Transcript Glacier, LC

Assessment of CCI Glacier and CCI Land
cover data for hydrological modeling of the
Arctic ocean drainage basin
David Gustafsson, Kristina Isberg, Jörgen
Rosberg, Berit Arheimer
SMHI
CCI-CMUG Integration meeting 5, 26-28/5, 2015
WHY modelling the Arctic basin?
~ 11% of world’s river-runoff flows to the Arctic ocean
(Gleick, 2000)
> 50% of freshwater to the Arctic Ocean is river runoff
(Barry and Serreze, 2000)
30% is Ungauged
Modified from:
Shiklomanov and
Lammers (2009),
Environ. Res. Lett.
WP3.9 Assessment of glacier, land cover and sea
level data for hydrological modeling of the Arctic
ocean drainage basin
Objectives
Assess the use of ESA CCI Glacier, Land cover (and Sea level)
in hydrological modelling for the Arctic:
• Compare CCI data with existing datasets: long term trends and
seasonal variations?
• Can use of CCI Glacier and Land cover improve simulated river
runoff to the Arctic Ocean?
• Is there a relation between changes in observed land cover and
the simulated river runoff?
WP3.9 Assessment of glacier, land cover and sea
level data for hydrological modeling of the Arctic
ocean drainage basin
Use of CCI data
Hydrological models depend on observations of e.g. glacier
area and land cover for setup, calibration and evaluation:
• Comparison with datasets used in the pan-arctic hydrological
model Arctic-HYPE
• Models using CCI data (Glacier, LC) will be compared to
previous model version
• CCI Glacier data will be used to improve the parameteri-zation
of the sub-basin based glacier mass balance model.
Large scale modelling using a
catchment approach
Model setup (short version):
1. Sub-basin delineation based on DEM (Hydro 1K),
Lakes (GLWD), and River discharge stations (GRDC
The Arctic-HYPE model:
etc) for the Environment (HYPE)
HYdrological Predictions
(Lindström et al. (2010), Hydrology Research)
23 million km2
2. Each sub-basin is divided in sub-units representing
599 sub-basins
combinations of soil type (HWSD) and land 32
cover
HYPE source code, WIKI, manuals, etc:
average spatial resolution
(GlobCover)
http://hype.sourceforge.net/
=715 km2
3. Calibration with various data (snow, evaporation,
galcier mass balance, discharge)
Available open global databases used
Trends in water balance and storages
Flow to Ocean - trends in annual mean
Pan-Arctic
Ob, Yenisei, Lena, Sernaja Dvina, Pechora, Kolyma
Yukon, Mackenzie, Peel, Back
Overall, similar trends to Shiklomaniv & Lammers (2009)
– but overestimated flow in North America (and no trend / declining)
Trends in water balance and storages
Modelled water balance for the Arctic basin
Trends in evaporation and precipitation < 1mm/year
Increase in Prec and Runoff in last decade
Storage change – 6 mm/year corresponding to glacier mass balance
Trends in water balance and storages
Modelled water storages in the Arctic basin
 Negative trend in Glacier storage = - 6 mm/year for the entire basin:
 Corresponds to -164 km3/yr in glacier volume
 Probably too high – Glacier model need more attention!
 Small changes in other storages
Annual change in water storage:
Assessment of CCI glacier, Land cover
Progress (preliminary results):
1) CCI Glacier (RGI v4.0) and CCI Land Cover (v1.3-4) used to evaluate Arctic-HYPE (v3)
glacier area and volume:
•
Glacier area:
•
Arctic-HYPE (v3) based on GlobCover 2004-2006
(v2.2) (“permanent snow and ice”)
•
CCI Land Cover (v1.3-4) gives more realistic snow
and ice distribution in the Arctic:
•
•
>100% difference
GlobCover
between
CCI
LC
and
However, CCI Land Cover (v1.3-4) gives ~30% larger
area than CCI Glacier (RGI v4.0).
Assessment of CCI glacier, Land cover
Evaluation Glacier Area [km2] per RGI regions:
• 95% of Region 1-10 (excl Greenland) is covered by
Arctic HYPE
• CCI clearly better than GlobCover (still +30%)
• Permanent snow and ice largely
overestimate glacier area in Arctic North
America and Arctic Russia/North Asia
RGI area RGI area
RGI area
Based on CCI LC
in model (fraction of total
(fraction of
(fraction of RGI Based on GlobCover LC
RGI region
(km2)
area in RGI region) model area)
area)
(fraction of RGI area)
01 Alaska
1232
0.12
0.007
1.359
1.569
02 Western Canada/US
315
0.15
0.001
3.152
17.114
03 Arctic Canada North
104074
1.00
0.291
1.272
2.4
04 Arctic Canada South
41303
1.01
0.036
1.607
2.647
06 Iceland
11060
1.00
0.109
1.009
2.079
07 Svalbard
32342
0.98
0.542
1.185
1.65
08 Scandinavia
1805
0.80
0.006
1.56
1.311
09 Russian Arctic
50756
0.99
0.346
1.198
2.286
10 North Asia
759
0.47
0
2.039
8.599
Total
243646
0.95
0.014
1.315
2.424
WP3.9 Assessment of glacier, land cover and sea
level data for hydrological modeling of the Arctic
ocean drainage basin
Volume from RGI areas[km3]
Progress (preliminary results):
1) CCI Glacier (RGI v4.0) and CCI Land Cover (v1.3-4) used to evaluate Arctic-HYPE (v3)
glacier area and volume:
Glacier volume based on non-linear area-volume
relationship (Bahr et al, 1997, Grindsted, 2003):
Volume from subbasin area [km3]
•
CCI Glacier outlines (RGIv4) was used to evaluate
error when aggregating (or splitting) area from
multiple glaciers in/between model sub-basins:
•
Volume overestimated by 28%
•
In combination with the area overestimation,
current Arctic HYPE model overestimate Glacier
volume with more than 200% compared to data
derived from CCI Glacier!!
Assessment of CCI glacier, Land cover
Evaluation Glacier Volume estimates per RGI regions:
•
•
•
Eyafjallajökul, Iceland
RGI region
01 Alaska
02 Western Canada/US
03 Arctic Canada North
04 Arctic Canada South
06 Iceland
07 Svalbard
08 Scandinavia
09 Russian Arctic
10 North Asia
Total
Volume estimated from total area overestimate volume
from individual area by 28%
Variation depending on:
• agreement glacier outlines and subbasins
• number of glaciers / subbasin
Combination of area and volume estimation errors add
up largely in current model!
Volume from individual RGI Volume from
Vol [total] / Vol [ind]
Vol [GlobCover]/Vol
areas (km3)
total area [km3]
Vol [CCI area] / Vol [RGI] [RGI]
1216
1654
1.361
1.901
2.077
113
224
1.987
7.168
59.204
15778
18488
1.172
1.518
3.135
4701
5868
1.248
2.067
3.541
1750
1811
1.035
1.041
2.164
4269
6244
1.463
1.768
2.543
134
217
1.614
2.627
2.313
7071
10225
1.446
1.78
3.703
78
152
1.952
4.231
26.872
35108
44882
1.278
1.693
3.377
WP3.9 Assessment of glacier, land cover and sea
level data for hydrological modeling of the Arctic
ocean drainage basin
Progress (preliminary results):
2) CCI Land cover used to evaluate and improve Arctic-HYPE surface water area:
Example from the area around the lower Ob:
• Globcover shows sparse vegetation
• CCI 2005 shows water bodies; Shrub or
herbaceous cover; flooded; fresh/saline/brakish
water.
On average 6-10% increase, depending on method
for combining information from CCI and GLWD.
3) CCI land cover time series 2000, 2005, 2010 show negative trend in Siberian deciduous needleleaf vegetation:
•
Due to increase in soil moisture as response to increased precipitation? To be evaluated
4) CCI Sea level - No results yet
Conclusions
Preliminary conclusions
•
•
•
•
CCI Glacier very useful for hydrological modelling:
• setup and evaluation of glacier area and volume
• Very high resolution and level of detail
• Global coverage
Current limitations:
• Time series missing in RGI dataset
• needed for initialization, calibration, evaluation of long term modelling
• Users may compile information from other data sets:
• GLIMS/WGMS/literture/etc.
CCI Land cover:
• Improved snow and ice area
• Increased water surface area
• Interesting trends in Arctic forest vegetation
Current limitations:
• Static Surface water and Glacier information
Thank you
Q&A