Applied climatology report

1
Applied climatology report

• Presentation dates
• November 5th
• November 12th
• November 19th
• E-mail me preferred date and topic within next 2
  days
• Report due by Wednesday, November 25th

2
Automated synoptic methods II

• Principal component analysis and cluster analysis

3
PCs on San Juan, Puerto Rico
4
Map-pattern correlations (Lund, Kirchhofer)

• Keyday 1 selected as map with most significant
  correlations with other maps all maps removed
• Keyday 2, etc. follow from rest of set
• Finally, all days reclassified into Keyday they
  best correlate with
• Can also be done as a hybrid method

5
Akron SSC, July
6
Akron SSC, January
7
Fuzzy clusters

• Assignment of a probability of a case (day,
  station) belonging to a particular cluster

8
Levels of analysis

• Surface
• 850-mb
• 700-mb
• 500-mb
• 200/300-mb

9
Variables of analysis

• Upper-level
• Usually height
• Temperature
• Vorticity
• Moisture in lower levels

• Surface
• Temperature
• Dew Point
• Pressure
• Wind
• Pressure
• Visibility
• Cloud Cover

10
On synoptic climatology

• Simple does not translate into worse
  complicated does not mean better - Yarnal (1993)
• Corollary More variables does not make things
  better
• An important distinction to make is theoretical
  synoptic climatology vs. applied synoptic
  climatology

11
Merrill et al. 2005

• Compared SSC air mass frequency with physical
  activity across the US

12
Dyer and Mote (2005)

• Looked at SSC air mass frequency and snow
  ablation events in the US Plains.

13
Vaneckova et al. (2008)
14
Vaneckova et al. (2008)
15
Vaneckova et al. (2008)
16
Kimmel et al. (2009)

• PCA / CA on sea level pressure summed up
  frequency of clusters over winter
• Synoptic type frequencies over all of winter
  explained 54 of variability of spring runoff,
  compared to T and P explaining only 17.
  Correlations with zooplankton and fish
  populations also noted.

17
Hondula et al. (2009)

• Analyzed causes of high pollution episodes in
  Shenandoah Valley using SSC and trajectory

18
Hondula et al. (2008)

• Trajectories AND synoptic categories (SSC)
  together produce the best predictions for air
  pollution in the Shenandoah Valley

19
Pook et al. (2008)

• 500-mb SLP analysis of south Australia
  precipitation types and relation to planting
  times for wheat

20
Frank et al. (2008)

• Pine Rust Transport from California to New Mexico
• 500mb heights used for upper-level classification
• Lower level moisture and temperature conditions
  used to evaluate pine rust survival

21
Bednorz (2009)

• Extreme snow ablation episodes in Poland and
  northern Germany
• Selected 96 days where snow loss gt 5 cm and did
  k-means clustering of SLP

22
Bednorz (2009)
23
Huth et al. (2008)

• Correlated spatial patterns of heat events by
  Grosswetterlagen

24
De Pablo et al. (2007)

• Hospital admissions compared with Lamb-type SLP
  patterns in southern Spain

25
Seluchi and Chou (2009)

• Landslides in Brazil
• Examined mean meteorological fields on 39
  landslide days
• Minus several days which were due to isolated
  convective events

26
Seluchi and Chou (2008)
Later subdivided into frontal systems and
convergence zone
27
Jiang et al. (2005)

• NO2 and NO pollution levels in morning rush hour
• K-means clustering of 1000-mb heights

28
Jiang et al. (2005)

• Large scale and small scale components to
  pollution identified

29
Cheng et al. (2007)

• Evaluated freezing rain likelihood in climate
  change scenarios
• Did PCA/CA on 240 variables (reduced to 18 PCs)
  per day per station!

30
Bentley and Stallins (2008)

• Composite maps of excessive dew point events in
  the midwest and their evolution