Trends in Long Term Solar Activity

1
Trends in Long Term Solar Activity

• Jenna Rettenmayer
• 2005 Solar Physics REU
• Montana State University

2
Why do we care about the sun?

• Understanding solar activity is important for
  prediction of space weather that impacts
  satellites and power grids on earth
• The climate on earth may be affected by solar
  activity in the long term
• The sun is neat

3
What is the goal of my project?

• To understand long term solar activity trends
• To do so, we looked at sunspot data and solar
  proxy data, such as 10.7 cm flux, total solar
  irradiance, and cosmic ray flux
• The first step is to make sure we get well known
  solar cycles out of short term data

4
Data Analysis

• All data I used is monthly averaged
• Data has been smoothed with a 13 month smoothing
  algorithm
• (R-6)/2 (R-5) R (R5) (R6)/2 /
  12
• Unsmoothed monthly averages used to find
  periodicities via fast fourier transforms and
  power spectrum analysis

5
Fast Fourier Transforms

• IDL function FFT is defined as
• Results in a complex array that doesnt do us
  much good by itself
• But

6
Power Spectrum

• A power spectrum tells us the dominant
  frequencies of a time series
• Each time series has a true power spectrum that
  can be estimated by the square of the modulus (or
  absolute value) of the Fast Fourier Transformed
  vector
• The resulting plot of frequency vs. power is
  called a periodogram

7
Example Power Spectrum
8
Statistics

• Found 99 confidence limit with a bootstrapping
  method
• Randomly shuffle each data set, perform FFT, and
  find tallest peak on power spectrum
• Do this 10,000 times and the 100th tallest peak
  is the cutoff for the believable peaks in the
  Power Spectrum

9
Well Known 11 Year Cycle
10
SIDC - RWC Belgium

• 250 years of data (1750-2005)
• Monthly Averaged and Smoothed data
• Smoothing algorithm comes from this data
• Data from RWC Belgium World Data Center for the
  Sunspot Index

11
Results

• See 11 year periodicity as well as Gleissberg
  cycle

12
10.7 cm Radio Flux

• 50 years of data (1947-2004)
• High correlation between sunspot numbers and 10.7
  cm Radio Flux (r .995)
• Data from Dominion Radio Astrophysical
  Observatory courtesy of National Research Counsel
  of Canada

13
10.7 cm Radio Flux
Radio flux lags sunspot numbers by 4.5 months on
avg.
14
R0.995
15
Results

• 11 year cycle is very evident in 10.7 cm Solar
  Radio Flux data

16
Cosmic Ray Flux

• 50 years of data (1953 - 2005)
• Anti-correlation (r -.879)
• Sunspot minima gt greater magnetic dipole on the
  sun gt more protection for our solar system gt
  fewer cosmic rays bombard earth
• Time required for magnetic field to permeate
  heliosphere ( 1/2 solar cycle)

17
Magnetic Butterfly Diagram
http//science.nasa.gov/ssl/pad/solar/sunspots.htm
18
Interplanetary Magnetic Field
http//solarsystem.nasa.gov/multimedia/display.cfm
?IM_ID282
19
Cosmic Ray Flux
Cosmic Ray Flux lags SSN by an average of 4.7
years
20
R -.879
21
Results

• 11 year cycle is evident in Cosmic Ray Flux data
  as well

22
Total Solar Irradiance

• 26 years of data (1978 - 2004)
• PMOD composite data averaged monthly
• Positive correlation to sunspot numbers (r
  .957)
• Data courtesy of the World Radiation Center

23
TSI
TSI lags sunspots by an average of 3.2 months
24
R .957
25
Results

• We do not see the 11 year cycle in this data, but
  only a 9 year cycle
• Artifact of the data?

26
What next?

• Look at reconstructed time series on the scale of
  thousands of years to search for longer solar
  cycles
• Where do we get reconstructed time series?
• 10-Be and 14-C

27
Beryllium 10 Isotope

• Formed in earths upper atmosphere by cosmic rays
• Deposited and stored in ice on earth
• Ice cores allow us to reconstruct cosmic ray
  activity (and hence solar activity) for thousands
  and millions of years
• 1.6 million year half-life

28
Carbon 14

• Cosmic rays strike Nitrogen 14 and produce Carbon
  14 in the atmosphere
• Half life of 5730 years
• Stored in living things (e.g. plants)
• Can be difficult to find good samples
• Thus, we have fewer years of data than from 10-Be

29
(No Transcript)
30
for the feature presentation

• Perform same analysis on these reconstructed time
  series to look for solar activity cycles on the
  order of thousands of years
• Then we can begin to predict solar activity in
  the long run

31
A Big Thank You To

• Charles Kankelborg and Dibyendu Nandi for being
  supercool advisors
• Dick Canfield for being The Man
• The other REU students for a being the coolest
  people in the visible Universe
• Trae Winter for his expert guidance on Beer and
  Wine

32
The Sun, with all the planets revolving around
it, and depending on it, can still ripen a bunch
of grapes as if it had nothing else in the
Universe to do. Galileo Galilei