Spatial and temporal mapping of temperature variability in Iceland since the 1870s

1
Spatial and temporal mapping of temperature
variability in Iceland since the 1870s

• Halldór Björnsson
• Icelandic Meteorology Office
• Co-author
• Trausti Jónsson

2
Structure of talk

• Anomaly maps
• 40 year timeseries of temperature anomalies
• 140 year extension using Stykkisholmur
• Recent warming in Iceland
• Temporal and spatial pattern
• Early 20th Cent warming
• 120 years of monthly anomaly maps on a coarse
  grid
• Temporal and spatial pattern of warming
• Discussion

3
Introduction

• How do we assess regional temperature change?
• Using 85 manned weather stations maps of monthly
  anomalies from 1961 to 2000 climatology
• Areal average yields timeseries of mean
  temperature anomalies
• Maps give insight into spatial pattern
• Station data from select stations can be used to
  extend the anomaly timeseries.

4
Data

• Data comes from 85 manned weather stations
  collected during 1961 to 1990 (blue circles)

5
Example of anomaly map

• Anomaly map for January 1992, and archived
  hand-drawn map for comparison

6
Anomaly maps - Long records

• For each month from 1961 to 2000 an anomaly map
  is calculated
• 12 x 40 480 maps.
• Anomaly is the departure from the 1961 - 1990
  climatology
• Anomaly maps can be used to construct timeseries
  of of the areally averaged monthly mean anomaly
• Also have a few stations with data extending into
  the 19th century.
• Ex. Stykkishólmur Starting in1823
• These overlap from 1961 to 2000

7
Evolution of areal average of anomalies
8
Anomalies from Stykkishólmur
9
Comparison of anomalies monthly (480 pts) and
annual (40 pts)
10
Comparison of anomalies 40 summers Winters
11
Timeseries of temperature anomalies for Iceland

• Using the linear relationships for the 40 year
  period 1961 to 2000, the anomaly timeseries can
  be extended to 1823.
• These will be similar to the timeseries from
  Stykkisholmur, but will also have a confidence
  envelope

12
Annual anomaly timeseries
13
Winter and summer timeseries
14
Two warming episodes

• The extended timeseries shows
• Warming episode in the 1920s
• Warming episode in the last 20 years
• The anomaly maps allow us to examine the spatial
  pattern of the latter one

15
Warming of the last 20 years
16
Always greater warming in The N and NE parts of
the island
17
Spatial pattern of recent warming

• All of the island warms
• North and Northeastern (depending on season)
  parts warm more.
• Resembles Föhn effects

SAT difference between NE and SW Iceland is often
observed (EOF 1) Warming pattern consistent with
warm SW moist winds
18
Early 20th Cent warming

• Sparser observation network
• Can use sparse network to calculate anomaly maps
  from 1880s onwards
• Can use sparse network to reproduce the time
  series of temperature anomalies
• Smaller net anomalies on sparse network than on
  dense network
• Timeseries of anomalies calculated from the
  sparse network remains within errorbounds on
  previous timeseries
• Sparse network reproduces the main features in
  the spatial signal of the recent warming

19
A few good stations
20
(No Transcript)
21
Anomalies Dense vs. Sparse
22
Annual anomaly timeseries
23
Warming of the last 20 years (sparse grid)
24
The last century at a glance
25
Spatial pattern of 1920s warming

• To examine the pattern of warming in early part
  of the 20th century a 10 station grid was used
• These stations were in continuos operation
  throughout the century

26
(No Transcript)
27
Spatial pattern of warming (phase 1)
More warming in the W SW parts!!
28
Winter and summer pattern of warming (phase 1)
Notice that in neither case is the warming
enhanced in the NE
29
Spatial pattern of warming (phase II)
More warming in the NE
30
Winter and summer pattern of warming (phase 2)
Warming in N and NE
31
Discussion

• Temperatures increased in the 1920s and also in
  the last 20 years
• Spatial pattern for the last 20 years suggests
  warmer and moister south- westerlies
• Spatial pattern for the earlier warming is more
  complex
• Early part sees more warming in S and SW
• Latter part has a pattern more similar to that
  above

32
Discussion Station histories

• There is no obvious evidence of a spatial
  progression of anomalies
• Same goes for SST data that has been tentatively
  analyzed

33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
Discussion Warm, moist south-westerlies

• One can speculate what makes SW winds more moist
• Higher frequency of these wind directions
• Same frequency but warmer winds have cooled less
  en-route
• Hence warm moist!
• Warming could be due to
• Less radiative cooling (a la enhanced green house
  effect)
• Warmer SSTs south west of Iceland
• Has precipitation been enhanced?

38
The End

• Stop!

39
Warming of the last 20 years (10 station grid)