Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Suggests for stuff to model

  1. Mar 5, 2013 #1
    Hi, I'm taking an ecology course and I need to model some kind of population/system for a final project. The problem is, I'm a physics major and know almost nothing about real biological systems, so I'm looking for suggestions of real systems that are simple enough to model fairly realistically in a school project.

    Well-studied systems with a good amount of literature behind them would be ideal, but anything I can find good data on will work. There's no real guidelines, it could be aquatic, terrestrial, whatever! It doesn't even have to be strictly biological. I'm open to suggestions, so if you have any ideas please share!
     
  2. jcsd
  3. Mar 5, 2013 #2

    jedishrfu

    Staff: Mentor

    predator - prey models like rabbit and wolf.

    THis one actually had a problem where the model didnt agree with reality due to the fact that human hunters weren't included in the mix.

    You could make it more exotic like antelope to cheetah...

    Also these models bring in chaos theory at some point...
     
  4. Mar 5, 2013 #3
    Well predator-prey is ok, but its been beat to death and a lot of them can be solved analytically - i'm looking for something a little more complicated.

    What I'm looking for though is suggestions about specific biological systems - some system I can model, then compare to real life data. I just don't know enough examples of biological systems have many to choose from on my own.
     
  5. Mar 5, 2013 #4

    jedishrfu

    Staff: Mentor

    Well how about looking into pathogens and modeling how they attack the body?
     
  6. Mar 5, 2013 #5
    The hard part of modeling is finding an appropriate data base. The models are generally simple, but finding data appropriate to a simple model is difficult. There are tricks for finding data to analyze.

    One thing that I find useful when searching for data is the “images” option in the google search. If you key in a quantitative subject, and then do an “image search”, what often comes up is a selection of graphs. Each graph has data. The data can be modeled any way you like. You don’t have to model it the way the author modeled it. In fact, it could be more fun to model it in a way that has nothing to do with the theory that the original investigator intended.

    Although the predator-prey model is over-used, it is very general. I don’t know what your programming skills are like. However, maybe you would prefer to vary an overworked problem rather than to construct a model from scratch.

    One thing you could try is an inverse problem. Start with a mathematical model that is fairly well known. Find a data set where the dynamics are unknown, but where the data vaguely resembles solutions to the well-known model. Then, try to find input parameters that approximate the data. Different ways of “trying out” parameters could be part of your report. The data you are modeling does not have to have a clear physical connection to the physical model.

    Predator-prey model is a classic. You know that strange oscillations are typical of the solutions. Find a data set with oscillatory data for any system at all. Try to fit the data set to by varying input parameters of the predator prey-problem. The data set does not have to have a known connection to predators or prey.

    This is a slightly different approach than what you are asking for. I am suggesting that you start with a predator-prey model, and search for a graph that can be “fitted” to it.


    I did a search using the image option on Google of “turbidity versus time”. I got a large selection of graphs. One graph that made me think of population studies was in a study of “aqueous mixtures of pectin versus chitosan.” The abstract is in this link.
    http://www.sciencedirect.com/science/article/pii/S0014305705001072

    You can write a predator-prey program, but keep the source code. You can vary any line of the program that you like.

    You can also generate solutions to the predator-prey problem. Make lots of graphs. Then, look through the literature for an experimental graph that happens to look like one of the graphs from your predator-prey program. Then read the article where the data came from, and then make up a physical explanation of why the two match.

    This is not cooking. Try to find new insight into the data this other investigator found. His model may be correct. However, any investigator can miss something obvious.

    One scientist that I know was working on a program using molecular dynamics to model chemical reactions. The reactions generally resulted in stable molecules. He made a copy of the program and erased an arbitrary line. The simulated molecules exploded. So he started a successful project simulating explosives.

    Maybe you could simulate explosives with a predator-prey model.
     
  7. Mar 7, 2013 #6
    Hi, dipole.

    I think you have an excellent opportunity here to make a very simple, but nevertheless interesting experiment.

    Consider the following quote: "The most exciting phrase to hear in science, the one that heralds new discoveries, is not 'Eureka!' (I've found it!), but "That's funny..." " -- Isaac Asimov

    Secondly, I would like to point out that a model in ecology can be extremely simple, but still impressive. One model that has fascinated me a lot is this one:
    http://img41.imageshack.us/img41/6106/1s20s1569904804000679gr.gif [Broken]
    You see imediately what is going on here, and yet you may be suprised, because perhaps you have never before considered this relationship.

    These two points contrasts e.g. the lotka-volterra model (predator-prey model), which can be quite difficult to grasp. It definitely is useful, but I feel that it lacks some of the suprising edge that simpler models have, which I would say makes it less interesting in educational settings.

    So from here, it can be quite easy to find an interesting (i.e. suprising) relationship, which is possible to model. For some relationships the data is easy to gather yourself, which is useful.

    Just an idea I had right now was to make a model of the relationship shown above, and see if I could predict the basal metabolic rate of other animals not listed here, based on their body size. And the first animal I would test would be the sloth (of course). And as predicted, it does have a basal metabolic rate much lower than expected, based on its body size ( source ). You wouldn't even have to limit yourself to mammals, e.g. to see if this relationship holds for plants or jellyfish.

    From here it would be relatively easy to make a good, and insightful essay. But my experiment was only meant as an example, as there are plenty of other relationships to explore. You say that you would like system to be a bit complicated? I suppose it can be as complicated as you want it. In my example, you could look at different taxonomic groups, different geographical regions, with or without fuhr, etc. Your imagination is your limit, not the biological systems :)

    Hopefully you found my thoughts on the matter interesting, at least, even if it might not have been the answer you were hoping for :)

    Regards,
    Adrian

     
    Last edited by a moderator: May 6, 2017
  8. Mar 7, 2013 #7

    atyy

    User Avatar
    Science Advisor

  9. Mar 7, 2013 #8

    Mute

    User Avatar
    Homework Helper

    What level course is this?

    I saw an interesting paper on the arxiv a while back: http://arxiv.org/abs/1301.5465

    It's a model of (gazelle) population dynamics taking into account resource gradients and communication ("acoustic interactions") between members of the herd. The model looks at what calling frequencies lead to the least amount of time it takes to find the region with the highest quality vegetation.

    The model is basically a Langevin diffusion model with drift/bias terms that cause the gazelles to move in the direction of higher resource quality or calls from other members of the herd. The gazelle aspect is actually just a case study - the model is applicable to other populations of animals, so if you could get some data you could compare their model predictions to data from other kinds of animals. You could also consider adding some more complicating factors to the model, such as birth/death processes, predators, or resource depletion over time (which are mentioned in the paper but not explored).
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Suggests for stuff to model
  1. Water stuff (Replies: 6)

  2. Suggest me please (Replies: 2)

  3. Smelly stuff? (Replies: 1)

Loading...