Welcome to the Water Pollution website that I've made for my models. I made two models that are simulating the process of Eutrophication, which is a process where huge amounts of fertilizer or nitrogen goes into bodies of water and disrupts the balance of the aquatic ecosystem. I decided to research this topic because this type of water pollution is a serious problem in North Carolina. This usually occurs around the beaches of North Carolina, some ponds and and sometimes even lakes. The models on the webpage can be used to study how these bodies of water are or could be potentially polluted by the households or the agricultural industries in North Carolina.
This is a google document that I used as a HLD for my AgentCubes and Vensim Models. It has the information for how the model should be made, the components needed to make these models, and what these models are expected to do.
This model was made with AgentCubes, a website that allows you to create 3D worlds that have 3D agents. This AgentCubes model starts out with a group of farm agents on dry land, a group of dam agents in the lake, and the populations of a fish species and a plant species in the lake. When the simulation starts, the fish move around randomly in the water and the farms start spawning fertilizer agents that also move randomly in the water. The dam agents will be regular blocks that will stop some of the fertilizer agents from going through. There is a percent chance that a fish agent can die if it's next to a fertilizer agent. The farms will stop spawning fertilizer agents after a certain amount of time. If a fertilizer agent gets to a plant agent, the fertilizer agent will disappear and the plant agent will become a polluted plant agent. The polluted plant agents will reproduce rapidly towards the land and will kill the fish agents that's next to it. The polluted plant agents will also die when it's on top of a land agent.
As I was making the model, I noticed a few problems with the polluted plant agents. The first problem was that the polluted plant agents couldn't spawn in a random direction, which was a major problem that that made me redo my model. The second one was that the polluted plant agents kept stacking on top of each other when they reached the end of a column. This problem kept breaking the AgentCubes model after a few seconds and it crashed the AgentCubes software every time. From this model, I've learned that in order to make an AgentCube model, the story has to be broken down in simpler terms for the program to understand.The AgentCubes program can't do what it's told if it can't understand the directions it's given.
I made this other model with Vensim. Vensim allows you to simulate a graph of one or more populations that are affected by variables that the user can change. In my Vensim model, there is a fish population and a plant population being modeled on the graph. The two populations each have a birth rate and a death rate. There is also a fertilzer variable that's affected by the amount of fertilzer the farm releases, the amount of fertilzer that's blocked, and the effectiveness of the fertilizer itself. The Fertilzer varible affects the birth rate of the plant population and the actual variable for the plant population while also affecting the death rate of the fish population. When the model is run in SyntheSim mode, the birth rates, the death rates, the farm variable, the dam varible, and effect of vertilzer varible are all adjustable by using the slider bars under them.
When I was making the Vensim model, there was only one major problem with the model and that was setting the parameters. Originally, the Vensim model was breaking so easily after changing the slider bars by a little bit. When I saw this happening, I knew that I had to adjust to the sensitivity of the the Vensim model in order for the model to be more realistic. After I spent a few days adjusting the parameters, I finally got the model to stop breaking with the new parameters. I've learned from making this Vensim model that random numbers can't be used in for a model in Vensim to actually make sense. The parameters have to be precise for making the poputlations on the graph change as expected and any wrong equation or number can mess up the entire model.