SO503 Spring 2005

Test 2

Exam makeup

 

Short answers

4 of 5, 10 points each

40

 

Problems

4 at 15 points each

60

 

Total

 

100

 

   


Short Answers:  Answer 4 of the following 5 of the questions with a few sentences.  Each is worth 10 points.

 When is lossy compression worthwhile?

 Why are so many scientists concerned about the possibility of a manned mission to Mars?

 Why has EOS had so many references to the great megathrust earthquake of 2004?

 What events can cause tsunamis, and what are the challenges in setting up warning systems?

 How can tidal signals show up in rocks millions of years old?  


The graph below shows a time series, with a reading taken every day for 1255 days.

 

Problem 1, 15 points.  The next page shows the results of an FFT run on the time series.

Results of an FFT run on the data, with the bins sorted by Power (top) and Period (bottom).

 

 

 

Bin    Frequency        Period              Power

    1     0.0            infinite         83.638870

    6   0.00293        341.333333          5.451275

    5   0.00244        409.600000          5.094719

    3   0.00146        682.666667          5.062107

    4   0.00195        512.000000          3.227708

  293   0.14307          6.989761          3.187808

    2   0.00098       1024.000000          2.551096

  292   0.14258          7.013699          2.136037

   10   0.00488        204.800000          1.647327

    9   0.00439        227.555556          1.577684

  585   0.28564          3.500855          1.176392

  294   0.14355          6.965986          1.150024

    7   0.00342        292.571429          0.967016

   26   0.01270         78.769231          0.757571

  586   0.28613          3.494881          0.667365

 

 

 

Bin    Frequency        Period              Power

    1     0.0            infinite         83.638870

    2   0.00098       1024.000000          2.551096

    3   0.00146        682.666667          5.062107

    4   0.00195        512.000000          3.227708

    5   0.00244        409.600000          5.094719

    6   0.00293        341.333333          5.451275

    7   0.00342        292.571429          0.967016

    8   0.00391        256.000000          0.351580

    9   0.00439        227.555556          1.577684

   10   0.00488        204.800000          1.647327



Problem 2, 15 points.  This shows the result of fitting Fourier curves to try to replicate the time series, sorted by amplitude.  Note the large number of periods near 1 day, the sampling interval.  Are these meaningful according to Nyquist?

 

 

Component   Period   Phase    Amp   % SS  Cum % SS

 

     3     418.333   141.2  20.703   5.485    5.49

  1252       1.002    38.8  20.703   5.485   10.97

   179       7.011   300.1  20.513   5.385   16.36

  1076       1.166    59.9  20.513   5.385   21.74

  1253       1.002    34.0  19.102   4.670   26.41

     2     627.500   146.0  19.102   4.670   31.08

  1075       1.167    86.8  19.003   4.621   35.70

   180       6.972    93.2  19.003   4.621   40.32

  1254       1.001   306.6  17.947   4.122   44.45

     1    1255.000    53.4  17.947   4.122   48.57

  1251       1.003   279.2  15.631   3.127   51.70

     4     313.750    80.8  15.631   3.127   54.82

   358       3.506     2.4  12.341   1.949   56.77

   897       1.399   177.6  12.341   1.949   58.72

  1239       1.013   312.6  11.953   1.829   60.55

    16      78.438    47.4  11.953   1.829   62.38

 

 


Problem 3, 15 points.  Explain what this graph showing an autocorrelation means.   The second graph is a blow up of the left side of the first graph.  Could it be the same data set seen in the last problem?  Why or why not.

  

 

 


Problem 4, 15 points.  The graphs below show the trends of average yearly and monthly sea level at Annapolis from approximately 1930 to 2000, from  pol.ac.uk