Transcript FMI

Assessment of fire risk
in boreal forests
under the present-day
and future climate
Andrea Vajda, Ari Venäläinen
and Kirsti Jylhä
Finnish Meteorological Institute (FMI)
ENSEMBLES Annual Meeting, Lund, 20-23 Nov 2006
WP6.2 – Linking impact models to probabilistic scenarios of climate
Outline
• FMI contribution to deliverable D6.9 “Report on
an intercomparison study of modelled, Europe-wide
forest fire risk for present day conditions”
by Giannakopoulos et al. (NOA & FMI) (month 24)
• The impact of climate change on forest fire risk
in northern Europe: some first estimates
• Future plans in ENSEMBLES
Finnish fire statistics 2001-2006
Annual number of fires and area burned
7000
2500
No. of fires
Area burnt
No. of cases
5000
1500
4000
3000
Number of fires
60
50
40
30
20
10
0
Apr 1 May 1 June 1 July 1 Aug 1 Sep 1 Oct 1
Average daily area burned
140
120
100
80
60
40
20
2000
Burned area (ha)
6000
70
Area burned (ha)
• The average annual number of fires:
about 3100
• The average annual area burned:
about 800 ha
=> The mean area burned per event: 0.25 ha
18% of fires ≥ 0.1 ha
0.3% of fires ≥ 10 ha
Average daily number of fires
in April - October
0
Apr 1 May 1 June 1 July 1 Aug 1 Sep 1 Oct 1
frequency of fires is
Boreal forests coverThe
nearly
highest and the total area
78% of total land area in
Finland
burned
is largest in May
1000
2000
500
1000
0
0
2001
2002
2003
2004
2005
2006
Photo
© A. Drebs
Photo
© Krister
Sanmark
Smoke in Helsinki in summer 2006 due to forest fires from abroad
One hour later
(a shift in wind direction)
Photo © Pia Anttila
Helsinki 9 Aug 2006 at 1 pm
Photo © Pia Anttila
Helsinki 9 Aug 2006 at 2 pm
Evaluation of two forest fire danger indices
in the boreal forests environment (Finland)
 The Finnish Forest Fire Index (FFI):
- Soil surface moisture as an indicator of the fire risk
 The Canadian Fire Weather Index (FWI):
- Forest fuel moisture content; adjustment to the Finnish conditions
Fire danger
classes
FFI (6 classes)
FWI (5 classes)
6
>25
very high
extreme
5 – 5.9 high
18 – 24.9 high
4 – 4.9 moderately high
10 – 17.9 medium
3 – 3.9 moderately low
2 – 2.9 low
1 – 1.9 very low
2 – 9.9 low
0 – 1.9 very low
Fire potential threshold for
Pinus sylvestris stand
Correlation between the Finnish and Canadian fire indices
based on station data in Finland in 1961–2005 (April–September)
A station in northern Finland
Canadian index FWI
Canadian index FWI
A station in southern Finland
Finnish index FFI
Finnish index FFI
• The forest fire risk decreases northwards
• In northern and central Finland FWI indicated a fire risk more often than FFI,
and vice versa in southern Finland
• FWI responded more rapidly to the meteorological variations
(i.e. precipitation) than FFI.
Monthly distribution of days with
FWI and/or FFI indicating fire risk
(FWI>10, FFI≥4; central Finland)
Relative frequency (%)
100%
90%
28
33
80%
70%
40
51
70
8
76
8
60%
7
50%
8
40%
63
59
30%
53
41
20%
8
10%
16
4
26
0%
Apr.
Apr
May
May
FWI and FFI
FWI and FFI
June
June
July
July
only FFI
only FFI
Aug
Aug
only FWI
Sept.
Sep
only FWI
• In about 50-60% of the cases both indices indicated a fire risk
• Largest deviations between FFI and FWI in early spring and in autumn
FFI
50
45
40
35
30
25
20
15
10
5
0
no fire risk
FFI < 4
a fire risk
FFI ≥ 4
size0.1-0.9
of the
fires:
ha
≥1 ha
0.1-0.9 ha
≥ 1 ha
Frequency
distribution (%) (%)
Frequency of distribution
distribution
Frequency
distribution (%)(%)
Frequency of
Comparison of fire indices and observed fire events
50
45
40
35
30
25
20
15
10
5
0
FWI
no fire risk
FWI < 3
a fire risk
FWI ≥ 3
size0.1-0.9
of hathe fires:
≥1 ha
0.1-0.9 ha
≥ 1 ha
• 54% of the fires ≥0.1 ha occurred when FFI indicated fire risk
• 64% of the fires ≥0.1 ha occurred when FWI indicated fire risk
 The fire indices predict the favourable meteorological conditions for
the occurrence of fires, but they are not predictors of fire events
Human behaviour, etc
The impact of climate change on forest fire risk
in northern Europe – a preliminary study
MAM and JJA changes (%) in the max 1-day precipitation (left)
and max length of dry spells (right) by 2071-2100
For R1d:
R1d
MAM
CDD
MAM
For CDD:
increase
increase
decrease
decrease
R1d
JJA
CDD
JJA
Multi-model means
based on
RCM-H-A2 runs
The impact of climate change on forest fire risk
in northern Europe – a preliminary study (cont.)
Temporal variation of the annual
number of days with a fire risk
(FFI≥4) in Helsinki during 1961-2100
based on:
observations
RCA3-E-A2 scenario
RCA3-E-B2 scenario
Latitude range in Scandinavia
and the Baltic countries
Changes in time of the average annual
number of days with a high or a very
high forest fire risk (FFI ≥ 5)
The annual number of days with a forest fire risk (FFI ≥ 4)
– preliminary results
2001–2025
Based on the RCA3-E-A2 simulation
The annual number of days with a forest fire risk (FFI ≥ 4)
– preliminary results
2026–2050
Based on the RCA3-E-A2 simulation
The annual number of days with a forest fire risk (FFI ≥ 4)
– preliminary results
2051–2075
Based on the RCA3-E-A2 simulation
The annual number of days with a forest fire risk (FFI ≥ 4)
– preliminary results
2076–2100
Based on the RCA3-E-A2 simulation
Future plans in the ENSEMBLES project
 To produce new estimates of the impact of climate change on
forest fire risk in northern Europe on the basis of
- various future climate projections and
- the Finnish Forest Fire Index (FFI)
 To evaluate the impact of the extreme climate events on
soil temperature and soil moisture
- Simulations using a coupled heat and mass transfer model
for soil-plant-atmosphere system* (COUP model)
- Measured meteorological data / climate models’ output data
- Findings to be compared against results from the Finnish
forest fire danger forecasting system
___
* Jansson, P.E. & Karlberg, L., 2001: Coupled heat and mass transfer model for soil-plant-atmosphere systems.
Royal Institute of Technology, Dept of Civil and Environmental Engineering, Stockholm, 321 pp.
Year-to-year variation of the onset and end of the fire season
Calendar day
FFI start date*
FFI end date*
FWI start date**
FWI end date**
* first and last days with FFI=4
** 5 consecutive days with FWI>11,
using a 7 days moving average
FWI appeared to respond very quickly to the variation of precipitation events,
while FFI indicated smoother changes.