The Power of Models and Modeling
Models have evolved since the turn of the century. There are models for most any item or system that exists. Developments in mathematics have enabled the world of models to explode. Now most everything can be modeled using some sort of mathematics. Agriculture is field that must rely on predictions. The Farmers Almanac is a source of information that allows one to predict upcoming events. The data in the Almanac is used to model events, especially hydrologic events that can occur in the future. The models that are created today are complex and involve numerous computations that only a computer can do. These models are used in a wide variety of tasks. Models have long been a vision to people, now they are a reality, and have been used to create systems that produce cost-efficient and effective answers to everyday problems.
RESTRICTIONS ON MODELING
As with life, models come with restraints. These constraints must be weight to see the proper “give and take” ratio to ascertain a compatible model. There are numerous restraints exist among modeling. The obvious is the budget of the project. The more money available, the precision of the model increases. The use of complex mathematical relationships with multiple iterations leads to the need of vast amounts of computer time. The trade off is the computer time and cost is marginal compared the salary of an individual that is capable to do these mathematical computations. Educated personnel with extensive knowledge and understanding of the model are necessary to determine if the data created makes sense. Finally a model is of no use if there is not real world data to simulate. Simulation of this data checks the validity of the model and aids in its marketability (Johnson 1973).
CLASSIFICATION OF MODELS
Models are classified by their respective use. Watershed models can be classified by the process description, scale and technique of solutions. The process description can be further divided into lumped and distributed. A lumped model does not take into account spatial variability of processes, whereas the distributed model does take these into consideration. The processes can be divided into deterministic, stochastic, and mixed models (Johnson 1973).
The second way models can be classified is by space and time scales. The subclass of space scales is distributed, and then that is divided into small, medium, and large watersheds. The time subclass is divided into event-based, continuous-time and large time-scale classes. Models can be described by the solution technique. It can be either numerical analog, or analytical. The numerical can be related by; finite differences, finite elements, boundary element, boundary-fitted coordinate or mixed (which is just a combination of previous terms) (Johnson 1973).
FOREST HYDROLOGY (Plymouth, NC)
Multiple models have been developed to determine the...