Downloaded 23 times
Jared Wilson - Overview of survey, data collection and data analysis challenges
- 1. Overview of survey,data collection and data analysis challenges Jared Wilson Marine Ornithologist
- 2. • Survey Design •Data Collection • Data analysis Overview of issues
- 3. Information available fromrange of sources: Survey Design • MMO (2014). Review of post-consent offshore wind farm monitoring data associated with licence conditions. • Oedekoven, et al (2013). Statistical Modelling of Seabird and Cetacean data: Literature Review • SNH (2011). Draft Guidance on Survey and Monitoring in Relation to Marine Renewables Deployments in Scotland. Under revision? • Camphuysen, et al (2004) Towards standardised seabirds at sea census techniques in connection with environmental impact assessments for offshore wind farms in the U.K.: a comparison of ship and aerial sampling methods for marine birds, and their applicability to offshore wind farm assessments.
- 4. ESAS methods Kober, etal. 2010. JNCC Report 431. Survey Design
- 5.
- 6. Offshore wind issuesrelevant to ‘wets’: Survey Design • Scale - appropriate to variability of metric • Effort - driven by the question, variability of metric, magnitude of change of interest • Effort - within development area often insufficient • Monitoring- to tie responses to specific causes • Transect spacing – default may not be appropriate • No. surveys - unlikely to allow a change to be detected • Control sites - unlikely to be appropriate • Imperfect detection - often ignored • Best practice - depends on species, site, metric, baseline variability, the question.
- 7. Issues apparent from‘wet’ renewables projects Survey Design
- 8. Survey Design Sufficient effortin development site? How much effort is within lease area? Let alone the likely development footprint?
- 9. Survey Design Is bufferrelevant to lease area? How similar are environmental conditions across the survey area?
- 10. A B E D F C X Survey Design Post-hocsubdivision of survey area How much effort in each sub-division How does coverage probability and effort vary across sub-divisions?
- 11. Survey Design Also: •VP unitof effort often unclear/ poorly defined (not snap shot so flux measurement?) •Detectability decreases with distance from observer but often ignored. •For VPs, distance from shore confounds increasing distance from observer but often (always?) ignored. •Question being addressed and metric to be recorded often unclear. Issues apparent from ‘wet’ renewables projects
- 12. Data Collection • VPsurvey effort poorly/ inconsistently recorded
- 13. Data Collection • VPsurvey effort poorly/ inconsistently recorded • Survey effort confounded? What is true effort?
- 14. Data Collection • VPsurvey effort poorly/ inconsistently recorded • Survey effort confounded? What is true effort?
- 15. Data Collection • VPsurvey effort poorly/ inconsistently recorded • Survey effort confounded • Observers unclear on reasons for recording various aspects of observations (especially effort).
- 16.
- 17. Detection probability What istrue distribution? Data Analysis • Imperfect detection
- 18. Detection probability What istrue distribution? Data Analysis Confounding effect of increasing distance from coast • Imperfect detection
- 19. Data Analysis • Imperfectdetection • Uncertainty not presented • Unclear whether survey effort appropriate (no power analysis) • Post-hoc subdivision of survey area/effort • Lack of clarity over analyses methods used • Diagnostic outputs not presented so impossible to come to view on methods used (including Distance analysis)
- 20. MSS identified needfor: Data Analysis • Review of available methods • Guidance on most appropriate statistical analysis of bird and marine mammal data routinely collected • Freely available software to allow appropriate analysis
- 21. Clear need forguidance identified Data Analysis MRSea •Contract awarded to CREEM at University of St Andrews in 2013. •Completed August 2013 •Workshop held September 2013 •Methods review, Guidance and software package freely available to download: http://www.scotland.gov.uk/Topics/marine/marineenergy/Research/SB9
- 22. Summary • Range ofexperience to inform survey design, data collection and analysis. • Increasing body of guidance exists • Wave & Tidal Stream Energy sites have distinct challenges • Methods are evolving (and need to continue to do so)
- 23. Any Questions? Jared Wilson MarineOrnithologist [email protected]
Editor's Notes
- #3 Very brief overview of issues relating to survey design, data collection and data analysis of particular relevance to wave and tidal stream. Some issues have been identified by experience with offshore wind projects, Other issues identified over last few years during dealings with wave and tidal projects.
- #4 Guidance documents for offshore wind and wave & tidal available for some time. More recently several reviews undertaken. Whilst experience is greatest for offshore wind focus, also some valuable lessons for wave and tidal.
- #5 Line transect methodologies originated in Seabirds At Sea. Aim was to ID seabird areas sensitive to oil spills. However, survey design was not a dominant feature. Vessels of opportunity, as can be seen by dominance of e.g. ferry routes or areas of greatest O&G shipping activity.
- #6 COWRIE guidance provided methodology for offshore wind transect surveys. Relatively large sites within large areas of open water. Guidance did highlight need for appropriate scales and survey effort. Methods been applied over range of offshore wind projects (and to some extent wave and tidal)
- #7 Several reviews undertaken recently. Whilst offshore wind focus, some important lessons for wave and tidal also. “One size fits all” is not the case.
- #8 Also several issues encountered with wave and tidal projects perhaps less relevant to offshore wind
- #9 Application of ‘standard’ design may result in limited effort within the actual area of interest
- #10 For complex environments rather than large areas of open water, are buffers appropriate for providing wider context/ comparisons with lease area? Lease area is swamped by buffer. Wet renewables sites often selected due to very local conditions- is this reflected in the buffers? Does this cause issues for both design and model-based density estimates? Is comparison between lease area and buffer meaningful/ helpful? Would it be reasonable to apply a density estimate derived from the entire survey area to the lease or development site only? Has sufficient within lease area variability been captured to produce meaningful estimate? Lack of suitable effort may be compounded by tidal cycle. If you are interested in the lease area, design your survey accordingly.
- #11 Post-hoc subdivision of survey area. Appropriate effort or coverage probability in each sub-division? If you are likely to want to compare distributions/ abundance across the site, consider appropriate survey design OR analysis methods.
- #12 Several issues have also been encountered for wave and tidal projects that may be less relevant to offshore wind
- #13 Several issues have also been encountered for wave and tidal projects that may be less relevant to offshore wind
- #14 Movement of a vessel into the tide may result in increased encounter rates. But effort (km of transect surveyed) does not account for this.
- #15 Movement of a vessel with the tide may result in reduced encounter rates. But effort (km of transect surveyed) is the same as if vessel travels against the tide. Surveys across the tide may complicate this further. Does it matter? It does if you are wanting to compare seasons, years, sites or pre/post construction. It does if you want robust absolute abundance estimates to allow effect to be quantified e.g. collision risk modelling
- #16 It is difficult for surveyors/ observers to record data effectively if they do not understand the importance of what the are recording. Make life as easy as possible for observers- think about them when designing surveys and recording protocols.
- #17 The final topic.
- #18 VP surveys often ignore the issues of declining detectability with increasing distance from observer.
- #19 For land based VPs, further complicated by compounding effect of increasing distance from coast. If these are ignored, can you say anything meaningful about distribution and abundance at the site?
- #20 Several issues have also been encountered for wave and tidal projects that may be less relevant to offshore wind
- #21 Clear need from advisors to ensure provision of robust advice and developers/practitioners to ensure best practice is followed.