Metrics

The available metrics include the Reef Condition Index (RCI), the Reef Tourism Index (RTI) and Reef Fishing Index (RFI). There are also several versions of the RCI available, including raw RCI and RCI, which measures the total reef area in good or very good condition.

Raw RCI

The RCI can be calculated using process_RME_data.raw_rci(). This metric comprises 5 reef metrics:

  • Relative coral cover

  • Relative shelter volume

  • Relative abundance of juveniles

  • Complementary relative crown of thorns population (relative to outbreak levels)

  • Complementary rubble cover

These values are categorised as within Very Good, Good, Fair, Poor or Very Poor condition depending on their values and a set of categories compiled by expert elicitation. The expert elicitation process was carried out at a workshop in October 2021 with 8 coral reef experts elicited, resulting in the summary table below.

Expert elicitated reef condition metrics

Condition

RelCover

Sheltervol

CoralJuv

CoTSOutBr

Rubble

VeryGood

0.45

0.45

0.35

0.95

0.85

Good

0.35

0.35

0.3

0.85

0.8

Fair

0.25

0.3

0.25

0.75

0.75

Poor

0.15

0.25

0.2

0.7

0.7

VeryPoor

0.05

0.15

0.15

0.65

0.65

The raw RCI is calculated by assessing the condition of a reef at each timestep for each metric according to the above categories. The reef is assigned the condition of the highest category for which 3 or more metrics satisfy that category’s threshold. The category thresholds can be used as means across all 8 experts without uncertainty. Alternatively, the raw expert condition categories of data collected from the 8 experts can be sampled when calculating metrics by setting expert_uncert=1.

RTI

The RTI, or Reef Tourism Index, is a continuous version of the RCI condition categories. RTI can be calculated using process_RME_data.raw_rti().

RFI

The RFI or Reef Fishing Index, estimates the total fish biomass in kg km^2, based on a linear regression of total cover. This is based on digitisation of Fig 4A and 6B in Graham and Nash, 2012. The RFI can be calculated using process_RME_data.rfi().

RCI

The RCI looks at the total area of reef for which the condition is within the Good or Very Good categories. RCI can be calculated using process_RME_data.rci().

Uncertainty sampling

If ecol_uncert=1, ecological uncertainty is sampled in the results by sampling climate model repetitions for a particular set of results (stochastic samples within one or more climate model). If ecol_uncert=0 the mean over all ecological reps is instead used. ecol_uncert=0 also uses the mean across all climate reps.

If expert_uncert=1, expert uncertainty is incorporated in the results by sampling a set of expert opinions on what thresholds of the 5 metrics incorporated in the Reef Condition Index should be considered as “Poor”, “Good”, “Very Good”, etc. condition. If expert_uncert=0 the mean of the 7 experts opinions is used (see ./datasets/ExpertReefCondition_AllResults.csv).

Currently, shelter volume uncertainty sampling has not been incorporated (shelt_uncert=0 is the default), as it needs access to number of corals in each taxa and size class in the ReefModEngine.jl resultset. This is currently not available in the result sets from ReefModEngine.jl, but could be incorporated in future versions.