Ron Sandland recently wrote about the new phenomenon of 'big data' - weighing up the benefits and concerns. Terry Speed reflected on the same issue in a talk earlier this year inGothenburg, Sweeden noting that this is nothing new to statisticians. So what's all the fuss about? Here's another take on the 'big data' bandwagon.
Turbidity Monitoring clouded by dubious science February 17, 2014
Regulators and industry around the country are using a potentially flawed method to set environmental limits on water column turbidity.
How do you set a 'safe' turbidity level so as to have high confidence that marine flora and fauna will not be compromised during dredging and oil and gas exploration activities? Conventional wisdom is to determine a high-order percentile from 'background' turbidity data and use that as the threshold. Another approach uses the (related) concepts of 'intensity', 'duration', and 'frequency' as espoused by U.S. EPA workers Christopher McArthur and Roland Ferry together with John Proni (formerly of NOAA). Their approach was outlined in a conference paper delivered at a regional dredging conference in Florida in 2002 (McArthur et al. 2002). As far as we are aware, the work has not been published in a peer-reviewed journal, yet in Australia consultants are promoting it; regulators are relying on it; and industry is using it. This wouldn't be such a bad thing if the methodology was sound and had at least been 'road-tested' by the authors. Trouble is - it hasn't.
Back in 2011, Environmetrics Australia was asked by the LNG industry to provide comment on the McArthur et al. (2002) approach. On the basis of that initial investigation we concluded "there is sufficient doubt about the veracity of assumptions, application of statistical methodology, and demonstration of success to warrant deferring the implementation of a turbidity monitoring program that relies in whole or in part on the methods outlined in McArthur et al. (2002)". Before publishing this recommendation we contacted Christopher McArthur to seek clarification on a number of technical issues. In his reply, McArthur noted "we are very curious as to how this is being applied as we never actually ended up applying it here". This was also confirmed in a separate email from co-worker Roland Perry (received 3/11/2011) who stated "as I am the guy who came up with this concept and developed the statistical basis for it you might think I'd have a ready answer but Chris may have mentioned there was little interest in this concept here and I haven't thought much about it in the last 10 years".
So, here we have a situation where an untried, untested, and un-refereed piece of work that was put together for a conference in the United States 12 years ago and which has never seen the light of day since, is enshrined in environmental permits in this country.Despite our best attempts to alert industry and regulators about the short-comings of the McArthur et al. (2002) paper, their 'intensity-duration-frequency' approach continues to gain traction and momentum - particularly in Western Australia where it has been endorsed for use on large projects in the North-West Shelf.
We are concerned by this trend and strongly suggest state and federal environmental regulators cease relying on and/or accepting proposals based on McArthur et al. (2002) for setting turbidity limits for environmental protection purposes. Moreover, recent Australian projects such as Gladstone Ports Western Basin Dredging Project have raised the bar in this respect. Environmetrics Australia has developed near real-time forecasting of water column turbidity and benthic light on a wide range of spatial and temporal scales (click herefor more information from Dredging Today).
Reference McArthur, C., Ferry, R., Proni, J. (2002)Development of guidelines for dredged material disposal based on abiotic determinants of coral reef community structure. Proc. Third Specialty Conference on dredging and dredged material disposal, Coasts, Oceans, Ports and Rivers Institute (COPRI) of ASCE, May 5, 2002, Orlando, FL USA.