bonding to substrates. Dioxins in the Columbia River, for example, could be detected and quantified only after the substrate on which they were bound was subjected to very high temperatures and pressures in concentrated acid for several hours in an autoclave (a sterilizing instrument). While the separated dioxin concentrations could be measured the chemical was not biologically available; that is, even if some aquatic organism ate sand and gravel their digestive systems would not be able to separate the dioxin from the substrate.
Beneficial uses are set by regulators over broad areas of a river network and most permit compliance monitoring is at only a few specific locations. Without knowing the concentration of the chemical constituent of concern along the whole reach of the designated beneficial use it is not possible to claim harm to that use. Further complicating this issue is the infrequent measurement of chemical concentrations and (very importantly) no clear description of what specific harm is caused to the beneficial use or how that is measured. Broad generalizations are not technically sound or legally defensible.
Natural ecosystems vary over time and at different scales. Flowing water systems are inherently highly variable. For example, water temperature in headwater reaches of a river can vary by 40F from 4:00 am to 4:00 pm in the summer. Flow rates vary seasonally (high in winter and when snow melts and runs off, low at the end of summer),
annually (the western US has been in a drought for more than a decade), and weekly (the Columbia River at Portland, OR, varies in level from a low Monday through Thursday then rises about 2 feet Friday through Sunday as water behind the Bonneville Dam is diverted through the turbines to provide the increased electrical demands of the weekend. Any observed change in chemical component concentrations cannot be assumed caused by anthropogenic activities. Separation of natural variability and anthropogenic input can be done with the proper set of data and certainly allows finders of fact to better understand ecosystem dynamics.
When your case depends on correct analyses of environmental chemistry or biota, and you are not certain that you have the most technically sound and legally defensible statistical analyses and ecological interpretations of the results, contact me to discuss your concerns.
About the Author:
Aquatic Ecology Hydrology Fluvial Geomorphology Expert Witness No. 1109 is a river ecologist, hydrologist, fluvial geomorphologist (an expert in rivers and the watersheds they drain), and an expert in environmental impact assessments and environmental data analyses using advanced statistical and spatial models. His aquatic ecosystem expertise and experience includes fish and invertebrate biology, aquatic ecology, wetlands, hydrology, hydraulics, sediment transport, and other processes that shape river systems and determine their structures and functions. The Expert has extensive experience modeling natural systems using spatial analytical (GIS) software, statistics, and numeric process models. He has been an aquatic ecology/hydrology/fluvial geomorpholgy expert witness for the coal mining company defendant in a preliminary hearing and modeled terrain and hydrological processes to support the defense in another lawsuit involving flooded property.