|
气候与环境研究 2000
Dealing with Imperfect Data to Improve EstimationPrecision of Turbulence Flux
|
Abstract:
The observational data are seldom perfect, they are usually short noisy with wild points and nonlinear trends. In this paper, we talk about how to delete noise, wild points and filter nonlinear trends, how nonlinear trends contribute fake turbulence flux by having influence on self-relation func- tion and integral scale.