5 June 2006Canberra, Australia Global Trends in LAI, F APAR, 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)
5 June 2006Canberra, Australia Scientific Context Source: Sarmiento & Gruber, 2002, Physics Today, 55, Accumulation rate in atmosphere Accumulation rate in ocean and on land 1 Pg C/yr 3.5 Pg C/yr 2.5 Pg C/yr 6.5 Pg C/yr Net Uptake Ocean & Land Large interannual variation in the annual atmospheric CO 2 growth rate Fossil fuel emissions Growth rate of carbon reservoirs
5 June 2006Canberra, Australia Inter-annual variations in CO 2 growth rate
5 June 2006Canberra, Australia Secular increase in primary productivity from satellite NDVI over the past 25 years Nemani et al., Science 2003 (% per year) Regional greening and browning?
5 June 2006Canberra, Australia Global Trends in LAI, F APAR, Burned Area and Fire using Earth Observation Scientific Context Leaf Area Index F APAR Burned Area Active Fire Conclusions
5 June 2006Canberra, Australia Leaf Area Index GLOBCARBON Deng et al. in press, TGARS Feng Deng, Jing Chen
5 June 2006Canberra, Australia GLOBCARBON vs. MODIS - 1 Degraded to 10km, Barton Bendish (UK) 2001
5 June 2006Canberra, Australia GLOBCARBON vs. MODIS - 2 Degraded to 10km, Mongu (Zambia) 2001
5 June 2006Canberra, Australia GLOBCARBON vs. MODIS - 3 Degraded to 10km, Oregon (USA) 2001
5 June 2006Canberra, Australia Scatterplot over FLUXNET sites Roselyne Lacaze GLOBCARBON closer to CYCLOPES than to MODIS CYCLOPES saturates early GLOBCARBON low values in winter CYCLOPES closer to MODIS GLOBCARBON CYCLOPES GLOBCARBON MODIS CYCLOPES MODIS Product inter-comparison - 1
5 June 2006Canberra, Australia Seasonal profiles over some instrumented sites Ground 2002 Ground not 2002 Crops, France Ever. Needle. Forest, Estonia 5 0 J Dec Ever. Needle. Forest, Canada J Dec Ever. Broad. forest, Brazil 3 0 J Dec Grass, France Dec. Broad. forest, USA 7 J 0 Dec 0 6 J Crops, Argentina Dec J 3 0 J Crops, Romania Dec J 0 2 Roselyne Lacaze Product inter-comparison - 2
5 June 2006Canberra, Australia Global Trends in LAI, F APAR, Burned Area and Fire using Earth Observation Scientific Context Leaf Area Index F APAR Burned Area Active Fire Conclusions
5 June 2006Canberra, Australia The 2003 European Heat wave Philippe Ciais
5 June 2006Canberra, Australia Changes in FPAR between 2003 and former years Philippe Ciais
5 June 2006Canberra, Australia Vegetation ‘Trend’ 1999–2003 decreaseincrease DecreaseIncrease Biosphere Energy-Transfer HYdrology model Knorr et. al. (2005) ‘ Global-Scale Drought Caused Atmospheric CO 2 increase’, EOS, Transactions 86(18):178 & 181, Nadine Gobron
5 June 2006Canberra, Australia NINO3 measures the strength of an ENSO event as the SSTA averaged over [5S,5N] and [150W,90W]. Image from: 99% significance Niño3-SST Anomalies…1 Precipitation Chen et al./NCEP-Climate Prediction Centre
5 June 2006Canberra, Australia 99% significance Precipitation 3 month lag Chen et al./NCEP-Climate Prediction Centre Niño3-SST Anomalies…2
5 June 2006Canberra, Australia ‘Trends’ at global scales Nadine Gobron
5 June 2006Canberra, Australia Global Trends in LAI, F APAR, Burned Area and Fire using Earth Observation Scientific Context Leaf Area Index F APAR Burned Area Active Fire Conclusions
5 June 2006Canberra, Australia GLOBCARBON uses the experience of these and some of the algorithms to produce a single burned area product – multi-annually. Burned Area – No product = No trend GBA-2000Globscar Year 2000 – two independent demonstrators of global burned area: GLOBSCAR and GBA-2000
5 June 2006Canberra, Australia Results – 1 km (Angola) July 1998 GLOBSCAR onlyGBA only Both algorithms Algorithm Detection (GLOBSCAR, GBA, Both)
5 June 2006Canberra, Australia MODIS Comparison – 1 km July 2000 MODIS GLOBCARBON June 29
5 June 2006Canberra, Australia Percentage of burned pixels in a 10*10km box 2000 May km – Mongolia 2000
5 June 2006Canberra, Australia NW Australia – 10km May 99 May overlain with May Vectors
5 June 2006Canberra, Australia Global Trends in LAI, F APAR, Burned Area and Fire using Earth Observation Scientific Context Leaf Area Index F APAR Burned Area Active Fire Conclusions
5 June 2006Canberra, Australia World Fire Atlas 4308 or 312K – 3.7μm channel 4Global present 4ATSR-2 + AATSR 41km*1km 43-day repeat 4monthly files in ascii format (Date, Lat and Long) 4Underestimation, industrial sites not masked, night-time
5 June 2006Canberra, Australia Satellites do not see everything TRMM v ATSR-2 (Jan 98)
5 June 2006Canberra, Australia Diurnal Schematic Time (hrs) Normalised solar position NOAA-12 NOAA-14 MODIS ENVISAT ERS-2 NOAA-14 ERS-2 NOAA-12 MODIS ENVISAT Fire 1 Fire 3 Fire 1 (next day) MODIS Fire 2
5 June 2006Canberra, Australia Seasonal – Inter-annual
5 June 2006Canberra, Australia Continental
5 June 2006Canberra, Australia 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. F APAR derived from space has been shown to reliably exhibit strong signatures of climate and other stress impacts on vegetation.
5 June 2006Canberra, Australia 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
5 June 2006Canberra, Australia Canberra Fires Jan Hot spots detected by ATSR-2 (left) and AATSR (right) with ½ hour spacing. Below actual scene zoom with saturated pixels in blue (>312deg)