Western Atlantic Ridge
Global Models tend to have a much lower resolution than Mesoscale models. Global models in the long range have a variety of issues. To simplify things, lets look at it this way: Global models are used to show the thousand foot view of all the weather players which are interacting with each other. Mesoscale models, on the other hand, are used to look at the finer details and to help pinpoint small scale features such as banding during snow storms. Mesoscale tend to be short range models, however, due to the fact that they tend to have a much higher resolution than that of their counterparts (globals). This resolution, though, gives rise to several issues. First, its the chaos theory. Within this theory is the butterfly effect which states that a simple mistake early on leads to significant mistakes later on. With higher resolutions, these mistakes are amplified further due to the fact that higher resolutions use a much higher computation speed (more equations). On the other hand, while globals may not have these limitations, they do not have the resolution to "see" many of the intricacies which the meso models can. As such, globals tend to miss things which are limited by their resolution. This, consequently, allows globals to miss intricacies which can likewise lead to large discrepancies towards the end of the run, though this does not happen as often.