ATMO 632: Week 2
Time Series Analysis:
Regression & Principal Component Analysis
URL: http://www.cgd.ucar.edu/~svn/atmo632/week2.htm
Reading Material
- Wilks:
Chapter 6.2 (regression), 9.1-9.3 (Principal Component Analysis)
- North:
Chapter 2 (regression), 3.1-3.2
(Empirical Orthogonal Functions)
Concepts
Regression
Uncertainty in one variable
Least square curve fitting
Correlation
“Explained” variance
t-statistic (null hypothesis of zero
correlation)
z-statistic (null hypothesis of non-zero correlation)
Multiple regression
Uncertainty in both variables
(PCA/EOF)
Matrix
algebra
Norm
Eigenvalues and Eigenvectors
Eigenvectors of a real symmetric
matrix
Singular Value Decomposition
Principal
Component Analysis/Empirical Orthogonal Functions
Covariance matrix
Scaling conventions
ATMO 632: Homework for week 2 (due Sep. 14)
NOTE: You can obtain the datasets
used below by visiting the course home page
http://www.cgd.ucar.edu/~svn/atmo632
and clicking the appropriate links.
You do not need to type in the
URLs given below, which are provided for reference.
- Obtain
the winter (Dec-Mar) time series of the North Atlantic Oscillation (NAO)
index for the period 1864-2003 from the URL http://www.cgd.ucar.edu/~svn/atmo632/nao-djfm-1864-2003.txt
This is a text file which contains two columns (year of the January, and
NAO index value). The NAO index is computed as the difference of
normalized sea level pressure between Lisbon, Portugal and
Stykkisholmur/Reykjavik, Iceland, and is described in http://www.cgd.ucar.edu/~jhurrell/nao.stat.winter.html
- Plot
the time series as a function of time
- Compute
the mean and standard deviation.
- Obtain
the winter (Dec-Feb) time series of the El Nino (NINO3) index for the
period 1851-2003 from the URL http://www.cgd.ucar.edu/~svn/atmo632/nino3-djf-1951-2003.txt
This is a text file which contains two columns (year of the January, and
NINO3 index value in degrees Centigrade). The NINO3 index is computed as
the average of sea surface temperature in the region (150W-90W, 5S-5N),
and is obtained from http://www.cdc.noaa.gov/ClimateIndices
- Plot
the time series as a function of time
- Compute
the mean and standard deviation.
- Restrict
the NAO index to the period 1951-2003, which overlaps with the NINO3 index
period. Normalize the NAO and NINO3 indices by subtracting the mean and
dividing by the standard deviation.
- Plot
both time series in the same plot, using two different line styles or
colors for the two indices
- Display
the two time series in a scatter plot, with the NINO3 index on the X axis
and the NAO index on the Y axis
- Compute
the regression of the NAO index on the NINO3 index
- Compute
the correlation between the two indices
- Test
the null hypothesis that the two indices are uncorrelated, with 95%
confidence limits