The Effect of Stochastic Noise on Predictability
Abstract
LONG-TERM GOAL. Our long-term goal is to improve the accuracy of numerical prediction models of weather and climate. We shall concentrate on sources of prediction error involving interactions between physical phenomena having different timescales. OBJECTIVES. We examine how inadequate representation of rapidly-varying (i.e., stochastic) atmospheric effects in General Circulation Models (GCMs) can systematically affect prediction on a variety of scales, and how errors arising from this inadequacy may be ameliorated. Concurrently, we develop useful, efficient and accurate methods of accounting for these systematic effects of stochastic forcing, and also provide methods for estimating the spread of ensemble predictions taking into account multiplicative stochastic effects. APPROACH. We approach the problem both analytically and numerically. Simple barotropic models present cases that are nontrivial, yet simple enough to solve analytically, and we use these models to identify and solve some of the numerical problems that will arise in realistic, more complex models. We intend to apply our results to studying the impact of stochastic forcing in more complete models, culminating with long runs of stochastically forced, full-fledged GCMs and numerical weather prediction (NWP) models. Although routinely ignored in current models, the theory of stochastic white noise is well known and analytic solutions are available, at least in principle. We have evaluated the effect of white stochastic forcing on the mean response of a global barotropic model to steady forcing. In addition, we have considered the more realistic problem of stochastic colored noise forcing, deriving analytic estimations of the systematic change in the mean response, and have compared these analytic approximations with results from stochastically forced numerical models.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 30, 1999
- Accession Number
- ADA630131
Entities
People
- Cecile Penland
- Matthew Newman
- Prashant Sardeshmukh
Organizations
- National Oceanic and Atmospheric Administration