Predictability and Nonlinear Modelling in Natural Sciences and Economics [Paperback]

Researchers in the natural sciences are faced with problems that require a novel approach to improve the quality of forecasts of processes that are sensitive to environmental conditions. Nonlinearity of a system may significantly complicate the predictability of future states: a small variation of parameters can dramatically change the dynamics, while sensitive dependence of the initial state may severely limit the predictability horizon. Uncertainties also play a role.
This volume addresses such problems by using tools from chaos theory and systems theory, adapted for the analysis of problems in the environmental sciences. Sensitive dependence on the initial state (chaos) and the parameters are analyzed using methods such as Lyapunov exponents and Monte Carlo simulation. Uncertainty in the structure and the values of parameters of a model is studied in relation to processes that depend on the environmental conditions. These methods also apply to biology and economics.
For research workers at universities and (semi)governmental institutes for the environment, agriculture, ecology, meteorology and water management, and theoretical economists.
1. Geophysics.- Karl Popper and the accountability of scientific models.- Evaluation of forecasts.- The Liouville equation and prediction of forecast skill.- An improved formula to describe error growth in meteorological models.- Searching for periodic motions in long-time series.- Comparison study of the response of the climate system to major volcanic eruptions and el nino events.- Detection of a pertubed equator-pole temperature gradient in a spectral model of the atmospheric circulation.- A simple two-dimensional climate model with ocean and atmosphere coupling.- Climate modelling at different scales of space.- 2. Agriculture.- Simulation of effects of climatic change on cauliflower production.- Validation of large scale process-oriented models for managing natural resourclƒB