Global Trends in LAI, FAPAR, Burned Area and Fire using Earth Observation

1
Global Trends in LAI, FAPAR, Burned Area and Fire
using Earth Observation

• Stephen Plummer (IGBP-ESA)
• Jing Chen and Feng Deng (U. Toronto)
• Philippe Ciais (LSCE)
• Nadine Gobron (JRC)
• Roselyne Lacaze (MEDIAS)
• Tristan Quaife and Martin De Kauwe (CTCD)

2
Scientific Context
Growth rate of carbon reservoirs
Fossil fuel emissions
Net UptakeOcean Land
Accumulation rate in ocean and on land
Accumulation rate in atmosphere
Large interannual variation in the annual
atmospheric CO2 growth rate
Source Sarmiento Gruber, 2002, Physics Today,
55, 30-36
3
Inter-annual variations in CO2 growth rate
4
Regional greening and browning?
( per year)
Nemani et al., Science 2003
Secular increase in primary productivity from
satellite NDVI over the past 25 years
5
Global Trends in LAI, FAPAR, Burned Area and Fire
using Earth Observation

• Scientific Context
• Leaf Area Index
• FAPAR
• Burned Area
• Active Fire
• Conclusions

6
Leaf Area Index GLOBCARBON
Feng Deng, Jing Chen
Deng et al. in press, TGARS
7
GLOBCARBON vs. MODIS - 1
Degraded to 10km, Barton Bendish (UK) 2001
8
GLOBCARBON vs. MODIS - 2
Degraded to 10km, Mongu (Zambia) 2001
9
GLOBCARBON vs. MODIS - 3
Degraded to 10km, Oregon (USA) 2001
10
Product inter-comparison - 1
Scatterplot over FLUXNET sites
GLOBCARBON closer to CYCLOPES than to
MODIS CYCLOPES saturates early GLOBCARBON low
values in winter CYCLOPES closer to MODIS
02-2002 05-2002
08-2002 11-2002
Roselyne Lacaze
11
Product inter-comparison - 2
Seasonal profiles over some instrumented sites
Ever. Broad. forest, Brazil
Roselyne Lacaze
Crops, France
12
Global Trends in LAI, FAPAR, Burned Area and Fire
using Earth Observation

• Scientific Context
• Leaf Area Index
• FAPAR
• Burned Area
• Active Fire
• Conclusions

13
The 2003 European Heat wave
Philippe Ciais
14
Changes in FPAR between 2003 and former years
Philippe Ciais
15
Vegetation Trend 19992003
decrease
increase
Decrease
Increase
Nadine Gobron
Knorr et. al. (2005) Global-Scale Drought
Caused Atmospheric CO2 increase, EOS,
Transactions 86(18)178 181, 2005.
16
Niño3-SST Anomalies1
99 significance
NINO3 measures the strength of an ENSO event as
the SSTA averaged over 5S,5N and 150W,90W.
Image from http//ingrid.ldeo.columbia.edu/descr
iptions/.nino3.html
Chen et al./NCEP-ClimatePrediction Centre
17
Niño3-SST Anomalies2
3 month lag
99 significance
Precipitation
Chen et al./NCEP-ClimatePrediction Centre
18
Trends at global scales
Nadine Gobron
19
Global Trends in LAI, FAPAR, Burned Area and Fire
using Earth Observation

• Scientific Context
• Leaf Area Index
• FAPAR
• Burned Area
• Active Fire
• Conclusions

20
Burned Area No product No trend

• Year 2000 two independent demonstrators of
  global burned area GLOBSCAR and GBA-2000

GBA-2000
Globscar

• GLOBCARBON uses the experience of these and some
  of the algorithms to produce a single burned area
  product multi-annually.

21
Results 1 km (Angola)
July 1998
Algorithm Detection (GLOBSCAR, GBA, Both)
GLOBSCAR only
GBA only
Both algorithms
22
MODIS Comparison 1 km
MODIS
GLOBCARBON
July 2000
June 29
23
10km Mongolia 2000
2000
May 2000
Percentage of burned pixels in a 1010km box
24
NW Australia 10km May 99
May overlain with May Vectors
May overlain with May Vectors
25
Global Trends in LAI, FAPAR, Burned Area and Fire
using Earth Observation

• Scientific Context
• Leaf Area Index
• FAPAR
• Burned Area
• Active Fire
• Conclusions

26
World Fire Atlas

• 308 or 312K 3.7µm channel
• Global
• 1995-present
• ATSR-2 AATSR
• 1km1km
• 3-day repeat
• monthly files in ascii format (Date, Lat and
  Long)
• http//dup.esrin.esa.int/ionia/wfa/
• Underestimation, industrial sites not masked,
  night-time

27
Satellites do not see everythingTRMM v ATSR-2
(Jan 98)
28
Diurnal Schematic
1.2
ERS-2
MODIS
MODIS
NOAA-14
1
ENVISAT
NOAA-12
0.8
NOAA-12
0.6
Normalised solar position
Fire 1
Fire 1 (next day)
0.4
ENVISAT
MODIS
NOAA-14
Fire 3
Fire 2
MODIS
0.2
ERS-2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Time (hrs)
29
Seasonal Inter-annual
30
Continental
31
Conclusion - 1

• While the products are becoming available, LAI
  currently does not have sufficient consistency
  for long times series analysis globally. Attempts
  to do so are possible but they should be viewed
  with caution.
• FAPAR derived from space has been shown to
  reliably exhibit strong signatures of climate and
  other stress impacts on vegetation.

32
Conclusions - 2

• Burned area represents a similar story to LAI but
  with fewer products and there are problems at
  regional scales. GLOBCARBON will provide 10 years
  once reprocessed and MODIS is coming
• Active fire data exist as long time series
  (ATSR-2 WFA, TRMM, MODIS) but they represent
  snapshots (no one product is better than
  another). They provide a means to examine climate
  trends and regional variation but ultimately it
  requires high resolution geostationary for
  continuous diurnal monitoring
• Continuity of biophysical products over long time
  series are needed with various instruments
• Same type of high level products complete with
  quality values
• Validation and comparison exercises for quality
  assessments.
• Consistency over time and between products

33
Canberra Fires Jan 15 2003
Hot spots detected by ATSR-2 (left) and AATSR
(right) with ½ hour spacing. Below actual scene
zoom with saturated pixels in blue (gt312deg)