## MBARI Offsite ## Day 1: Strategic Planning # ## Mon Mar 21, 2005 [bellingham] Introduction Objectives: -Building partnership btwn engr and sci -Work toward action list for influencing the proposal process -Work toward common vision of what technology is important, how they impact sci, how to pursue them, common language for ongoing dialog. balance of projects: -science pull -tech push -infrastructure feedback to mteam during proposal process - must push the envelope - must support science at MBARI ? what is good breakdown of pull, push, infra projects? science driven enterprise technology driven enterprise generating support might be harder. cant predict how it will be used by sci, but need to define how it provides new capabilities. Economics typically not a strong argument, but if vague or unfavorable, will be a strong argument against. defining the project: func reqs vague, must be careful of scientist steering who is vocal but not representative. Early adopter may not be representative of eventual users (that's ok) (early adopters have $, willing to accept risks) doing the project: all usual challenges, plus skepticism or at least no buy in. anything that generates science buy-in will increase chance of success and decrease pain. must be vigilant for clues to key needs and functional reqs. Must make effort to entrain and retain users. infrastructure activities: ideas origin in egr, dmo, sci, itd. support: enhancement of efficiency, etc. defining project: usually clear what is needed. often hard for mgmt to see impact: eocnomy of scale also, uaully takes a lot of resources (again makes hard for mgmt to get on board) Institutional Process: preparing for 2005-2006 decisions mid year: retreat of sci, sel engr for discussion of future dirs. wide engr involvement encouraged. proposals 2005: exploratory pjts probably proposals 2006: build-up for new thrusts begin [chavez] science themes: -global ph change, carbon, Nitrogen cycle, pH -Harmful Algal bloom -who is there, what they are doing, how they are surviving (or not) and changing -horizontal transport to the seafloor [paull] unifying theme that cuts across mbari science themes, which are very wide compared to other places -gas hydrates -sea floor venting still trying to understand importance of this is in global processes. chemistry and possibly atmosphere determined by these processes. still not well understood. chavez,johnson, ryan, scholin : upper ocean chemistry haddock, robison: midwater barry, brewer, clague, paull, vrijenhoek sun, wind, rotation->mixing->nutrients, exchange w/ atmosphere, primary prduction topography -> circulation, distribution venting->benthic animals *ocean observatory initiative science drivers ocean circulation, climate variability, biogeochem cycling ecosystem dynamics turbulent mixing, and biophys interactions earth struct and dynamics of lithosphere Role of iron in the ocean sofex, ironex,bfex,soiree,sage,seeds,feep,eisenex industrialists have idea that it would be cheapest way to reduce CO2 in atmosphere ocean chemistry of greenhouse gases CO2 in the deep sea Benthic biology and ecology Submarine volcanism: volcanos may produce a lot of CO2 Continental Margin Processes [paull] Canyon Dynamics ESP next gen [scholin] AUV-mounted bioluminescence detector SCOPE robison: Trophic: refers to creation of food; levels of food chain Mauna Loa CO2 curve: CO2 and global warming: 6-7 GTon of CO2 from fossil fuel annually. ocean absorbs 2 GTon. Still a lot of discussion about what affect this will have on temp, how, and what effects would be (e.g, temp change, ocean pH etc) How do we measure Global Change? Natural changes to climate is similar or greater than manmade perturbations. we disagree over the sources of variability, they all agree on... why are long time series so rare? they get boring. they are expensive. ocean science is very underfunded and that is unlikely to change. mbari growth steady at 225 -- resources per capita is decreasing (flat resources over last 6 yrs). we have lots of things operating in the ocean...How to invest our resources what are a good number of projects [oreilly] open to interpretation, avoid overlap (e.g., instrument controllers) and duplication of efforts. Getting together more (sci, ops, engr) to decide HOW to do projects. [roman]is 25% soft $ a good thing? $$ is drying up...it will be hard to maintain our 25% [chavez] harbor branch had complete control of charter, finances. Their endowment shrank. at 40% they started to lose control of destiny. There is money, but it it's from military, oil.[bellingham] $ for raises? research $ decrease in a flat funding environment. [software and data systems] [graybeal] science inputs overlap bwtn science interests integration and cross cutting between projects video repository, sample archiving, benthic mapping featured strongly Community Needs: Visualization/integration (lot's of feedback): QC quality control: inputs hard qc requires domain expertise, is hard/impossible flagging data, not deleting qc of operations important first step -- IS there data? Visualization/integration : LOTS of feeback many different "needs", values visualization: different scales, seeing data in context, integrative data organization: putting together different categories of data * The tools we have don't give full functions, aren't as easy to use as they need to be. Tools and applications: Many domain specific requests Real-time event response: real time: would be nice hard to envision revolutionary tools Event Response important in some cases: first importance is interaction with deployed systems Need to have a process for determining what's important: who makes the process? Is there a lack of interest in experiment control? Control, NO, autonomous response, there are more important things to do first. SDS expertise is less in real-time systems, adaptive sampling, event response. [au] some parts may occur in data systems [gomes] this is an area where platforms may overlap with sds [davis] SWOT [gomes] Possible Directions [gomes] Issues single roadmap:sum of 3 group roadmaps Systems Engineering Group: inforcing interfaces between groups [chavez] SIAM influenced by external communities, and is disconnected from internal MBARI needs. we have to be careful about listening too much to external interests. [oreilly] we did go through an engineering process for a year...how could we do this more effectively? we did a lot of unecessary thrashing. also, we weren't focused on big picture, but little things. The point is: We could benefit from consistent, long term, systems engineering effort. DMO buy in: how to get this? Community interfaces: not likely that community would/could define such interfaces LOOKING, Grid making things that might work. [paull] VARS: example project > 300 ROV dives/yr > 11000 hrs of video obsevations in 16 yrs > 1.2 Million individual observations a picture worth a thousand words, if you can find it VARS -Knowledge base -Annotation Frame capture (QuickTime for Java) -Query Search and retrieve frame grabs Cold Seep (bad term) seep implies fluid flow chemosynthetic communities - energy from chemicals, not sun Chemosynth Bio Communities called cold seeps map of these communities in MBay c. 1999 was not good paradigm: seeps imply fluid flow faults are fluid conduits widespread fluid venting assoc w/ margin difficult to get useful data from ROV dive data (shotgun)...had to do a lot of filtering, correlation with We have observed only a tiny fraction of MBay seafloor. Difficult to know what we've missed, but make some assumptions about things being obvious, less likely to be missed did find that 'seep' communities DON'T happen near faults. May prefer steep slopes (implies recent erosion) Conclusions: -they are common, not rare -don't occur where fluid flow paradigm predicted -would have been almost impossible to learn this w/o sustained effort also it was a large effort, first real test of database. required bridging to another data set getting vehicle positions much more difficult than getting vehicle positions we've had several embarassing incidents where we didn't know we were collecting garbage data. two issues: data logging is broken (is there data, is it the right data, undetected, wrong lat/lon), real time qc (is the data good) third issue: infrastructure fixes, vs quick fixes to fix specific problems. what's being punched in daily is weak link probably. what is the quality of the data and lexicons of data that is being punched in hour by hour. how important is the video in this (vs annotation). its the initial thing that is looked at. there are many annotations w/o frame grabs: requires looking at lots of tapes manually to find relevant. we are an organization that should be able to do tech push. w/o tech push, there must be sci pull. where that point is varies. Engineering should be more vocal about deciding when to cut off projects. *what is level of sci commitment? the question is: how much of your next 5 yrs are you (sci) willing to commit to an engr idea? the question is not: is this a good idea? There is some assumption that there is at least a little sci pull... mteam doesn't see it that way. they see that there is sci pull, or can see that it may occur. proposal process: does proposal process favor taking risks, is it transparent,how are science themes communicated to proposal proc? There needs to be more communication into the process. Not clear (transparent) how internals of proposal selection works. It's hard to manage hi risk and long time scale projects w/o damaging What projects are successful? how is that measured? successful projects seem to be grass roots efforts [chavez] * a review of projects would be good onsite engr/sci would be good (15 min ea summary of crazy ideas, emerging trends, new tech) needs to be followed on with substantive conversations on ones that are interesting Time for this is hard to find...must be positive feedback that will make sure that activities are selected. Project updates already not well attended. would preview be helpful? does cookie break timing help?