April 12-13, 2017 Slides
Cesar Blanco, Rod Wittler, Shelly Hatleberg, Mike Urkov, Dick Pool, Michelle Workman, JD Wikert, Russel Perry, Michael Wright, Felipe Carrillo Matt Brown, Julie Zimmerman, Josh Israel, Chris Hammersmark, Mark Thompkins, Chuck Hansen, Li-Ming He, Dan Kratville, Rene Henery, Jim Peterson, Adam Duarte,
Jim started by reviewing the SIT prioritization process. During the SDM process, he is here to help the group simplify the problem using available SDM tools. But we can only use those tools if SIT is willing to use them. Jim discussed the current model, the number of parameters, and how the key uncertainties (juvenile survival) did not change from the original model because we never improved our knowledge of juvenile survival. Jim suggested we try to get at estimating juvenile survival before SIT invests in restructuring the current model. We have a functioning model, but Jim suggests we simplify it to make it more manageable. Jim focused on better classifying means and fundamental objectives. Jim discussed to use of calculating the expected value of information, which is an SDM tool to evaluate uncertainties and the expected return (i.e., increase in returning adults, etc.) of reducing that uncertainty. Jim provided an overview of the use of policy plots to create a set of strategies and lookup tables.
Dick would like to improve the transparency of the process, so that he can communicate to the salmon industry what and how the CVPIA is doing.
The meeting was then focused on reviewing the SIT guidance document. The document was sent out for comments, but only Michelle and Dan responded. Julie mentioned that the document was meant to be a living document. JD said that he read it and thought it seemed reasonable. It was mentioned that the "developing a communication strategy" never happened. It was envisioned USFWS was going to hire a full-time science coordinator and that would be one of the responsibilities of that position. It was asked if SIT wants to keep it. Rod said he would also like to hear the Core Team's thoughts too, but he doesn't want it to take time away from priorities. Rene mentioned that the model complexity should not be included in the outreach, but that, based on Dick's comments, a bridge of communication is important. Michelle asked if there is a place to give a summary of the annual activity and also mentioned the SIT meetings are open to those interested in attending. Rod mentioned that maybe we should give a half hour summary each meeting. JD asked about if it is easier to write a memo for that. Rod wonders who would write it. Josh thinks we need a communication/collaboration strategy with the Core Team. He doesn't think communication to the public is the SIT's priority. He thinks it is the CVPIA's priority. Cesar mentioned it might be a simple as putting a quarterly bullet list of what the SIT has achieved to the email list serve he uses for the priorities. Josh liked that. Julie mentioned that it was supposed to work that way, since the science coordinator would be a CVPIA employee. It was proposed that we have 2 scheduled meetings with the Core Team each year. Dick thinks the fall meeting is a good place to communicate with the public. He would like to see actions linked to abundance results at the meeting. Rene mentioned that we might reframe it to communicate what the model says we should do, the actual costs of those projects, and show the current budget that we actually have to make those projects happen. Dick agreed that would be better. Rod mentioned that the CVPIA produces the tech memo, work plan, and CVPIA accomplishments. We are meeting with the Core Team today to discuss the tech memo. Heather would like to modify the information distributed during the Core Team summer meeting to include more information and that might include the CVPIA accomplishments/returns. Rene would like to help the Core Team frame those accomplishments based on what SIT members hear from stakeholders, the model, and the current budget. Rod discussed the importance of the SIT finishing the adaptive management loop, so that we are gathering the data/information to better inform our efforts. Dick asked if we are making progress for a science coordinator. The government hiring freeze has hindered that, but Shelly has been filling some of the coordination side of things. In the interest of time, Jim and Adam will draft some bullet points on how the Core Team can communicate to the outside world and SIT (particularly Rene and Dick) can provide feedback on that. Michelle would like to see the tech memo as it is sent to the Core Team and would like to provide feedback (possibly beforehand) or the guidance document should better reflect that process. Dan has the same comment. Julie mentioned that the tech memo has also changed from the initial process. Rod mentioned FACA and how the SIT must deal with this as it relates to the tech memo. Dan wants it clear in the tech memo that the SIT did not write it and it should be made clear that the tech memo is not the direct view of the SIT's members but the interpretation of the science coordinator. Rod would like some language with how the tech memo relates (or does not relate) to the 5-year plan. Julie mentioned the tech memo is the annual evaluation but the call for charters would be based on the 5-year plan, ideally. But we have to get to a point that we are comfortable with the model(s). The priorities each year would be an expression of the 5-year plan. The 5-year plan would be the overall objectives/goals and strategies. The tech memo would be the model document with a set of strategies. Rod would like the first draft of the 5-year plan by June of 2018. For the SIT responsibilities, Josh would like to add an attendance responsibility. Dan would like the amount of workload be expressed to the cooperating agencies or management groups. He thinks the workload is more than what is outlined in the guidance document. Rene mentioned that the workload really deters NGOs, particularly smaller ones, to participate because of the time costs. Rod would like the NGOs to share their thoughts on that via email so that he can digest and relay the information. There was discussion of the proposal section of the guidelines. It was discussed that the proposal process is beneficial, as long as the group can agree on the conceptual model or the structure of the quantitative model first. Jim will flesh out the bullets for the proposal process and will send out to SIT for comments.
Dave Zezulak showed up during lunch.
Core Team Participants:
Jim Smith, Kevin Reese, Dave Zezulak, Heather Macintyre, Heather Casillas joined at 340, Brian Ellrott (will join shortly if can) – couldn't make it, Kim Webb joined at 340
Levi Johnson CVPIA project manager Reclamation
Dave Van Brian
Rene thinks juvenile survival is important because of the way the model was built. That is, we are "only modeling what is leaving". He also thinks the monthly time step is limiting. Rene likes the conceptual model but thinks the quantification of the model biases the results. Until we agree on that, he thinks we cannot apply tools like policy plots.
Cesar stated the Core Team wanted less priorities 2 years ago but the Core Team thought the SIT provided too few priorities this last cycle. Cesar would like more time for the Core Team to provide comments on SIT priorities this year.
Core Team feedback on SIT priorities
Jim Smith had difficult time understanding how the winter run PWT priorities translated over to the SIT priorities. The tech memo helped with this. He gathers that the SIT team was asked to crunch down the priorities into smaller bits and not as broad as what the winter run PWT provided. Jim Smith advocated to add the winter run PWT priorities in addition to the SIT priorities. The winter run PWT was not asked to focus vs. making it broad and he does not think that was fair to the winter run PWT. He also thinks the SIT is not as knowledgeable as the winter run PWT on winter run. He thought the fall run followed the requested process, but not the winter run. He also knows the Sturgeon and Steelhead were more involved with the SIT, so he agreed with those. He had a different view with respect to the winter run. Jim Peterson asked if we developed a winter run DSM who needs to be there for their expertise. Jim Smith said he and Matt sent out a list that they both worked on. Specific names included: Doug Killam, himself, Brian Portz, Dan Kratville
Shelly Hatleberg, Rod Wittler, Mike Urkov, Russel Perry, Michelle Workman, Chuck Hansen, JD Wikert, Cesar Blanco, Dan Kratville, Chris Hammersmark, Josh Israel, Michael Wright, Matt Brown, Julie Zimmerman, Jim Peterson, Adam Duarte, Felipe Carrillo, Rene Henery
Rod discussed the potential for a Bay Delta collaboration. Chuck gave a, presentation on the Salmon Scoping Team Gap Analysis Report. The report is available to SIT members, if they would like. The group discussed the need for outmigrant survival estimates from wild fish and frys as well as longer term studies. There is also a need for rearing survival estimates.
Does SIT want access to the DSM and if they do, what format do you want it in? Jim showed an example of a probabilistic network or graphical model in Netica. Jim discussed the difference in decision theoretic approach and the experimental decision approach. Using a theoretic approach, we can have competing models on how the system works. SIT members would like access to the model in a user-friendly format as well as access to the code. Mike Urkov discussed creating a Shiny app to do this but that it would help him if he knew what output people were interested in seeing. It was also discussed that there needs to be a system in place for version control. Julie mentioned how this was originally supposed to be handled by the Center for Data Management. Cesar is concerned that if it is made available to folks outside the SIT that it may be used for purposes outside what it was built for. Russel suggested that outside use of the model could be on a case-by-case basis so that we are aware of the alternative uses. Rod mentioned that if you need help from Mike U. and Mike W. for CVPIA needs that you should definitely use them as a resource. Rod also thinks we need to revisit the role of Center for Data Management. Cesar was also concerned with outside users using the model for publishing. That is, people who built the model should have first crack at using the model for publishing purposes. Rod was concerned with the restriction of use. Jim mentioned that we need to start organizing a place that SIT members can access the data that is going into the model and provide version control. Rene would like the data to be easy to query and match up with the different submodels in the current DSM as well as find a way to make the data accessible maybe before we have the perfect data structure for it. Mike Urkov discussed some options in R that they have been looking at so that it is easy to incorporate within the model and improve version control. As far as access, Julie brought up that those data are already out there. Michelle mentioned that it should probably be a case-by-case basis by the CVPIA managers. Julie mentioned that we also need to be concerned with version control. Julie wants it to be clear that once the model is modified it is not referred to as "the SIT model". Mike Urkov stressed that that is the distinction they are using. Matt mentioned that we should probably not distribute more than it already is till we are happy with it. Rod mentioned that there are different pieces, including the source code, the executable code, the data, etc. Mike Urkov asked about using the model and sharing the model to help people understand it when making charters. Felipe mentioned the use of creating logs/notes on the steps used to build the model so that people can see how/what was done without modifying the base code. Jim asked that SIT members send an email to him, Shelly, Rod and Cesar if they have more thoughts on this. There is a meeting on this on Monday at 230-330PM. Rod said the meeting is open to those interested.
The next discussion focused on fall run DSM issue that we had tabled before. Rod discussed the lighting project that is focused on anthropogenic light influencing prey and predator behavior. Dan discussed the uncertainty associated with saying predation is a problem for fish populations. He wants to make sure it is clear that we are evaluating IF predation is a problem. We are not going into this assuming predation is a problem. Russel mentioned the use of an experimental approach to estimate the influence of lighting on survival to disentangle different processes. Rod also discussed some other charters that are coming that are centered on predation. Rod asked if the SIT members would benefit from seeing charters and providing feedback. It was discussed that we need to be careful that folks who are on the charters are not reviewing them. It was also discussed that the SIT is not proposing projects, so that could be a problem. Michelle discussed how that is the responsibility of the project manager, not the SIT – which multiple SIT members agreed with. JD discussed some approaches he has been involved with in other efforts. Rene thinks for next cycles the SIT could help by defining some of these objectives more clearly in the priorities to help make sure charters are fulfilling the SIT's needs. It was discussed that the SIT was originally meant to look at the larger picture and if a watershed was identified then a team centered on that watershed would use their local knowledge (and possibly their higher resolution model) to identify what area or what method is best for that and submit that in the charter. The local experts would also be expected to estimate/quantify the benefit of their project within the charter. Jim asked for SIT members to identify a definition we can all agree on for what is a floodplain for next meeting. Rene offered to make some current projects/papers available to the SIT that is centered on classifying floodplain. Josh will also send out a paper concerning what a floodplain is that was done in the Central Valley.
END OF MEETING