ADRIA.jl

Appearance

Exporting to RME Format ​

The ADRIA.RMEExport.export_to_rme() utility allows ADRIA ResultSet objects to be exported into a format compatible with ReefModEngine.jl. This is primarily used to link ADRIA ecological simulations with the CREAM (Comprehensive Reef Economic Analysis Model) suite via the cost-eco-model-linker.

Domain Compatibility ​

The export targets Great Barrier Reef (GBR) domains. Both a standard ADRIA GBR Domain and an RMEDomain (load_domain(RMEDomain, ...)) are supported. It requires:

  • A GBRMPA_ID column in the domain spatial data (rs.loc_data), used to name the yearly reefsets in scenario_info.json. Current ADRIA GBR DataPackages and current RME data packages both provide it. Older packages without GBRMPA_ID, and non-GBR or synthetic domains, are not supported.

  • The standard 5-group functional coral model and the standard seeding factors (N_seed_*, seeding_devices_per_m2). The evenness normalisation hard-codes 5 groups.

  • An RCP in the results that also exists in the domain's DHW data (needed for the GCM-name lookup).

Overview ​

Because ADRIA and RME use different internal standards, the export process performs several critical "translation" steps: 2. Unit Scaling: Proportions (0.0 - 1.0) are scaled to percentages (0 - 100) where required by the economic model.

  1. Coordinate Variables: NetCDF outputs include human-readable coordinate variables for timesteps (years) and locations (Unique Reef IDs).

  2. Intervention Realization: ADRIA's spatial intervention logs (seed_log) are translated into absolute coral counts and realized intervention areas (km²) for logistical cost calculations.

Usage Workflow ​

1. Generate Paired Scenarios ​

The economic model calculates benefits by comparing an intervention to its corresponding counterfactual. Use sample_pairwise() to ensure your scenario set contains the necessary pairs in the correct order.

julia
using ADRIA

# Standard ADRIA GBR domain
dom = ADRIA.load_domain("path/to/domain", "45")

# ...or an RME data package
dom = ADRIA.load_domain(RMEDomain, "path/to/rme_data_package", "45")

# Generate 16 paired scenarios (16 Counterfactuals + 16 matching Interventions)
scens = ADRIA.sample_pairwise(dom, 16)

2. Run Simulations ​

Run the scenarios normally. Ensure you are using a domain compatible with the Great Barrier Reef (GBR) if you intend to use the standard CREAM cost models.

julia
rs = ADRIA.run_scenarios(dom, scens, "45")

3. Export Results ​

Call export_to_rme() to generate the RME-compatible directory structure.

export_to_rme() takes the Domain used to run the scenarios, the ResultSet, and the output directory.

julia
using ADRIA.RMEExport

out_dir = "./economic_analysis_input"
export_to_rme(dom, rs, out_dir)

Optional keyword arguments: map_counterfactuals, pdp_geopackage_path, scen_spec.

Exported Files ​

The utility produces the following files in the target directory:

  • results.nc: NetCDF containing total_cover, relative_juveniles, relative_shelter_volume (also written as relative), evenness, and zero-filled rubble / cots placeholders, plus timesteps / locations / scenarios coordinate variables.

  • iv_yearly_scenarios.csv: A year-by-year log of absolute coral counts, seeding densities, and deployment areas per scenario.

  • scenario_info.json: Metadata mapping scenario indices to counterfactual status and defining the unique "reefsets" used each year.

  • reef_information.csv: Spatial metadata for the reefs involved in the simulation.

Technical Notes ​

Shelter Volume Scaling ​

The downstream economic model applies a 9.33x scaling factor to the exported relative_shelter_volume. This reconciles ADRIA's reference standard (relative to 1 m²) with the historical GBR standard (relative to a 95cm diameter circle).

Counterfactual Detection ​

A scenario is flagged as a counterfactual in the export if all intervention factors (N_seed, fogging, SRM, etc.) are set to zero. This bitmask is stored in scenario_info.json.

This page was generated using Literate.jl.