PowerPointPrsentation

1
SIBERIA II Use of Remote Sensing to infer
Ecological Processesfor Carbon Fluxes Estimates
5th Framework Program of the European Commission,
Generic Activity 7.2 Development of generic
Earth Observation Technologies (EVG1-CT-2001-00048
)
2

• Thuy Le Toan, Nicolas Debart, Manuela Grippa,
• Sergo Vicente-Serrano and Laurent Kergoat
• CESBIO, Toulouse, France
• Nelly Mognard
• LEGOS, Toulouse, France
• Richard Kidd, Wolfgang Wagner
• IPF, Vienna, Austria
• Ghislain Picard, Shaun Quegan
• CTCD, UK
• Philippe Peylin, Philippe Ciais
• LSCE, France

3
Applicability of simple spring phenology model?
4
Remote Sensing
1987
Model
5
Regions of similar temporal variations
byPrincipal Component Analysis
Trend 82-04 D/decade
-2.20
-4.04
-5.72
-1.12
-4.06
-0.52
-0.03
-1.62
-2.22
-1.04
Strongest advance in Siberia central (components
1 and 4)
6
Remote Sensing Model
7
Trend from remote sensing
42
-42
42
-42
Trend from model
8
Comparison Remote Sensing- Model
Other models needed
9
How do phenological events (thaw start, snow
melt, phenology dates) coincide with start of
carbon efflux and carbon uptake?
.Correspondence with in situ flux measurements
? Correspondence with atmospheric inversion data?
10
Materials and methods

• Development of EO methods to determine
• spring phenology dates (CESBIO)
• snow melt dates and snow water equivalent
  (CESBIO)
• thaw freeze dates (IPF)
• Analysis of correspondence with seasonal
• variations of CO2 measurements (TCOS-Siberia)

Inferring vegetation processes and questions
11
Snow products in SIBERIA

• Snow melt dates using SSM SSM/I (1978-2001)
• Snow depth (snow water equivalent)
• SIBERIA-2 Development of new method using
• SSM/I (1987-2003)

• - M. Grippa, N. M. Mognard, T. Le Toan and E. G.
  Josberger (2004)
• Siberia snow depth climatology derived from
  SSM/I data using a combined dynamic and
• static algorithm.Rem. Sens. Environ. 93 (2004)
  30-41
• 2 - M.Grippa, N.M. Mognard, T. Le Toan (2005)
• Comparison between the interannual variability
  of snow parameters derived from
• SSM/I and the Ob river discharge Rem. Sens.
  Environ.

Data base of Snow products from 1987 to 2003 25
km, Siberia and global
12
SIBERIA end of snow melt dates (1988-2002)
13
(No Transcript)
14
(No Transcript)
15
Freeze thaw dates in SIBERIA
SIBERIA-2 Development of a new method
using Quikscat in Siberia
Data base of Freeze Thaw dates from 2000 to 2003
25 km Siberia and Eurasia
16
Freeze/Thaw Indicator by Quikscat
SIBERIA II Start of Thaw 2000
17
In situ soil CO2 efflux and EO beginning and end
snow melt
Soil respiration starts at end of snow melt
18
NEE of bog, in 1998, 1999, 2000, at at 6045 N,
8923 E (Schulze et al., 2002)
19
1999
1999
Thaw start Snowmelt Budburst
day 128 day 150
6045 N, 8923 E (Schulze et al., 2002)
20
NEE at 6045 N, 8923 E of bog (Schulze et
al., 2002)
2000
2000
Thaw start Snowmelt Budburst day 100
day 133 day 146
(Schulze et al., 2002)
21
1998
1998
Thaw start Snowmelt Budburst
day 148 day 153

• Bog
• CO2 efflux starts at thaw dates,
• turns from source to sink at greening date

22
NEE at forest stands of Betula, Abies and mixed
stands at 6101 N, 8934 E (Roser et al.,
2002)
EO data Thaw start day 100 (10April)
Snowmelt 143 (17 May) Budburst 146 (20 May)
Forests beginning of carbon release at thaw
start Deciduous forestcarbon source to sink
after greening Evergreen already carbon sink at
greening
23
Pinus Sylvestris
Almut Arneth
Evergreen forest Carbon sink at EO greening
24
Phase in CO2 measurements and EO dates
1. There is an indication that thaw start -
beginning of snow melt - coincides with onset of
respiration for all cover types 2. For bogs and
deciduous forests greening up occurs in
transition from C02 efflux to CO2 take up 3.
Evergreen forest is already a sink at EO
greening up dates
Determination of growing season according to
temperature? according to EO vegetation indices?
25
Preliminary comparison with atmospheric
inversion data
26
Atmospheric Inversion data (from P.Peylin, LSCE)
Snow melt dates
Occur at the first decrease
27
Atmospheric Inversion data (from P.Peylin, LSCE)
Greening up dates
Occur at 0 crossing, but data are monthly!
28
Correspondence snow melt, greening up dates
29
Interannual variations of the negative (sinks)
and positive (sources) of CO2
Data from P. Peylin, LSCE
Interannual variations dominated by variations of
the positive part in spring, autumn and
winter Interannual variability caused by
respiration?
30
Impacts of change in snow melt dates and snow
depth on ecosystem functioning in Siberia?
Insitu studies deeper snow pack increases -
summer vegetation productivity - summer leaf
nitrogen content - soil nitrogen mineralization
(Schimel et al., 2004, Dorrepal et al., 2004,
Grogan and Jonasson, 2003) -carbon uptake by
vegetation 6 months later (Welker et al., 2000)
Large scale observations?

• M. Grippa, L. Kergoat, T. Le Toan, N. Delbart,
  J. LHermitte and S. Vicente-Serrano (2005)
• On the relationship between vegetation and snow
  indicators derived from remote sensing
• over central siberia Geophys. Res. Lett.

31
Valuesgt0.4 are significant above 90

• Negative correlation
• Late snowmelt Less cumulative NDVI
  
• lenght of the growing season
• correspondence between snowmelt and positive
  temperatures
• effects of the snowmelt on the reflectance used
  to calculate the NDVI

32

• Positive correlation
• Late snowmelt more NDVI in summer
• water availability after snowmelt (steppe)
• late snowmelt proctection from the snow layer
  against freeze events (fine root mortality)

33

• Positive correlation
• more snow depth in winter more NDVI
• more water available after snowmelt ?
• soil insulation provided by a thicker winter
  snowpack (enhancement of microbial activity
  during winter-gtenhance leaf nitrogen content,
  soil nitrogen mineralization?

34
Highlights from EO
Siberia advance in spring phenology and snow
melt dates in the last 23 years which may result
in opposite effects on NPP