atlantisFleets.RmdUsing Atlantis code base: v6681
Several options to parameterize fleet effort:
Forced effort time series (days \(day^{-1}\)) by box - effortmodel = 11
Forced effort time series system wide (Global). Similar to forced catch. Development on new branch
Specify fleet total effort (scalar) that is fixed over the run. YYY_effort > 0 (days \(day^{-1}\))
We will implement box level forced time series effort by fleet (option 1)
CAMS data was pulled, QAQC’d, then processed to clean the raw data.
This data set was then used to identify communities at sea (fleets of fishers with similar behavior, gear, target species, port etc). For example:
A fleet is defined by trips using similar gear, targeting similar species, and landing in a similar port.
To select fleets the data was used to:
To force by box level effort we need to define fleets along the northern part of the eastern seaboard.
We will focus primarily on:
forcing.prm file
Effort_hdistribYYYd, no forcing, effortmodel = 9). This is
limiting and prevents looking at scenarios with changes in effort
through time based on management decision or climate changesHarvest = Effort * YYY_sweptarea * f(selcurve_YYY,sel_XXX) * management_scalar * (vertical overlap * positional availability) * g(habitat overlap, q_XXX) * Biomass (1-p(escapement) )
where
escapement = h(flagesacpement_XXX, p_escape_XXX ,ka_escape_XXX ,kb_escape_XXX)
Effort is a forced box level time series we supply
(effortmodel=11) or or Effort is fixed (effortmodel = 9)
where catch is proportional to population density (biomass)
YYY_effortmodel (at_harvest.prm)
flagimposecatch_XXX(at_harvest.prm)
catchall) is a catch all fleet. This fleet has value 1 for
all species. All other fleets are set to 0.flagincidmort (at_harvest.prm).
target_YYY
target_YYY (at_harvest.prm).
flag_access_thru_wc_XXX (at_harvest.prm)
XXX_flagdempelfishery (at_harvest.prm)
flaghabitat_XXX (at_harvest.prm)
q_XXX.habitat_XXX (at_biology.prm)
Does a value of zero indicate fishing is excluded? Current model has text indicating which habitats each group prefers. Value of 1 = group lives on habitat. Need to sort this out
YYY_mindepth,YYY_maxdepth
(at_biology.prm)
All mindepth values = 0 Maxdepth values most are 5000 or 1900. Some 255,800,25,50,100,200. These are currently specified by species. Think they should be by fleet
YYY_sweptarea (at_harvest.prm)
flagchangeswept =1. Turns on temporal changesYYY_changeswept_parameters - indicate which fishery has
changesYYY_swept_changes - the number of changes through
timeSWEPTchange_start_YYY - the start day of each change
for each fisherySWEPTchange_period_YYY - the periods over which each
change occursSWEPTchange_mult_YYY - the multiplier used to get the
final value at the end of the change periodStarting value of 1E6 m3 = .001 km3 per day by the fishery.
eg. If a ship trawled for 1km with a net opening of 10m x 10m it would sample = 1E5 m3 water = 10% of starting value.
NOTE: Manual 15.8.12. To convert the swept volume per unit effort of fishing pressure to fishing pressure per day per m3 Atlantis divides the total swept volume (in m3 per day per box) by the box volume available to the fishery based on its access to habitats, i.e. the propfish value. This is the same proportion of accessible box volume that is used in the calculation of the habitat scalar, described in chapter 15.7.3.
\[F_{m3}=\frac{volume_{swept}}{volume_{box}prop_{fish}}\]
where \(prop_{fish}\) = is the proportion of box area accessible to the fishery according to the habitats they can access, which can also be seen as a probability that the fishing gear has access to the individuals of species in this box (note, this is NOT a proportion of area fished based on effort allocations, but just the habitat overlap) \(volume_{box}prop_fish\) = total volume of available to fishery
NOTE: NOTE! Ensuring meaningful representation of fisheries habitat
access and vertical effort distribution Note, that the habitat dependent
propfish scalar is applied to each layer of the box and not
to the bottom layer only. This means that the users have to make sure
that the vertical distribution of fishing effort and habitat access by
fisheries makes sense. The habitats probably should have no effect on
fisheries that operate in pelagic areas only. In this case the
habitat_YYY should be set to 1 for all habitats to ensure
that the fishery has access to the entire volume of the cell. The
habitat_YYY vector makes sense for demersal fisheries only
and the users should make sure that the vertical distribution of the
fishing effort is also concentrated (or occurs entirely) in the bottom
water column layer.
Need to better understand the range of a tow to estimate if this value has any meaning
As a reference the average swept area for a standard tow in the bottom trawl survey = xxx?
q_XXX. (at_harvest.prm)
flaghabitat_XXX = 0 for all fleets this is a simple
multiplier.sel_XXX. (at_harvest.prm)
YYY_selcurve. (at_harvest.prm)
sel_XXX)From the manual: The logistic selectivity curve is one of the most
commonly applied length-based selectivity options in fisheries (most
often used for trawl nets or other gear where the likelihood of capture
grows to plateau with larger sizes). It is defined by two parameters –
the length at 50% selectivity, also defined as the inflection point
lsm (YYYsel_lsm, given in cm) and the spread or steepness
of the selectivity curve selb (YYYsel_b). See Manual for
the parameter selb affects the shape of the selectivity
curve. The selectivity of age group i is calculated as
\[psel_i = \frac{1}{1+exp(-selb(length_i - lsm))}\] where
\(lsm\) = length at 50% selectivity
\(selb\) = steepness of the selectivity curve
\(len_i\) is the length of the age group i.
Parameter estimates of \(\hat{selb}\) and \(\hat{lsm}\) can be found in the NE groundfish selectivity
If length-based selectivity is selected for a fishery, it will be applied to both age-structured groups and biomass pools. The application of length-based selectivity to biomass pools is a new feature in Atlantis and it is important to get it right, or else invertebrates may not be caught by the gear (see wiki post here)
Length of an age group in age-structured groups is calculated using
length-weight conversion parameters (li_a_XXX and
li_b_XXX) given in the biology.prm file. For this, first,
the structural and reserve nitrogen (RN+SN, in mg) are added up and
converted to wet weight in grams (wgt) as:
\[wgt = (RN + SN) . wetdry. X_{CN}/100\]
where wetdry is the wet weight to ash free dry weight ratio
(k_wetdry in biology.prm) typically set to 20 in many
Atlantis parameter files, based on using C as a proxy for dry weight
(but noting that in many models and literature the wet : dry ratio used
is often closer to 6-8) and X_CN is the carbon to nitrogen Redfield
ratio (X_CN in biology.prm) typically set to 5.7.
The wgt (in grams) is then converted to length (in cm) using the standard equation
\[wgt = a.length^b\] or \[length = \left( \frac{wgt}{a}\right)^{1/b}\]
where a is li_a_XXX and b is li_b_XXX in
the biology.prm. Note that these parameters could be set in such a way
to return length in metres instead, but by convention it is in cm (as
are any other faunal lengths).
YYYsel_lsm (at_harvest.prm)
YYYsel_b (at_harvest.prm)
NOTE: from Manual. Chapter: 15.4 It might be easier to track the fishing outcomes if catchability or escapement are modified one at a time. So, for example, the catchability can be set to 1, but escapement can be set to a different age or length specific proportion. Alternatively, the user might want no escapement and modify the available biomass through catchability only. Finally, the simplest option would include no escapement and 100% catchability; then the available biomass of the target and bycatch species of fishery YYY, in the swept volume of the cell, would only be limited by the fishery’s gear selectivity.
Provisions are made for escapement in the NEUS model but all fisheries set to have no escapement
flagescapement_XXX (at_harvest.prm)
p_escape_XXX)Ka_escape_XXX,
Kb_escape_XXX)p_escape_XXX (at_harvest.prm)
Ka_escape_XXX,Kb_escape_XXX - All
values = 0
YYY_flagmanage (at_harvest.prm)
With these settings the Harvest should collapse to something like:
Harvest = Effort * sweptarea * Biomass
Therefore included estimated Effort in forcing file Effort = Harvest/(Biomass*sweptarea)