Dendroclimatology

1
Dendroclimatology
2
Dendroclimatology
Dendroclimatologists are interested in past
climate so that the variation and trend of modern
climate can be put into perspective Synoptic
climatology the study of climate from the
perspective of atmospheric circulation
circulation patterns can be inferred from
reconstructed patterns of precipitation, temp,
and pressure
Climate conditions can be estimated from a
variety of natural environmental records going
back to well before humans existed. Information
of this kind is called proxy data because it
substitutes for, or acts as a proxy for, the
actual data we are seeking.
3
Dendroclimatology

• Climate reconstruction starts with a site-level
  analysis of a tree species climate response
• Site selection trees chosen from sensitive
  sites
• Many ages of trees chosen because trees may
  change in their climatic response with age
• Older trees yield longer chronologies, but older
  trees may have a weakened climate signal due to
  senescence
• Injured trees or those in the understory are
  avoided
• Sample depth
• - Standardization removes non-climatic trends

4
Dendroclimatology
PRISM data- takes data from individual climate
stations and modles the signal ofer the landscape
based on a physiological model. This data set
provides accurate climatological info for
location not previously monitored Climate
division data from National Climate Data Center
5
Response function analysis
6

• Correlation analysis
• Correlation analysis complements results from
  response function analysis.
• RFA primarily concerned with temp and precip.
  Correlation analysis can be done on ALL climate
  variables (PDSI, ENSO, PDO, etc.)
• Correlation analysis best done with stats
  packages (SAS, Systat) or PRECON.
• Range of values -1.0 lt r lt 1.0
• Associated with each r-value is its p-value which
  tests for statistical significance.
• In general, we want p-values less than 0.05, or p
  lt 0.05.
• As in response function analysis, we also analyze
  months from the previous growing season (why?).
• As in response function analysis, we look for
  groupings of monthly variables to indicate
  seasonal response by trees.

7
Correlation analysis
Graphical output from PRECON. Any value above
0.2 or below -0.2 is significant.
8
Note how response function analysis (top) and
correlation analysis (bottom) are complementary
(but different).
9

• Correlation analysis
• R-values also known as Pearson correlation
  coefficients
• SAS output below r-value (top), p-value
  (middle), n size (bottom)
• How do you interpret negative correlations?

Pearson Correlation Coefficients
Prob gt r under H0
Rho0
Number of Observations lmayt ljunt
ljult laugt lsept loctt
lnovt -0.08019 -0.03131 -0.34233
-0.16914 -0.29516 -0.09849 -0.02712
0.4941 0.7897 0.0023 0.1414
0.0096 0.4071 0.8173 75
75 77 77 76
73 75
10
Reconstructed Climate Variables
1) Temperature
Trends in reconstructed Northern Hemisphere
temperatures.
11
Reconstructed Climate Variables
2) Precipitation
Trends in reconstructed precipitation for the
American Southwest.
12
Dendroclimatology
Teleconnection the interconnection of physical
parameters over long distances
Primary Variables - Temperature (ave, min,
max) - Precipitation (total) - Palmer Drought
Severity Index (PDSI) - Palmer Hydrologic
Drought Index (PHDI) Derived Variables - El
Nino Southern Oscillation (ENSO) - North
Atlantic Oscillation (NAO) - Atlantic
Multidecadal Oscillation (AMO) - Pacific Decadal
Oscillation (PDO)
13
ENSO
14
(No Transcript)
15
Pacific Decadal Oscillation

• Discovered late 1990s by salmon fisheries experts
• Long-lived El Niño-like pattern of Pacific
  climate variability
• Operates on longer 20 to 30 year periods
• Note when shifts occurred

16
Pacific Decadal Oscillation
Warm phase PDO
Cool phase PDO
17
North Atlantic Oscillation

• Large scale seesaw in atmospheric mass between
  the subtropical high and the polar low
• Dominant mode of winter climate variability in
  the North Atlantic region ranging from central
  North America to Europe
• Operates on 20 to 30 year periods

18
(No Transcript)
19
Atlantic Multidecadal Oscillation
20
(No Transcript)