Analysis of extreme weather events in Arctic region

1
Analysis of extreme weather events in Arctic
region

• Dag Johan Steinskog
• Nansen Environmental and Remote Sensing Center

2
Introduction

• My motivation
• Identify and explain extremes in the Arctic
  region in gridded datasets
• Try to explain future changes in extremes and
  explaining why using GCMs.
• Why Arctic region?
• Impacts on ecosystems and humans
• Explaining results of global warming
• Fundamental understanding of high latitude climate

3
Extremes what is it?

• From a statistical perspective
• Unique technique develop models for describing
  the unusual rather than the usual
• Looking at the tails of a distribution rather
  than the whole distribution
• There is a need of enough data
• Both minimum and maximum extremes are possible to
  look at (lower and upper tail)

4
Tail of a distribution
NASA Goddard Institute for Space Studies (GISS)
and Columbia University.
5
Two main techniques

• There are two main techniques for identifying
  extremes in
• extreme value analysis
• Generalized extreme value (GEV) family of extreme
  distributions
• Dividing the data into blocks
• Problem identifying size of blocks
• Generalized pareto distribution (GPD)
• Analysis of excesses of a given threshold
• Problem in which threshold to be selected

6
Methodology

• Three approaches on analysing extremes on
• gridded datasets
• Extreme value analysis of the grid points
• Coles (2001), Kotz and Nadarajah (2000) and
  Beirlant et al. (2004)
• Analysis of climate indices
• Easterling et al. (1997), Frich et al. (2002) etc
• Statistical testing (t-tests, goodness-of-fit
  tests, etc)

7
Possible extremes in climate

• The extremes in climate are normally divided into
  two
• different groups (IPCC 2001)
• Simple extremes (single variable)
• Heavy rain- and snowfall
• Heat spell
• Cold spell
• Complex extremes (critical combination of several
  variables)
• Cyclone (rainfall and wind)
• Drought (little precipitation and high
  temperatures)

8
(No Transcript)
9
Definitions of extremes in climate

• IPCC 2001 defintion of an extreme event
• An extreme weather event is an event that is
  rare within its statistical refrence distribution
  at a particular place. Definitions of rare
  vary, but an extreme weather event would normally
  be as rare or rarer than the 10th or 90th
  percentile

10
Not the only definition.

• Other IPCC TAR 2001 definitions
• Extreme climate event
• An average of a number of weather events over a
  certain period of time which itself extreme (e.g.
  rainfall over a season)
• Simple extremes
• Individual local weather variables exceeding
  critical levels on a continuous scale

11

• Complex extremes
• Severe weather associated with particular
  climatic
• phenomena, often requiring a critical combination
  of
• variables

12
Scientific questions so far

• What is the spatial structure of extremes?
• How is the large scale flow connected/affecting
  the extremes?
• How is the over all static stability influence
  extremes?
• Sea ice extent and extremes?
• What about past and future trends?

13
Why are extremes interesting?

• Heat wave summer 2003 in Europe
• 14000 excess deaths estimated by French
  authorities
• ICRC World Disasters Report Ch. 2
• Heat waves in winter time in the Alps
• Economical problems because of lack of tourism in
  ski resorts
• Beniston (2005)
• The hurricane Katarina New Orleans
• Disaster that will probably cost several billions
  USD and years to rebuild the city

14
The first experiments The spatial structure?

• Data NCEP/NCAR Reanalysis 6 hour mean
• Northern Hemisphere (20-90N, -180 180E)
• Any trends over ten days filtered out
• Ivar Seierstad lowpass filter
• Time period 1948-1953
• Season winter (120 days from October 1st)
• Threshold of 90-percentile of the data at each
  gridpoint selected
• Pareto distribution fitted to the tail.

15
(No Transcript)
16
(No Transcript)
17
Which factors are important for the creation of
extremes?

• Dynamical factors
• Pressure?
• Temperature gradients?
• Weather regimes
• NAO/AO?
• Scandinavian Pattern?

18
Connection to NAO?

• Beniston (2005) pointed out a possible connection
  between NAO and heat waves in winter time in
  Switzerland
• Difference between the period 1901-1970 and
  1970-2001 detected

19
Future climate?

• Warmer, wilder and wetter?
• There have been a significant decrease in the
  number of of frost days and increases in the
  number of very warm nights over much of the
  Northern Hemisphere (Kiktev et al. 2003)
• Will based on the results of the former questions
  try to identify and look at extremes in future
  (and past) climate
• Climate models
• Variables Pressure and temperature

20
Summary

• Extremes in climate are of great interest.
• A statistical approach will provide useful
  diagnostics about the extremes

21
References

• Stuart Coles (2001) An Introduction to
  Statistical Modeling of Extreme Values
• Samual Kotz and Saralees Nadarajah (2000)
  Extreme Value Distributions Theory and
  Applications
• Jan Beirlant et al. (2004) Statistics in
  Extremes Theory and applications
• Beniston (2005)