search

nmfs-swfsc-ast / ast-tasks

0 stars 0 forks source link

Develop Logistical Plan-2307RL #23

Closed kstierhoff closed 1 year ago

kstierhoff commented 1 year ago

We'll use this task to develop the transect plan, identify ports of call, locate and schedule mid-leg transfers, etc.

The first iteration of checkTrawls is located on Google Drive, along with the first iteration of the Coastal Explorer file (2307RL.nob). I typically create copies of the .nob file before making any revisions (e.g., 2307RL_20221209.nob), to make it easier to revert to previous versions as needed.

alicebeittel commented 1 year ago

Checklist for myself:

kstierhoff commented 1 year ago

I'm placing the route_plan_fsv.csv file in Drive. I think it's fairly straightforward, but it doesn't appear to consider the survey.direction for computing the transect order or the day length.

FYI, this file gets created each time you run checkTransects, in: Output/routes.

I'm also placing a copy of the Sheet from the previous year.

alicebeittel commented 1 year ago

@kstierhoff Tried to run check Transects this morning but ran into a an issue in the chunk of code starting on line 49. 

# Get project name from directory
prj.name <- last(unlist(str_split(here(),"/")))

# Get all settings files
settings.files <- dir(here("Doc/settings"))

# Source survey settings file
prj.settings <- settings.files[str_detect(settings.files, paste0("settings_", prj.name, ".R"))]
source(here("Doc/settings", prj.settings))

image

alicebeittel commented 1 year ago

@kstierhoff Figured out my issue with line 49....wrong project open and file name, pulled new changes to 2307RL settings. New error I ran into is at line 60:

image

kstierhoff commented 1 year ago

Can you try pulling new changes from Github and see if it'll run? It's running here and I think I pushed the latest version to Github. If not, please point me to the latest .nob and/or .gpx file that you're using. It's failing somewhere in the processTransects.R script, I think.

alicebeittel commented 1 year ago

Looks like I've already got all the new changes from Github. C:\ABB\CODE\Github\estimATM\2307RL\Data\Nav\waypoints_2307RL.gpx

kstierhoff commented 1 year ago

It's possible that the code is not finding the GPX file, since it's named differently than what's defined (rosepoint_waypoints.gpx) in the settings file: image

kstierhoff commented 1 year ago

This error was a combination of 1) the NOAA bathy server being down and 2) a couple editing errors on waypoint names in the .nob file. I provided the bathy_data_2307RL.Rdata file to Alice to alleviate the problem with the missing bathy data. The waypoint errors were related to 210N, and should be fixed in the master .nob file.

kstierhoff commented 1 year ago

Didn't mean to close this issue...just update it. Reopening!

alicebeittel commented 1 year ago

@kstierhoff Preliminary time estimates are on the "platform coordination" tab in the route planning spreadsheet. See notes section on spreadsheet tab: https://docs.google.com/spreadsheets/d/1wYwtqfkxr-O3uU_gNCWZHiLYFmqbfgv1zDbT0ubcvz8/edit?pli=1#gid=1603117248

I have not yet looked at the wind farms or the hake survey mid-water trawl locations from 2019 (I have the lat/long data for the hake survey, just need to figure out how to plot it in either coastal or R).

kstierhoff commented 1 year ago

@alicebeittel, this is great. Could you take a few minutes to present this at the AST meeting today? A few things that come to mind with the way these plans have worked out so far:

Kevin

alicebeittel commented 1 year ago

To Do after AST meeting:

kstierhoff commented 1 year ago

@ddemer, would you like to respond to Isaac regarding shortening the transects to ~30 nmi off Baja? Or shall I?

kstierhoff commented 1 year ago

@alicebeittel, it looks like Isaac has agreed to ~30 nmi lines off Baja, but asked that they be as long as possible.

@ddemer, shall we make them 35 nmi long to be consistent with our minimum transect lengths N of the Border?

And should I offer Isaac and Roberto our transect lines off Baja, so that the Carranza transects interleave ours? Or allow them to develop their own plan? I'm not sure how far along they are with their planning.

alicebeittel commented 1 year ago

@kstierhoff @ddemer Updated the spreadsheet with some new estimates using SD + 30 nm MX lines and 20nm MX line spacing. https://docs.google.com/spreadsheets/d/1wYwtqfkxr-O3uU_gNCWZHiLYFmqbfgv1zDbT0ubcvz8/edit?pli=1#gid=1603117248

In short, the new MX line arrangement gives us a ~14 line gain for Leg 1 with survey work ending around SLO. I compared two scenarios going forward. 1) SD does even # lines from Cape Mendocino to SF Bay allows RL to arrive to Cape Mendocino to meet SH ~8/18. (neg) I think this will be too late and provide too much of a gap with SH given their preliminary port schedule (still waiting on survey plan). (pos) Using 3 SD at once gives us an 11 day max spread between RL and SD rather than a 17 day max spread using 2 SD. Starting SD 4 days earlier than when RL is expected to arrive to Cape Mendocino could reduce this to ~7-8 day spread. SD and RL would be working towards each other, passing off the coast of N. CA. (pos) SD use would be concentrated in an area where we saw mostly one species in past years.

2) SD does even # lines from Cape Mendocino to Pt Conception. (neg) There would be a larger temporal and spatial gap between SD and RL data acquisition, roughly ~2 week spread. (pos) RL gets to Cape Mendocino ~ 4 days earlier around 8/14 (not much)

I'm guessing having SD sample as close to RL (and trawls) is ideal.

Third Scenario: If RL did ONLY odd # lines in the SCB w/ SD working even # lines Cape Mendocino to Pt Conception, RL could get to Cape Mendocino as early as 8/7.

Fourth Scenario: If SD did all lines SF Bay to Pt Conception, frees up ~16 RL days (if RL were doing all lines). If SD did all lines Cape Mendocino to SF Ba, frees up ~13 RL days (if RL were doing all lines). Let's say 15 days for either - RL could get to Cape Mendocino around early Aug, similar timing to 3rd scenario.

Out of the box idea: SD does offshore portions of the lines or start SD further north of Cape Mendocino to allow RL to move up the coast faster.

alicebeittel commented 1 year ago

@kstierhoff I went to re-run the checkTransects with the edited nearshore lines, but ran into an error in the same section as before.

Image

I've added a tab for nearshore line estimates, the data currently in there is with the un-edited nearshore lines. Copied them into the spreadsheet before realizing they were still the un-edit version. A rough estimate with the nearshore un-edited lines is ~70 DAS.

alicebeittel commented 1 year ago

Nearshore code issue resolved. Remember to export waypoints, not route!

DAS estimate for all nearshore lines (islands and near coast) is still ~70.

alicebeittel commented 1 year ago

Updated survey proposal:

I think this gets us as close as possible to meeting SH. RL does 30nm MX lines at 20nm spacing + 20nm spacing through the SCB with lines shortened to 2022 CPS extent + 20nm RL spacing until meet up with SH (SD interweaves during this time).

Punta Baja to USA/MX Border:

USA/MX Border to Pt Conception

Pt Conception to as soon as RL meets SH

When RL meets SH (sometime leg 2)

9/13/2023 Hake Survey scheduled to end. RL continues north to do all lines, 10nm spacing off Vancouver Island (to complete Vancouver Island all lines is ~12 days of survey)

Couple notes:

https://docs.google.com/spreadsheets/d/1wYwtqfkxr-O3uU_gNCWZHiLYFmqbfgv1zDbT0ubcvz8/edit#gid=1603117248

ddemer commented 1 year ago

Alice,

Thanks, Please contact Julia Clemens for Shimada's cruise dates. She mentioned a schedule to me today, which seemed final?

Dave

On Mon, Jan 9, 2023 at 1:20 PM Alice Beittel @.***> wrote:

Updated survey proposal:

I think this gets us as close as possible to meeting SH. RL does 30nm MX lines at 20nm spacing + 20nm spacing through the SCB with lines shortened to 2022 CPS extent + 20nm RL spacing until meet up with SH (SD interweaves during this time).

Punta Baja to USA/MX Border:

RL 30nm lines, 20nm spacing BC interweaves every other line at 20nm spacing

USA/MX Border to Pt Conception

lines shortened to CPS extent of 2022 (avg line length 67 nm) RL does shortened lines, 20nm spacing Expected 7/18 RL arrives at Pt Conception

Pt Conception to as soon as RL meets SH

RL does full length lines, 20nm spacing SD interweaves every other line, 20nm spacing Expected 7/29 RL arrival to Cape Mendocino

When RL meets SH (sometime leg 2)

RL does full length lines, 10nm spacing in sync with SH if extra SD days, could use on offshore portions of lines if needed as confirmation no CPS

9/13/2023 Hake Survey scheduled to end. RL continues north to do all lines, 10nm spacing off Vancouver Island (to complete Vancouver Island all lines is ~12 days of survey)

Couple notes:

If RL meets SH by Newport, we can use the max # of SD days (180) and have RL do every line starting at Newport going North. Est max transects SD can do in 180 days w/ 20nm spacing = 30 We don't have final Hake survey dates (they are still trying to shift their inports)

https://docs.google.com/spreadsheets/d/1wYwtqfkxr-O3uU_gNCWZHiLYFmqbfgv1zDbT0ubcvz8/edit#gid=1603117248

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>

-- David Demer, Ph.D. Senior Scientist Leader, Advanced Survey Technologies Group Southwest Fisheries Science Center

alicebeittel commented 1 year ago

Looks like with the plan above, we could sync up with SH just south of Newport (~113, Umpqua River). There are variables to consider such as the potential of adaptively extending transects, wx, equipment or staffing issues, etc. I would say it is realistic that RL could meet up with SH at the start of their Aug 10th leg in the vicinity of Newport, OR. At this point, RL would switch to 10nm spacing proceeding northbound with SH. Having RL pick up every line at 113 would give SD 30 transect lines which is the estimated maximum for 180 SD days. Say everything goes to plan (ha), RL would finish survey ~9/15 (then add ~5 days for transit to SD), so we are working with about 10 days contingency. These 10 days could be eaten up if RL slows down to stay in sync with SH (see question 2).

Questions:

  1. Looks like Franklin will pick up transect lines for the Hake survey mid-WA through mid-Vancouver Island. Do we want to sync up with Franklin through this area and then sync back up with SH for the northern half of Vancouver?
  2. I need to confirm with NWFSC on their line spacing and Franklin dates, looks like their time estimates have them moving up the coast slower than my estimates for RL doing 10nm spacing and Franklin starts survey at line 77 before SH will arrive to line 76.

Backup plans could include:

alicebeittel commented 1 year ago

@kstierhoff @ddemer

Below is an image of what I've described above. Figured it would be easier to compare with the Hake survey plan with a similar image.

I confirmed with Julia - line spacing is 10nm throughout. Franklin will start survey before SH ends, so we could match up with Franklin by getting ahead of SH earlier. (SH estimated end leg 4 survey Aug 22 at line 76. Franklin starts survey at line 77 on Aug 19 and ends survey Aug 27 at line 88. SH starts leg 5 at line 89 Aug 28 and ends Sept 5 at line 102. Line #s reference Hake transects). If we stay with SH and Franklin and end the survey near Sept 15, we have about 10 survey days extra (add 5 days for transit from Vancouver to San Diego). We could use these 10 days to survey offshore portions of the shortened S. CA Bight lines. Alternatively, we've got 10 days contingency.

Before I send Julia the image -- Is this is plan we would like to propose?

Hake-CPS_SurveyPlan_13Jan.pdf

*Updated Image

alicebeittel commented 1 year ago

@kstierhoff @ddemer Updated nearshore survey plan for LM and LBC. See "Transect Compare" tab for DAS, start/end dates, proposed leg dates, and comparison between platforms: https://docs.google.com/spreadsheets/d/1wYwtqfkxr-O3uU_gNCWZHiLYFmqbfgv1zDbT0ubcvz8/edit#gid=1094284163

alicebeittel commented 1 year ago

An updated graphic: Hake-CPS_SurveyPlan_17Jan.pdf

kstierhoff commented 1 year ago

This looks good Alice, thanks for revising. Some thoughts:

I'll be free tomorrow to discuss further with you and Dave.

Kevin

alicebeittel commented 1 year ago

@ddemer A couple observations looking at the latest hake survey waypoints sent over by Julia last week:

  1. Hake survey transects are roughly 0.75-0.25 nm offset from ours due to differences in how we generate our transect lines. Is this an acceptable offset for the purposes of the data comparison this year? Steve de Blois mentioned this when we initially exchanged lines and it is still true with the latest waypoints. We did not come to a conclusion yet.
  2. Hake survey transects north of Cape Flattery continue in parallel with latitude lines, while ours are perpendicular to the shoreline. Are we ok with this difference in transect orientation for the purposes of data comparison this year? Both of these are food for thought for future joint-survey coordination and probably warrant a conversation down the road.

Saildrone proposed timeline: Once we decide what looks good, I'll send over this proposed plan to SD for their input. Since we have 3 SD, it might make sense to start two up north around the area we expect to begin overlap with SH/start RL on every line and 1 SD south. Here is what that could look like: Start 2 SD at line 112 near Reedsport, OR on 7/5, SD will have the closest geographical overlap with RL just south of Cape Mendocino and finish at SF entrance (line 74) around 8/15. Est. total DAS (both SD combined) at 1 kt survey speed is ~85 DAS. Start 1 SD at SF entrance line 72 on 7/10, SD will have the closest geographical overlap with RL around Monterey Bay and finish at Pt Conception (line 52) around 8/30. Est total DAS (1 SD) at 1 kt survey speed is ~52 DAS. Total SD DAS 137.

See 'SD_RL Coordination' tab in the google spreadsheet to see how the gap between SD and RL changes over these dates. https://docs.google.com/spreadsheets/d/1wYwtqfkxr-O3uU_gNCWZHiLYFmqbfgv1zDbT0ubcvz8/edit#gid=1444885606

Am I remembering correctly that our 180 SD days do NOT include transit time to survey locations? This would be 180 ON project days.

RE: RL SCB Lines With SCB lines extended to 75nmi, this adds about 15hrs of survey time for RL. Given that we will be adaptively sampling and the actual survey distance we cover may fluctuate, should we plan for a hard stop date of finishing the SCB on 7/19 to stay on schedule?

alicebeittel commented 1 year ago

1 SD at 1.4 kts avg speed could do about 18 transect lines spaced 20nm apart in about 60 days. An example of scale (Reedsport to Bodega Bay is ~ 18 lines 20nm apart). After talking with Kevin, another idea is to use the saildrones starting at a specific point and start each saildrone separately to survey the same area at different time stamps. In this case, we could compare the saildrone acoustic data at three different levels of temporal proximity to the RL data.

kstierhoff commented 1 year ago

Hi Alice, As we discussed, the ~30 d lag between sampling is probably unacceptable. After discussing with Dave and Juan, a process study looking at repeated sampling using Saildrones is probably not feasible this year. An alternative that we discussed was having one SD sample from SF to Pt. Conception, coincident with Lasker, then a second SD sample from Mendocino to SF, then a third from Cape Flattery to Mendocino, or something along those lines. This approach probably requires more thought too, but it seems possible. Kevin

On Mon, Jan 30, 2023 at 4:23 PM Alice Beittel @.***> wrote:

1 SD at 1.4 kts avg speed could do about 18 transect lines spaced 20nm apart in about 60 days. An example of scale (Reedsport to Bodega Bay is ~ 18 lines 20nm apart). After talking with Kevin, another idea is to use the saildrones starting at a specific point and start each saildrone separately to survey the same area at different time stamps. In this case, we could compare the saildrone acoustic data at three different levels of temporal proximity to the RL data.

— Reply to this email directly, view it on GitHub https://github.com/nmfs-swfsc-ast/ast-tasks/issues/23#issuecomment-1409562949, or unsubscribe https://github.com/notifications/unsubscribe-auth/AARG62R4GIICKSSDSXWDZGDWVBLOHANCNFSM6AAAAAASZUCSUA . You are receiving this because you were mentioned.Message ID: @.***>

--

Kevin L. Stierhoff

Research Fisheries Biologist, NOAA/NMFS/SWFSC

alicebeittel commented 1 year ago

@kstierhoff @ddemer If we split up the lines evenly between each SD the breakdown looks like this: SD 1: Pt Conception to N. of Bodega Bay (Salt Point), SD2: Salt Point to Coos Bay, SD 3: Coos Bay to Cape Flattery. I looked at what the spread would be between SD and RL if we did this scenario and the ~30 d lag is unavoidable at the start and end locations of the SD. Since SD is moving N to S and RL moving S to N there is about a 26-35 d lag at the start and end points for each SD. The lag is shorter in the southern region because the transects are shorter. IF we did this scenario, would RL still continue 10nm spacing, overlapping with SD lines from Umpqua River (T113) northward? We had chosen to bring RL back down to 10nm spacing in the vicinity of where we are closest with SH.

Spreadsheet Transect Compare Tab is updated with potential SD start and stop days that minimize the lag with RL as much as possible. https://docs.google.com/spreadsheets/d/1wYwtqfkxr-O3uU_gNCWZHiLYFmqbfgv1zDbT0ubcvz8/edit?pli=1#gid=1094284163

kstierhoff commented 1 year ago

It appears that working in opposite directions is the main problem, which will be difficult or impossible to overcome. Thinking I should reach out to Saildrone to see whether they'd be willing to sail from S to N along with RL They would only be slowed during the transits between transects, the orientation of the transects themselves would not change.

On Tue, Jan 31, 2023 at 10:55 AM Alice Beittel @.***> wrote:

@kstierhoff https://github.com/kstierhoff @ddemer https://github.com/ddemer If we split up the lines evenly between each SD the breakdown looks like this: SD 1: Pt Conception to N. of Bodega Bay (Salt Point), SD2: Salt Point to Coos Bay, SD 3: Coos Bay to Cape Flattery. I looked at what the spread would be between SD and RL if we did this scenario and the ~30 d lag is unavoidable at the start and end locations of the SD. Since SD is moving N to S and RL moving S to N there is about a 26-35 d lag at the start and end points for each SD. The lag is shorter in the southern region because the transects are shorter. IF we did this scenario, would RL still continue 10nm spacing, overlapping with SD lines from Umpqua River (T113) northward? We had chosen to bring RL back down to 10nm spacing in the vicinity of where we are closest with SH.

Spreadsheet Transect Compare Tab is updated with potential SD start and stop days that minimize the lag with RL as much as possible.

https://docs.google.com/spreadsheets/d/1wYwtqfkxr-O3uU_gNCWZHiLYFmqbfgv1zDbT0ubcvz8/edit?pli=1#gid=1094284163

— Reply to this email directly, view it on GitHub https://github.com/nmfs-swfsc-ast/ast-tasks/issues/23#issuecomment-1410903562, or unsubscribe https://github.com/notifications/unsubscribe-auth/AARG62RBTJPBAX6MCWMOC73WVFNZZANCNFSM6AAAAAASZUCSUA . You are receiving this because you were mentioned.Message ID: @.***>

--

Kevin L. Stierhoff

Research Fisheries Biologist, NOAA/NMFS/SWFSC

ddemer commented 1 year ago

Good plan. Hoping they agree.

They may have some experience working from S-N, as they have worked with the hake survey previously.

Dave

On Tue, Jan 31, 2023 at 12:20 PM Kevin Stierhoff @.***> wrote:

It appears that working in opposite directions is the main problem, which will be difficult or impossible to overcome. Thinking I should reach out to Saildrone to see whether they'd be willing to sail from S to N along with RL They would only be slowed during the transits between transects, the orientation of the transects themselves would not change.

On Tue, Jan 31, 2023 at 10:55 AM Alice Beittel @.***> wrote:

@kstierhoff https://github.com/kstierhoff @ddemer https://github.com/ddemer If we split up the lines evenly between each SD the breakdown looks like this: SD 1: Pt Conception to N. of Bodega Bay (Salt Point), SD2: Salt Point to Coos Bay, SD 3: Coos Bay to Cape Flattery. I looked at what the spread would be between SD and RL if we did this scenario and the ~30 d lag is unavoidable at the start and end locations of the SD. Since SD is moving N to S and RL moving S to N there is about a 26-35 d lag at the start and end points for each SD. The lag is shorter in the southern region because the transects are shorter. IF we did this scenario, would RL still continue 10nm spacing, overlapping with SD lines from Umpqua River (T113) northward? We had chosen to bring RL back down to 10nm spacing in the vicinity of where we are closest with SH.

Spreadsheet Transect Compare Tab is updated with potential SD start and stop days that minimize the lag with RL as much as possible.

https://docs.google.com/spreadsheets/d/1wYwtqfkxr-O3uU_gNCWZHiLYFmqbfgv1zDbT0ubcvz8/edit?pli=1#gid=1094284163

— Reply to this email directly, view it on GitHub < https://github.com/nmfs-swfsc-ast/ast-tasks/issues/23#issuecomment-1410903562 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AARG62RBTJPBAX6MCWMOC73WVFNZZANCNFSM6AAAAAASZUCSUA

. You are receiving this because you were mentioned.Message ID: @.***>

--

Kevin L. Stierhoff

Research Fisheries Biologist, NOAA/NMFS/SWFSC

— Reply to this email directly, view it on GitHub https://github.com/nmfs-swfsc-ast/ast-tasks/issues/23#issuecomment-1411011202, or unsubscribe https://github.com/notifications/unsubscribe-auth/A4DJS7O2YQVUKVKHTTL3DR3WVFXZ7ANCNFSM6AAAAAASZUCSUA . You are receiving this because you were mentioned.Message ID: @.***>

-- David Demer, Ph.D. Senior Scientist Leader, Advanced Survey Technologies Group Southwest Fisheries Science Center

alicebeittel commented 1 year ago

New Saildrone Proposal -- @ddemer @kstierhoff

Looked into some time estimates today and tinkering with the idea of saildrone "leap frogging". What do you think about "leap frogging" the saildrones up the coast? This would look like SD 1 starting at Pt Conception 5 days earlier that RL expected arrival (7/15), SD1 surveys the geographic span of 10 transects (which means it would actually survey only 5 transects, interweaving with RL transects). SD2 starts 5 days earlier than RL expected arrival at the next northern chunk of 10 transects. SD1 finishes, transits to SF. SD2 finishes outside SF Bay. Recover SD 1 and 2. Deploy SD3 5 days earlier than RL expected departure from SF inport. Then continue "leap frogging" the three SD up the coast at increments of 10 transect lines. Each started 5 days earlier than RL expected arrival. This would provide a max 7-10 day gap in data between RL survey lines and SD survey lines. It would probably take ~2.5 days for the SD to transit north to the next 10-transect chunk (~70mi), so we would need to chat with SD and figure out if that would come out of our DAS budget. This plan would allow us to retrieve data from SD 1 and 2 from everything south of SF mid-season.

Happy to jump on a call tmrw to explain this verbally. Check out 'SD_RL Coordination StoN' tab for est start and end dates for each sail drone. There are three evolutions of leap frogging. https://docs.google.com/spreadsheets/d/1wYwtqfkxr-O3uU_gNCWZHiLYFmqbfgv1zDbT0ubcvz8/edit#gid=333260667

kstierhoff commented 1 year ago

Hi @alicebeittel, thanks for putting this together. I'm intrigued by the leap-frog approach and would like to hear more about it. I'm free most of the day to discuss. If I had to guess, I'd say that SD would not count the days hove to waiting to start against our mission day count, but I'm guessing that the transits between 10-transect groups would count. They also do not typically count the days required to return from, say, Cape Flattery to Alameda. But for our planning, so we can assess how much progress we can expect to make, we should ask how they compute mission days.

alicebeittel commented 1 year ago

@kstierhoff Well it looks like my initial idea may have been too good to be true...glad I caught my error in the dates. Can't have one saildrone in two places! Went down quite the rabbit hole to see if there was a better combo of transect chunks out there, sometimes I guess you have to go in circles to realize the first option was the better one to begin with.

Transit time between each 10-transect chunk is about 6 days. Which seems like a lot, especially if those are coming out of our sea days budget. Transit time between 8-transect chunks (4 SD lines) is about 5 days, a bit better. Either way, the SD move so slow that even with leapfrogging up the coast, the more northern sites become farther and farther separated from when RL would survey the surrounding lines. Looking at a max 10 day spread in early southern lines, 27 day spread on line 130, and max 23 day spread on line140. I still think the leapfrogging is better overall and we get closer overall to the RL. Plus, realistically if RL is delayed at all, this will help push SD closer to match up with RL in the northern areas. In the end it probably doesn't make that big of a difference having two 6-day transits per SD vs three 5-day transits per SD.

Are we OK with a 6 day transit between each SD area?

Attached is an image. I've put a proposed start date of Sunday July 9, starting earlier allows us to minimize having larger gaps up north between RL and SD.

Saildrone Plan_22Feb.pdf

alicebeittel commented 1 year ago

If we want to pitch this plan to SD today before the meeting tmrw, here is a draft:

In order to minimize temporal gaps between Saildrone and Reuben Lasker data acquisition, we would like to propose a "leapfrogging" survey plan. In this scenario, the three Saildrones would each survey 5 Saildrone transects at a time and leapfrog up the coast starting at Point Conception, ending at Cape Flattery. Attached is an image illustrating this idea. Saildrone A starts at Pt. Conception surveys 5 transects (interwoven with Lasker transects) finishing at Big Sur where it will start a transit to SF for data offload. After data offload, Saildrone A will transit to Pt Arena to begin the next chunk of 5 Saildrone transects. Saildrone B starts at Big Sur surveys a chunk of 5 Saildrone transects ending in SF for data offload, then transits north to Cape Mendocino for another set of 5 transects. Saildrone C starts outside SF, surveys 5 transects, then transits north to the CA/OR border for the next set of 5 Saildrone transects. The leapfrogging continues up the coast.

The image includes some rough start time estimates for each Saildrone. We did our estimates assuming the Saildrone could make 1.4 kts for survey and transit, but are curious if you have any insights on expected Saildrone speeds heading northbound or ways to refine predicted progress of the Saildrones. We are open to hear your thoughts on this idea and if you think it might be possible.

kstierhoff commented 1 year ago

I think we could send this tomorrow am. Do you have the image? I didn't see it attached.

alicebeittel commented 1 year ago

Great! Yes. The image is attached to the post before.

alicebeittel commented 1 year ago

Updated image with last end dates and locations here: Saildrone Plan_24Feb.pdf

alicebeittel commented 1 year ago

Topics for discussion with SD team:

alicebeittel commented 1 year ago

Potential mid-leg transfer dates and locations. 2nd leg and 3rd leg transfer locations/dates may shift around pending Saildrone plan. @kstierhoff do you have any updates on starting SD farther north than Pt Conception given historical biomass distributions?

Transect | Potential Locations | Date Range -- | -- | -- 45 | Santa Barbara | 7/17-7/19 128 | Astoria | 8/17-8/19 JDF Strait? | Neah Bay or Port Angeles (long transit) | 9/17-9/19
kstierhoff commented 1 year ago

@alicebeittel my gut feeling is to have SD start at Pt. Conception at the expense of not having them sample all the way to Cape Flattery, given that we already have RL, SH, and LM sampling intensively off the coast of OR and WA. While biomass can be high between Pt. Conception and Monterey Bay can be large, diversity has been low and most of the biomass is anchovy, so I thought we could potentially sacrifice some sampling precision by sampling that area at 20 nmi spacing with RL only. @ddemer do you have any thoughts on this?

img_species_proportion_distributions_2015-2022

alicebeittel commented 1 year ago

@kstierhoff @ddemer I think either of those options (or combo of both) could be what we need to bring us within our DAS budget with the leapfrogging idea. Reducing SD lines north at Cape Flattery would look like cutting out 6-8 SD lines (so ending near Willapa Bay, WA ~130S).

kstierhoff commented 1 year ago

Alice, Let's make a few additional changes toward finalizing this plan: 1) set the nearshore extent of Saildrone lines to 5 nmi from shore; 2) limit lines to 30 nmi, which would allow a line to be completed in 2d. At that rate, the Salldrones could complete a cluster of 5 transects in ~16 d (2 on each transect x 5, and 1 d between transects x 4); and 3) removing the adaptive lines from the SCB. From that, we might hope to get a more firm estimate or progress by RL and SD and closer to nailing-down our transect plan for the project instructions. Kevin

On Fri, Mar 3, 2023 at 9:41 AM Alice Beittel @.***> wrote:

@kstierhoff https://github.com/kstierhoff @ddemer https://github.com/ddemer I think either of those options (or combo of both) could be what we need to bring us within our DAS budget with the leapfrogging idea. Reducing SD lines north at Cape Flattery would look like cutting out 6-8 SD lines (so ending near Willapa Bay, WA ~130S).

— Reply to this email directly, view it on GitHub https://github.com/nmfs-swfsc-ast/ast-tasks/issues/23#issuecomment-1453877072, or unsubscribe https://github.com/notifications/unsubscribe-auth/AARG62VF2DXTU5W5JSR3Y4LW2IUN7ANCNFSM6AAAAAASZUCSUA . You are receiving this because you were mentioned.Message ID: @.***>

--

Kevin L. Stierhoff

Research Fisheries Biologist, NOAA/NMFS/SWFSC

alicebeittel commented 1 year ago

Here are some new date estimates for the SD with 30nm lines. Knowing now that a data transfer mid-survey in SF for SD1 and SD2 would take about a week, I'd recommend that we do not do a data transfer as this will allow us to keep our maximum gap between RL and SD around 10-12 days instead of 18-20 days. These dates account for an estimated 6-day transit for each SD between each 5 transect cluster. My math checks out with roughly what you calculated above (~16 days per 5-transect cluster).

Total SD DAS with this plan is 165 (includes transits between SD clusters). Should I share these estimates with SD?

  | Start |   | End |   -- | -- | -- | -- | -- SD 1 | 52 | 7/10 | 60 | 7/22 SD 2 | 62 | 7/10 | 70 | 7/23 SD 3 | 72 | 7/14 | 80 | 7/27 SD 1 | 82 | 7/27 | 90 | 8/9 SD 2 | 92 | 7/28 | 100 | 8/11 SD 3 | 102 | 8/1 | 110 | 8/14 SD 1 | 112 | 8/14 | 120 | 8/27 SD 2 | 122 | 8/16 | 130 | 8/29 SD 3 | 132 | 8/19 | 140 | 9/3

https://docs.google.com/spreadsheets/d/1wYwtqfkxr-O3uU_gNCWZHiLYFmqbfgv1zDbT0ubcvz8/edit#gid=333260667

alicebeittel commented 1 year ago

@kstierhoff Updated .nob uploaded to google drive with appropriate SD transect edits. "2307RL_20230314.nob" https://drive.google.com/drive/folders/1-XWEnb9aKYH2pois1Fop9suR_JgMeUFq

kstierhoff commented 1 year ago

Updated the .nob file (2307RL_20230531.nob) and GPX file (rosepoint_waypoints_20230531.gpx) here: https://drive.google.com/drive/folders/1-XWEnb9aKYH2pois1Fop9suR_JgMeUFq

This replaces the file provided to NAV earlier this week, and should be final.