The Ocean Health Index (OHI) measures the sustainable delivery of ten ‘goals’ (key benefits and services) to people now and in the future. At any spatial scale, OHI assessments use the best available information to develop goal models and set targets, and goal scores are combined together to give a comprehensive picture of coastal ocean health.

The OHI Toolbox is an open-source, cross-platform resource to calculate OHI scores at any scale. Here, we will use the Toolbox to add a new variable to a goal model. We will use Belize as an example: this is for teaching purposes only.

Note: this presentation is made with Rmarkdown. It is viewable rendered online and in its raw form on GitHub.

We will walk through an example of adding a new variable to the Artisanal Fishing Opportunity (AO) goal model in Belize.

Assume your team has already discussed what is important in terms of AO in Belize. You have explored all data available and your team has decided that the best way to capture local characteristics is to add a new variable to the global AO model.

Your team has decided that the best way to capture local AO information is to add a new variable to the global AO model.

Steps:

1. Review global AO model
2. Prepare data layer & save in the layers folder
3. Register the layer in layers.csv
4. Update goal model in functions.r
5. Check and possibly update goals.csv
6. Calculate scores; view with WebApp

AO global model variables: need, access, and sustainability.

\[ D_{U} = (1 - P) * (1 - O_{AO}) \] \[ x_{AO} = (1 - D_{U}) * S_{AO} \]

\(D_{U}\) = unmet demand
\(P\) = need (PPP-adjusted per capita GDP)
\(O_{AO}\) = access to artisanal fishing
\(S_{AO}\) = sustainability of fishing methods

AO global model variables: need, access, and sustainability.

\[ D_{U} = (1 - P) * (1 - O_{AO}) \] \[ x_{AO} = (1 - D_{U}) * S_{AO} \]

The way this looks like in functions.r:

    Du = (1.0 - need) * (1.0 - access)
    statusData = ((1.0 - Du) * Sustainability)

Let's say we have data available about poverty in Belize, which can be incorporated as part of need (the global model is based on need and access).

Let's say we have data available about poverty in Belize, which can be incorporated as part of need (the global model is based on need and access).

Let's incorporate a second measure of need called \(P_{pov}\).

\[ D_{U} = (1 - (P + P_{pov}) / 2) * (1 - (O_{AO}) \] \[ x_{AO} = (1 - D_{U}) * S_{AO} \]

Let's say we have data available about poverty in Belize, which can be incorporated as part of need (the global model is based on need and access).

Let's incorporate a second measure of need called \(P_{pov}\).

\[ D_{U} = (1 - (P + P_{pov}) / 2) * (1 - (O_{AO}) \] \[ x_{AO} = (1 - D_{U}) * S_{AO} \]

The way this looks like in functions.r:

    Du = (1.0 - (need + need_pov) / 2 ) * (1.0 - access)
    statusData = ((1.0 - Du) * Sustainability)

What do our new data look like?

• Population by district
• Percent below poverty line for Belize

What do our new data look like?

• Population by district
• Percent below poverty line for Belize

What do the existing data layers look like?

• Explore using the blz WebApp
  • http://ohi-science.org/blz/layers
• Explore the blz repository
  • access
  • need

Working with data: good practices

• organize & prepare data in a github repository
  • ex: github/blz/blz-prep (ours is stored here)

Working with data: good practices

• organize & prepare data in a github repository
  • ex: github/blz/blz-prep (ours is stored here)
• document all metadata
  • README.mds are readable on GitHub (example)
• create a name-to-region lookup table
  • ex: rgn_district_lookup.csv
• script any data manipulations
  • prep_ao_need.r

Our R script prep_ao_need_pov.r prepares the data layer called ao_need_pov.csv with these steps:

• Setup
• Access and clean raw data
• Calculate percent poverty by district
• Add region identifiers
• Save final data layer

Since we've created a new variable, we'll add a new row to layers.csv, with information for the following columns:

• targets: AO
• layer: ao_need_pov
• name: layer name will be displayed on your WebApp.
• description: description will be displayed on your WebApp.
• fld_value: percent
• units: percent
• filename: ao_need_pov_blz2014.csv
• fld_id_num: rgn_id

Update the model in functions.r, (and some layer handling):

    Du = (1.0 - (need + need_pov)/2 ) * (1.0 - access)
    statusData = ((1.0 - Du) * Sustainability)

Look at goals.csv if there is anything to update (particularly in the preindex_function column)

After you make your changes, you can calculate and visualize your scores in several ways:

1. In RStudio, run calculate_scores.r and visualize scores by running launch_app_code.r
2. In RStudio, run calculate_scores.r, sync changes to GitHub and visualize scores on the WebApp
3. Sync your changes to GitHub: scores will be calculated online and can be visualized on the WebApp