People value marine waters that are free of pollution and debris for aesthetic and health reasons. Contamination of waters can come from land and ocean-based sources and cause many problems. Pollution can result in contaminated marine food, damage to mariculture, toxic blooms, beach closures, and mass kills of organisms. People are sensitive to these phenomena occurring in areas they access for recreation or other purposes but people even appreciate clean waters they do not directly use. The Clean Water goal captures the degree to which coastal waters are unpolluted by human-made causes. This goal scores highest when the contamination level is zero.

This goal should aim to capture the full spectrum of pollution types that can cause waters to become unsuitable for recreation, enjoyment, and other purposes. Categories of pollution can include eutrophication (nutrients), chemicals, pathogens, oil pollution, and marine debris. Ideally, data will be from direct measurements of pollutants from each category collected from on-going monitoring programs.

Practical Guidance

-STEP 1: Identify pollution data-

First consider the types and sources of pollution that may be in your area (see table). Types of pollution can include: trash/plastics, excess nutrients (e.g., nitrogen and phosphorous), and organic (e.g., pesticides, medications) and inorganic (e.g., heavy metals) chemicals. Sources of pollution can originate on land or in marine environments. For example: Are there known sources of trash and marine debris? Is wastewater effectively treated before it is discharged into the environment? How does urban and agricultural runoff contribute to your local coastal waters?

For the Global Assessment, we include 4 categories of pollution from multiple sources: nutrient pollution, organic and inorganic chemical pollution, pathogens, and trash.

Once the pollution types are identified, ideally you can find in situ measurements that directly measure pollutants. This could include data monitoring concentrations of pathogens, nitrogen, chemical contaminants, or trash. Pollutants can be measured in samples from the water column, sediments, or organisms, such as shellfish. For the Global Assessment, trash pollution was based on point sample data of offshore marine plastics that were interpolated to obtain complete global maps. Data at this scale will not capture local sources of trash pollution and isn’t great at describing coastal trash.

Alternatively, for some pollutants the data could reflect the impact of the pollution. For example, instead of describing nitrogen or phosphorous concentrations, data can describe the frequency and location of anoxic conditions or toxic algal blooms.

If direct measurements of water pollution are unavailable, the pollution levels may be estimated by modeling available data sources. For the Global Assessment, nutrient pollution was modeled from land-based agricultural inputs occurring within the watershed and then diffused into the ocean (e.g., Halpern et al. 2019).

If modeling is not a viable solution, proxy data could be used such as coastal population or shipping density. For the Global Assessment, pathogen pollution from human wastewater is based on the number of people who do not have improved access to sanitation.

You should attempt to use more refined data than the Global Assessment data, because it relies heavily on proxy data for water quality and may not be very accurate at smaller spatial scales.

Example sources of pollution

Source Pollutant Possible approaches
shipping organic pollution (oil spills) direct monitoring of oil spills, shipping density
shipping inorganic pollution (coatings, etc.) direct monitoring; shipping density
cruise ships nutrient pollution (sewage) direct monitoring, cruise ship density
cruise ships pathogens (sewage) direct monitoring cruise ship density
fishing trash and plastics direct monitoring and fishing effort and equipment
garbage disposal trash and plastics direct monitoring
agriculture nutrient pollution (fertilizer and manure) direct monitoring; modeling based on agricultural intensity
human wastewater microplastics direct monitoring; modeling based on human population
human wastewater nutrient pollution direct monitoring; modeling based on human population
human wastewater pathogens direct monitoring; modeling based on human population
human wastewater organic pollution (drugs, waste disposal) direct monitoring; modeling based on human population
land-based runoff inorganic pollution direct monitoring; modeling based on impermeable surfaces
point sources: shipbuilding, manufacturing, disposal organic and inorganic pollution direct monitoring; locating of known sources
mariculture nutrient pollution direct monitoring; modeling based on mariculture intensity
mariculture trash and plastics direct monitoring; modeling based on mariculture intensity
extraction: oil, minerals, etc. organic and inorganic pollution direct monitoring; locating of known sources
extraction: oil, minerals, etc. sedimentation direct monitoring
deforestation sedimentation direct monitoring; modeling based on land-use change, soil type, etc.

One potential problem to avoid is counting the same source of pollution in multiple ways. For example, if you have in situ measurements of excess nutrients, it would be unnecesary to include modeled nitrogen inputs from agriculture because presumably this would already be accounted for in the direct measures.

-STEP 2: Modeling the data-

For the Global Assessment, we focus on marine pollution occurring within 3 nm from shore. This is the region of most direct importance to humans, but this decision should be evaluated based on the objectives of your assessment.

Each of the 4 categories of pollution (nutrient pollution, chemical pollution, pathogens, and trash) should be rescaled to values from 0 to 1.

Fro the Global Assessment the reference point is no pollution, and consequently, a value of 1 indicates an environment with no pollution. You may also decided that an ocean completely rid of pollution is ideal, or you may use a different reference point. For example, you may use find that beach closure of less than 10 days per year due to E.coli contamination is acceptable.

The categories of pollution will need to be combined into a single score. For the Global Assessment, we combine the scores using a geometric mean which means that the final score will be driven by the worst performing pollutant category. This guarantees that if any one of the components scores poorly, the score will be low even if the other pollution types perform well.

Other Notes

The data used in the Clean Waters goal is also incorporated into the model as pressures layers. When the data are used as a pressure, the scores are inverted such that a high value of debris results in a high pressure score.

Examples of the Approach

Assessment Developing the Model Setting the Reference Point Other Considerations
Global 2012 The status was calculated as the geometric mean of four components, eutrophication (nutrients), chemicals, pathogens and marine debris. Reference point is when the contamination level is zero. The lack of direct measurements meant that modeled and proxy data were used. The status of this goal was also used in the pressures layers.
Global 2013 - 2015 The model was same as Global 2012, with a few simplifications. Revenue data were adjusted by dividing by GDP per region, reported in 2013 USD. The reference point was the same as Global 2012. The approach was the same as Global 2012, with simplifications.
Brazil 2014 The goal model was the same as Global 2012. Data used to model the components for eutrophication (nutrients) and chemicals was the same as in Global 2012, while pathogens and debris were localized to state level data. The reference point approach was the same as Global 2012. The study used better, or more local, data than the Global.
U.S. West Coast 2014 The model was the same as Global 2012, with regional instead of global data. The reference point was set as the number of days when beaches were closed to bathers because pathogen counts were higher than state standards. The study used more local data than the Global.
Israel 2014 The goal model was the same as Global 2012 with local data. The reference points for each category of pollutants are government-set standards. N/A
Ecuador-Gulf of Guayaquil 2015 The approach is similar to Global 2012. The same four indicators were used. However, trend data were added as follows: Coastal human population as a proxy for trend in trash. Fertilizer consumption was used as a proxy for trend in nutrient pollution. Trends in access to improved sanitation as a proxy for trend in pathogen pollution. Pesticide consumption as a proxy for trend in chemical pollution. The reference point is a zero pollution state for each of the components considered. N/A
China 2015 Status model is similar to global assessments. Pollution is based on nitrogen, phosphate, chemical oxygen demand, and oil pollution. The same reference point as Global 2012 was used that waters are free from all pollution. The study used all local data. Data on pathogens and marine debris are poor or unavailable and thus were ignored in the model.