Models are used to imitate or mimic the real life or field conditions. They can be mathematical, computational, graphical or conceptual in nature. Models are used to simulate reality and make predictions.
The major limitation of models is that they are 'idealizations' or 'simplification' of reality and thus cannot possibly replace reality. A number of assumptions are made during modeling and this causes differences between model and reality. For example, while studying movement of rain water from ground to subsurface aquifers, one can model the soil particles as spherical and uniformly distributed (which we know is far from truth, since soil particles are not spherical, they come in all shapes and sizes), etc.
A model is also only as good as the information you use to build it. If we use simpler information or wrong information, it will give us wrong results. In other words, the model input governs its output, no matter if it is correct or incorrect.
Hope this helps.