Introduction to Spectral Analysis光普分析导论.pdfVIP

Introduction to Spectral Analysis Don Percival, Applied Physics Lab, University of Washington • Q: what is spectral analysis? • one of the most widely used methods for data analysis in geophysics, oceanography, atmospheric science, astronomy, engineering (all types), . . . • method is used with time series • let xt denote value of time series at time t; for example, x10 = 43◦ = temperature at GHS at 8AM on 10th day of 2003 • four examples of time series x , x , . . . , x , x 1 2 127 128 4 (a) 15 (b) xt −1 0 −6 −15 12 (c) 7 (d) xt 10 0 8 −7 0 64 128 0 64 128 t t • Q: how would yo udescribe these 4 series? • spectral analysis describes xt ’s by comparing them to sines and cosines 1 Sines and Cosines: I • Q: what do sines and cosines have to do with time series? • plots of sin(u) and cos(u) versus u as u goes from 0 to 4π 2 0 −2 0 2π 4π 0 2π 4π u u • let u = 2π 2 t for t = 1, 2, . . . , 128 128 • plots of sin(2π 2 t) and cos(2π 2 t) versus t 128 128 2 0 −2 0 64 128 0 64 128 t t • can regard above as artificial time series: st = sin(2π 2 t) etc