Overview
The purpose of atlantisom
is to use existing Atlantis ecosystem model output to generate input datasets for a variety of simpler population models, so that the performance of these models can be evaluated against known (simulated) ecosystem dynamics. Atlantis models simulate a wide range of physical and ecological processes, include a full food web, and can be run using different climate forcing, fishing, and other scenarios.
Users of atlantisom
specify fishery independent and fishery dependent sampling in space and time, as well as species-specific catchability, selectivty, and other observation processes to simulate survey and fishery “data” from stored Atlantis scenario output. atlantisom
outputs internally consistent multispecies and ecosystem datasets with known observation error characteristics for use in individual model performance testing, comparing performance of alternative models, and performance testing of model ensembles against “true” Atlantis outputs.
Development of atlantisom
began at the 2015 Atlantis Summit in Honolulu, Hawaii, USA. Substantial progress since then was made possible by a NOAA NMFS International Fellowship and the Institute of Marine Research in Bergen, Norway.
Credits
Sarah Gaichas, Christine Stawitz, Kelli Johnson, Alexander Keth, Allan Hicks, Sean Lucey, Emma Hodgson, and Gavin Fay
Get atlantisom
https://github.com/r4atlantis/atlantisom
Detailed data flow
Have these Atlantis input files in a directory:
- The .bgm file defining model geometry
- The initial conditions/biomass pools .nc
- The functional groups .csv
- The fishery groups .csv
- The biology .prm
- The run prm (.xml version)
At present,atlantisom
uses the following Atlantis output files, where […] is a model-specific prefix. They should be in the same directory as the input files above:
- […]BiomIndx.txt
- […]Catch.txt
- […]CatchPerFishery.txt
- […]DietCheck.txt
- […]YOY.txt
- […].nc
- […]CATCH.nc
- […]PROD.nc
- […]ANNAGEBIO.nc (if available)
- […]ANNAGECATCH.nc (if available)
Get the truth:
Generate the data:
Background
Assessing the ability of models to predict key processes is essential if the models are to be used in decision-making dependent on those processes. Skill assessment compares model predictions with observations; good agreement between predictions and observations indicates high skill (see, e.g. Stow et al 2009). However, in many natural resrouce management contexts, the observations themselves are highly uncertain, so assessing model skill can be difficult. In these cases, creating an observational dataset with a simulation model can provide alternative information for model skill assessment.
End-to-end ecosystem operating models as dataset generators
Atlantis is a spatially resolved mechanistic end-to-end ecosystem modeling framework: Fulton et al. 2011, Fulton and Smith 2004. Atlantis models have been implemented for regional ecosystems around the world, including:
Norwegian-Barents Sea
California Current
Marshall et al. 2017, Kaplan et al. 2017
Atlantis models can incoroporate physical drivers from global change projections and simulate complex biological responses throughout the ecosystem: Hodgson et al. 2018, Olsen et al. 2018
Why use Atlantis?
- Mechanistic processes create internally consistent “truth”
- Include cumulative effects of multiple processes:
- Climate drivers
- Species interactions
- Spatial and seasonal variability
- Fisheries
- Oil spills, red tide, anything else Atlantis can do
- Implemented for many ecosystems worldwide
Why generate datasets instead of simulating within Atlantis?
- Not all analyses need computationally expensive model interaction
- Faster!
- Test many models or model configurations with the same dataset
- Many dataset realizations from same “truth”; compare:
- Different observation error and bias
- Changing temporal and spatial survey coverage
- Improved or degraded fishery observations
How does atlantisom
make “data” for stock assessments?
The atlantisom
user must specify uncertainty in assessment “data”:
Survey specification:
timing and spatial coverage?
which species are captured?
species-specific survey efficiency (“q”)?
selectivity at age for each species?
Survey uncertainty:
additional observation error (survey cv for index)?
effective sample size for biological samples?
Fishery uncertainty:
additional observation error (catch cv for total)?
catch sampled for length/age in all areas?
effective sample size for biological samples?
Using these specifications, atlantisom
generates survey data from the true biomass outputs of the Atlantis model, and generates fishery data from the true fishery outputs of the Atlantis model. The simulated data can then be used as inputs to a variety of single species or multispecies assessment models.
Make Atlantis output into assessment model input
Example atlantisom workflows:
Get true biomass, abundance, age composition, length composition, weight at age, fishery catch, fishery catch at age, fishery length composition, and fishery weight age age for a “sardine-like species”: https://sgaichas.github.io/poseidon-dev/FullSardineTruthEx.html
Format these outputs and get other life history parameters for input into a stock assessment model (Stock Synthesis, using
r4ss
): https://sgaichas.github.io/poseidon-dev/CreateStockSynthesis.htmlGet true and observed input data, format inputs, and run the assessment model: https://sgaichas.github.io/poseidon-dev/SardinesHakeatlantisom2SStest.html
In progress: compare assessment results with truth: https://sgaichas.github.io/poseidon-dev/SkillAssessInit.html
Simplified dataset extraction with wrapper functions: https://sgaichas.github.io/poseidon-dev/NOBAcod.html
Visualizing atlantisom
outputs
Survey census test NOBA
True length composition NOBA
Standard survey test CCA
Survey length composition CCA
Example: a “sardine” assessment
Need: assessment model data inputs and life history parameters
(model based on actual Sardine assessment in Stock Synthesis 3)
Data:
- survey biomass index
- survey length composition
- survey age composition (conditional catch at age)
- fishery catch (tons)
- fishery length composition
- fishery age composition
Parameters:
- natural mortality (from total mortality)
- growth curve (from survey length at age)
- maturity at age (true)
- unfished recruitment and steepness (true)
- weight-length curve (true)
A “sardine” assessment: setup
- California Current Atlantis run with and without climate signal
- Input data generated (e.g. sardine survey, below in green)
- Parameters derived; simpler recruitment distribution
A “sardine” assessment: fits to data
A “sardine” assessment: skill? (proof of concept)
Biomass
Fishing mortality
Recruitment
Key: True SS3 estimate