Scientific computing

Scientific computing uses algorithms and computers to model, analyze, and solve scientific problems, often with large datasets or computationally intensive simulations.

Numerical analysis: The study of algorithms for approximating mathematical problems and controlling their computational errors. It explains accuracy and stability limits in scientific calculations.
Computational fluid dynamics: The numerical simulation of fluid motion and its interactions with surrounding surfaces or forces. It applies numerical algorithms to predict flows that are difficult to study directly.
Mathematical modeling: The formulation of real-world systems as mathematical structures that can be analyzed or simulated. A computational solution begins with a mathematical description of the scientific system.
Analytical solution: An exact mathematical expression that satisfies a problem's equations under specified conditions. Numerical computation is often used when an analytical solution cannot be obtained.
Computational science: An interdisciplinary approach that uses computation as a central means of scientific discovery and investigation. It treats computation not merely as a tool but as a mode of scientific inquiry.
Numerical method: A computational procedure that approximates the solution to a mathematical problem. Scientific models often require numerical methods when exact solutions are unavailable.
Weather forecasting: The prediction of atmospheric conditions using observations and models of the atmosphere. Operational forecasts depend on repeated, large-scale numerical weather simulations.
Algorithm: A finite, well-defined sequence of steps for solving a problem or performing a computation. Algorithms specify the procedures that transform models and data into results.
Analog computer: A computer that represents quantities through continuously varying physical properties. Digital scientific computing uses discrete representations rather than continuous physical analogies.
Reproducible research: Research practices that make methods and evidence available so results can be independently reproduced. Reproducible computational work requires accessible code, data, and execution details.

Linked from 17 pages