Background: The Factors
Coliform bacteria, like E. Coli, generally grow in the intestines of warm blooded animals, but generally do not grow in recreational waters, making them a useful indicator of water contamination. We classified degree of water contamination by using a dataset that measured concentration of coliform bacteria. Increased water contamination was indicated by a higher concentration of coliform bacteria in the ocean water.
When it rains, part of the precipitation sinks into the soil, and part of it flows downhill, eventually ending up in a body of water. This runoff can carry fertilizer from agricultural lands, leading to elevated concentrations of pollutants such as nitrogen and phosphate in bodies of water including the ocean. The runoff can also contain bacteria from animal and human waste and petroleum by-products from leaking vehicles. Because this pollution cannot be classified or modelled as coming from a single point such as a pipe discharging industrial waste for example, it is called non-point pollution.
Factors that influence runoff can be classified into two categories: meteorological and physical characteristics of the landscape. The main meteorological factors that affect runoff include the type of precipitation (rain, snow, sleet,...), temperature, wind, relative humidity, season and the rainfall intensity, amount and duration.The main physical factors that affect runoff include agricultural land use, vegetation, soil type, drainage area, elevation, slope, topography, and drainage network patterns. Human activity increases runoff by replacing the soil that formerly absorbed some of the water with concrete. Effects of increasing runoff include increased erosion, groundwater pollution and increased concentration of agricultural pollutants in oceans. Increased rainfall increases the concentration of eroded material in ocean water, causing increased turbidity and contamination. Runoff transports water pollutants to groundwater and ocean water, causing a contaminated drinking source and ocean environment. Increased nitrates in the water cause eutrophication and decreased dissolved oxygen in the water, leading to large death of marine mammals, for example, fish kills. Runoff from pesticides such as DDT can genetically alter the gender of fish, transforming male to female fish. Flooding can also occur if the rate of rainfall exceeds the rate of soil absorption and runoff combined.
Turbidity is defined as the measure of the clarity of a fluid. It is measured by shining light through a sample of liquid. The higher the turbidity, the higher the intensity of the resulting scattered light. It is caused by particles such as “clay, silt, finely divided inorganic and organic matter, algae, soluble colored organic compounds, and plankton and other microscopic organisms” suspended in the ocean water. These contaminants are often caused by shoreline erosion, or inadequate sewage processing facilities. Turbidity is directly correlated with the degree of pollution of ocean water. A high turbidity adversely affects ocean environments because photosynthetic algae depend on light to photosynthesize, and the the light is largely blocked in ocean water with a high turbidity. Because photosynthetic algae are the energy basis of the marine ecosystem, this can have large effects across the marine food web. If there is decreased visibility, this can also cause decreased animal mating, decreasing animal populations. PH is defined as the concentration of free hydrogen ions in a solution. Since the Industrial Revolution, ocean water pH has fallen by 0.1. Because the pH scale is logarithmic, this corresponds to a 30% increase in ocean acidity. Various models predict that ocean water could around 150% more acidic by the end of the century. This is known as ocean acidification. Changes in pH are caused largely by two factors: changes in CO2 concentration and changes in alkalinity. Changes in CO2 concentration in oceans result from an increased amount of CO2 in the atmosphere. As CO2 dissolves in water, it reacts with water to form carbonic acid (H2CO3).
[CO2] + [H2O] <=> [H2CO3]
Carbonic acid then releases a hydrogen ion/proton (H+) and bicarbonate (HCO3-)
[H2CO3] <=> [H+] + [HCO3-]
The hydrogen ion/proton (H+) will react with carbonate ion (CO32-) to form carbonic acid
[H+] + [CO32-] <=> [HCO3-]
Overall, this process causes an increase in the concentrations of H+, H2CO3 and HCO3- a decrease in the concentration of CO32-. There are also more free hydrogen ions, causing the a decrease in pH and more acidic seawater. Overall, ocean contamination is both and cause and an effect of ocean acidification. Alkalinity refers to the capability of water to neutralize acid (buffering capacity). Hydrolysis of bicarbonate and carbonate helps to buffer (or minimise) pH changes caused by photosynthesis. As concentration of carbonate ions (CO32-) decrease due to increased CO2 concentration, so does their ability to minimize changes in pH due to photosynthesis, and acidity rises/pH falls. Ocean Acidification is important because it results in a decreased amount of free carbonate in ocean water, which means that calcium cannot bond to carbonate to form calcium carbonate. Marine animals have shells/skeletons made of calcium carbonate use the calcium carbonate in water to maintain this skeleton in a process called calcification. Less calcium carbonate in water is life threatening to these species because they can’t maintain their skeletons. Coral reefs are held together by calcium carbonate, and a decrease in calcium carbonate decreases the resistance of coral reefs to bleaching, disease and death. Because coral reefs a provide habitat that increases biodiversity, less coral reefs would result in less biodiversity. Plankton would also be affected by the decrease in calcium carbonate concentration in ocean water. Because they are one of the bases of the food chain in the ecosystem, decreasing plankton would decrease food and energy supply across the marine ecosystem. Other organisms affected by a change in calcium carbonate concentration include sea urchins, crabs, and lobsters, all of which are important to fisheries and local economies.
Global Warming has raised ocean temperatures by about 0.18 degree Fahrenheit/0.1 degrees Celsius over the last century. This can be incredibly detrimental to both marine mammals and people. Due to thermal expansion, an increase in temperature of the oceans causes the sea level to rise. This in turn causes increased soil erosion, increasing turbidity and ocean contamination. Increasing water temperature also has a detrimental effect on marine life because temperature acts as a regulator of ocean environments. All species can survive in a certain temperature range, and native species’ specific temperature range is different from invasive species temperature range. Being unable to survive in certain ocean climates often prevents a certain species from becoming an invasive species. However, if ocean temperatures rise, this may trigger a large forced migration of a certain species and/or cause extinction of the species. It may also allow some invasive species and bacteria to thrive, where they were unable to previously. Warmer water temperatures may also interrupt the “ocean conveyor belt” that regulates temperature fluctuations of the earth, causing a higher rate of atmospheric temperature rise.
The Process
In order to begin our process, we had to find past data on the factors we were attempting to base our model on. This proved to be difficult because much of the data we found was incomplete, from different location, and difficult to access directly. We also had difficulty finding data from a specific date range in order to relate our different factors to one another. We found turbidity and water temperature data from a sensor under the Santa Cruz wharf that had been recording since around 2013. In order to relate other factors to turbidity and water temperature, we tried to find data from the same time frame for the other factors.
We then tried to find water contamination levels, which we modelled by levels of coliform bacteria from 2013 to 2017. We contacted the county government of Santa Cruz and asked them for data from the different beaches around Santa Cruz. The data they sent us was taken weekly, spanning several years, from about 15 beaches in Santa Cruz. We decided to use the data from Cowell Beach because it had the correct time frame that allowed us to correlate it with the other data we had found, and it had a large amount of recordings per month. If we were to continue work on this project, we could use our model to predict water contamination for many different locations, simply by obtaining this dataset from that location.
The water contamination data given to us consisted of the number of colony-forming units of both Enterococcus and E. Coli. In order to use this data for a three dimensional graph, we needed to condense these two dataset into a singular dataset.
Using country and state health limits, we generated a color for each level of contamination as indicated by the combined coliform levels. Green indicates a safe level of contamination, yellow indicates caution, and red indicates an unsafe level of contamination. This was overridden if either or the two were over a limit, because sometimes the combined can be in the green, even if one of the two is in the red.
After attempting to combine all of the datasets into one graph, Emily realized that the datasets had different time steps for taking data, making the combined graph essentially meaningless. We wanted all of our factors to be parameterized by time (t), however the program was making it parameterized by the index. This essentially means that the program was acting as if the data recorded every hours matched the data recorded every 5 minutes just based on how the columns were aggregated. As a result, when Emily graphed the data the x, y, and z of each point did not match up by time, but by which row number the were in their data set. At first Emily tried to write a program that combined only the data that had matching time stamps, hoping that the data being graphed would then correlate by time, however, this seemed to be overly complicated and created even more problems. She then decided to manually combine the data we had into a new data set that had a column for each of the data we wanted to graph by time(t).
With this new data set, we used the time step from the the water contamination date because it was the largest and for each day there was water contamination data, we added five recordings of temperature, rainfall, and turbidity data. We did this to get more data points and to try and get a more accurate reading because data like turbidity fluctuates rapidly. With this she was able to generate a 3D graph with a color map representing water contamination.
With this graph we looked for qualitative trends in water contamination. If there was a cluster of red dots, was it at a high or low temperature, high or low turbidity, high or low rainfall? By doing this, we could determine what factors generally correlated with higher or lower levels of contamination
We also searched for constantly updating data, which took a really long time to find, but allows us to build a program that is constantly updating. We found a site that contains information on all of the data in our function that is being recorded at the Santa Cruz Wharf. We then built a program to grab that data off of the website and input it into the function we created. The function then outputs a color, which represents water contamination levels.