Can NMFS Develop Acoustic Effects Guidelines Without Models?
"None of the models currently under development are yet capable of prediction,
nor have they been corroborated with field data."
The quote above is from the Draft Marine Mammal Advisory Committee Report on Acoustic Impacts on Marine Mammals ("Draft Report").1
The final MMAC Report needs to state whether the National Marine Fisheries Service ("NMFS") can develop, apply and enforce national guidelines on acoustic effects on marine mammals when there are no accurate, reliable models for those effects.
If the answer is yes,' then MMAC should explain how to achieve this goal.
The MMAC is a FACA advisory committee that is statutorily mandated under the Omnibus Appropriations Act of 2003 (Public Law 108-7) to "fund an international conference or series of conferences to share findings, survey acoustic 'threats' to marine mammals, and develop means of reducing those threats while maintaining the oceans as a global highway of international commerce." 2
The MMAC's final Draft Report will present findings and recommendations to the public and to NMFS. It is expected that NMFS will use the final MMAC Report during its development of national guidelines for acoustic effects on marine mammals.3
Computer models are often used in developing federal agency risk assessments such as the one NMFS is performing. According to the Draft Report, there are currently no accurate, reliable models for assessing acoustic effects on marine mammals:
Scientists can develop models to describe and predict various phenomena, including biological and environmental systems (e.g., responses of marine mammals and their habitat to certain events) and their interactions with human activities (e.g., sound production and its impacts). By combining such models, it may be possible under some circumstances to predict the likely exposure of an animal or population to a sound and, if enough behavioral information under the full range of possible conditions is available, also the probable response of the marine mammals. However, data needed to create and validate such models are lacking for marine mammal impact analysis, and the ecological variables involved are complex, making predictive modeling more difficult and prone to a low probability of accuracy, and therefore present challenges to managers. 4
According to the Draft Report, two models that might be useful in an acoustic effects risk assessment cannot be used yet. These two models are the Acoustic Integration Model ("AIM") and the Effects of Sound on the Marine Environment ("ESME"). The Draft Report states that AIM and ESME, "like other models, have limited data on which to predict the response of marine mammals to sound exposure. Both of these models are still under development and have not been field tested or validated." 5
Another potential problem with the AIM model is that its software is apparently proprietary. Proprietary software means that stakeholders would be unable to check the model results themselves if NMFS uses the AIM model in its development of acoustic guidelines.
NMFS stated in during the San Francisco public meeting on its development of acoustic guidelines that the Agency will put online and make publicly available everything the public needs to use the AIM model.
The Center for Regulatory Effectiveness has an opinion about what NMFS must do before it uses any model in developing, applying, and enforcing acoustic guidelines for marine mammals. CRE has stated that opinion to NMFS and to other agencies, and will continue to do so. Rather that reiterate that opinion here, CRE refers the reader to the original documents.6
At this point, CRE believes that the threshold question is whether NMFS develop, apply and enforce national acoustic guidelines for marine mammals when there are no computer models that are sufficiently accurate, reliable and transparent to use.
The Draft Report does not appear to answer this question directly. It does, however,
recommend that the NMFS risk assessment underlying the acoustic guidelines should meet the following criteria:
"1) explicitly discuss the limitations of using any predictive risk models;
The MMAC concludes in its Draft Report that none of the known models has "been suitably evaluated through field verification and by outside experts."
2) explicitly discuss and evaluate underlying assumptions and their consequences;
3) be suitably evaluated through field verification and by outside experts; and
4) make all elements, including models, publicly available." 7
Assuming the Draft Report's conclusion is correct, then MMAC's final report should address the essential question whether NMFS can develop, apply and enforce national guidelines on acoustic effects on marine mammals without any reliable or accurate models? If yes, then how?
Click here for Draft final MMAC Report
Click here for MMAC April 19- 21 meeting agenda
Click here for CRE Comments to NMFS on L-DEO sound propagation model (model comments begin on page 5)
Click here for NMFS's response to CRE's comments on models
Click here for other comments on Models
1The full Report is available at http://mmc.gov/sound/plenary5/pdf/reportforplen5discussion.pdf.
The quote occurs at Draft Report, Management of Acoustic Impacts on Marine Mammals ("Management"), Section IV. B.2.
2The quoted passage is from the MMAC website.
3More information on NMFS's development of national marine mammal acoustic effects guidelines is available at the following link: http://thecre.com/theipd/regactions_nmfs.htm
4Draft Report, Management, Section IV.A.2.d.
5Draft Report, Section IV.B.2.
5CRE's comments to NMFS regarding the requirements for agencies's use of models can be accessed at the following link: http://thecre.com/pdf/NOAA/CutPages/CRE.1.2.7-9.pdf.
Appendices A, B and C, which are referenced in CRE's comments, can be accessed at the following link: http://thecre.com/forums/showthread.php?s=f4d1d4f70f256f892bdd694d79492b3b&threadid=1875
5Draft Report, Section IV.B.5.