Agribusiness’ report critical of Bay Model called flawed, misleading
Review panel said LimnoTech made 'notable errors' when comparing two models
A scientific panel has concluded that a highly critical review of the EPA's Bay Model that was paid for by agricultural organizations was deeply flawed and provided no justification to halt the implementation of Chesapeake cleanup plans.
Last December, the Agricultural Nutrient Management Council, which is funded by the American Farm Bureau Federation and other farm organizations, released a report that compared results from the EPA's Watershed Model - which was used to set Bay nutrient reduction goals - to results from a model developed by the U.S. Department of Agriculture.
It argued that substantial differences existed between the two models and that the EPA should withdraw its Chesapeake Bay Total Maximum Daily Load until those differences were resolved.
LimnoTech, a Michigan-based environmental consulting firm, developed the report, and it was widely cited in agricultural newspapers and websites, during congressional hearings and in a lawsuit by the Farm Bureau seeking to block the implementation of the TMDL, or "pollution diet," which set legally enforceable nutrient and sediment limits for the Bay.
But a panel convened by the Bay Program's Scientific and Technical Advisory Committee said in a sharply worded report that LimnoTech's review was "flawed and does not provide sufficient evidence to suspend implementation of the Chesapeake Bay TMDL."
The panel said LimnoTech's report had "poor scientific merit" and misinterpreted some model results, which had resulted in the appearance of greater differences between the EPA and USDA models than was actually the case.
The EPA, with support from the USDA, requested the review. It hailed the conclusion. "We're pleased this independent panel has confirmed what we have said all along," the agency said in a statement. "Now it's time to continue moving forward in supporting the Chesapeake Bay Watershed states' efforts to restore the Bay."
The Chesapeake Bay Foundation recently hired its own outside consultant to review the LimnoTech report and reached similar conclusions as the STAC review.
The LimnoTech report was part of a "coordinated attempt by national agribusiness lobbying groups to block implementation of the Chesapeake Bay pollution diet and to delay efforts to clean up the region's rivers, streams and the Chesapeake Bay," said Beth McGee, CBF senior water quality scientist. "It was an effort to mislead the public, the farm community and Congress using flawed science."
The STAC report said that LimnoTech's report was based on a "misunderstanding" that models developed for different purposes should produce the same results. Different models, the STAC panel said, use different input information, assumptions, mathematical algorithms and processes which make it nearly impossible to reach the same results.
"As modelers, we know that no model is perfect; that they are only at best an approximation of the real world," said Donald Weller, a senior scientist at the Smithsonian Environmental Research Center, who chaired the STAC panel. "Depending on what your goal is, you are going to make different choices in building your approximation. You are going to have to emphasize some processes, and de-emphasize others. You are forced to do that because you can't put everything in a model, so you are going to be making trade-offs. People, who make different choices because they have different goals, are going to get slightly different answers."
The Watershed Model was developed over a 30-year period to account for nutrient and sediment loads from all major sources in the watershed. In contrast, the USDA's model was designed specifically to identify the effect of conservation practices applied to cultivated cropland between 2003-2006.
As a result, the STAC report said, the USDA model has more detail about cropland, but less detail about other sources than the Watershed Model. But the Watershed Model has a finer level of detail, and is calibrated against monitoring data at 237 sites throughout the Bay watershed, as opposed to five locations for the USDA model.
But the differences within the models were appropriate based on the different uses for each, the STAC review said. "Consequently, the review panel concludes that it is scientifically unreasonable to expect the two modeling efforts to be in agreement to the extent suggested by LimnoTech."
It said LimnoTech had also made "notable errors" when comparing results from the two models that made differences seem greater than was the case. For instance, LimnoTech estimated that the difference in the amount of nitrogen and sediment from agricultural lands between the two models was 28 percent and 67 percent respectively. But when errors in LimnoTech's interpretation were corrected, the differences were 15 percent and 29 percent, "which are within the likely margin of error" for agricultural land estimates for the two models, the report said.
Overall, the report said the differences in the loads attributed to agriculture and croplands between the two models was "small."
"More importantly," it said, "the results of the two models are similar in their assessment of the need for implementing more management practices on cropland."
The STAC report did say that the USDA's model included information that could improve the Bay Watershed Model and urged modelers from the two agencies to work together, something that is already happening.
It also said the EPA Bay Program should move to a modeling framework that uses multiple models, instead of relying on single models to produce results.
In a statement, the Agricultural Nutrients Policy Council said it would review the STAC report and hoped to meet with the review team, but contended that while STAC's wording was "harsh," its actual conclusions were consistent with corrective actions recommended by LimnoTech.
"We disagree with the STAC that these findings did not justify delaying the finalization of the TMDL," it said. It said it would issue more detailed comments after it had a chance to thoroughly review the report.
STAC is an independent advisory committee to the Bay Program that provides technical reviews to policies and programs and at times in the past has had its own criticism of the Bay Program's model. The review team included both members of STAC and outside modeling experts.
The STAC report, "Review of the LimnoTech Report 'Comparison of Load Estimates of Cultivated Cropland in the Chesapeake Bay Watershed'," is available at www.chesapeake.org/stac.
- Category: Politics + Policy
Comments are now closed for this article. Comments are accepted for 60 after publication.