Exploring the Roles of Climate and Land Surface Changes

1

• Exploring the Roles of Climate and Land Surface
  Changes
• on the Variability of Pan-Arctic River Discharge
• Jennifer C. Adam1, Fengge Su1, Laura C. Bowling2,
  and Dennis P. Lettenmaier1
• Department of Civil and Environmental
  Engineering, Box 352700, University of
  Washington, Seattle, WA 98195
• 2. Department of Agronomy, Purdue University,
  West Lafayette, IN 47907

Photo http//gallery.maiman.net/terragen/arctic
3
Precipitation and Temperature Effects on Runoff
and Land Surface Albedo Variability
2
Improvement of Precipitation and Temperature
Inputs
ABSTRACT The export of freshwater to the Arctic
Ocean plays a key role in both regional and
global climates (e.g. via effects on the strength
of the North Atlantic Deep Water (NADW) formation
that drives the thermohaline circulation).
Observed changes in streamflow may be linked both
to direct and indirect effects of climate change.
For example, a general warming leads to earlier
spring runoff, however changes in high arctic
vegetation such as increased incidence of brush
can lead to increased snow accumulation, and
hence sustained runoff later in the summer. Also,
vegetation changes can substantially change
evapotranspiration. Changes in snow cover extent
and the distribution of vegetation and wetlands
over the pan-arctic domain affect land-atmosphere
energy exchanges, and the seasonality of river
flow. Furthermore, recent research has suggested
that changes in permafrost extent and the active
layer depth may also be affecting river flow. We
report a 60-year (1930-89) run of the Variable
Infiltration Capacity (VIC) macroscale hydrology
model over the pan-arctic land domain, designed
to offer insights into the nature and causes of
observed long term trends in river discharge. VIC
is a semi-distributed grid-based model that
parameterizes the processes occurring at the
land-atmosphere interface. The most recent
version of the model includes several recent
improvements specific to cold-land regions. We
summarize a set of model runs from which we have
estimated the inflow to the Arctic Ocean from all
pan-arctic land areas (including the Canadian
Archipelago) and an assessment of the capability
of the land surface model to simulate the
observed changes in gauged streamflow. These
results utilize precipitation and temperature
fields that incorporate a method of adjustment to
reflect the best current understanding of
long-term precipitation and temperature trends
over the pan-Arctic domain. Finally, we describe
an exploratory analysis in which we use the model
to evaluate the effects of changes in climate
(precipitation and temperature) and active layer
depth on streamflow variability and trends over
the last half century.
Precipitation and temperature inputs to
the model were adjusted for spurious trends
according to the method of Hamlet et al. (2004)
using temporally homogenous index stations from
the GHCN (http//cdiac.esd.ornl.gov/ghcn).
Designed after Hamlet et al. (2004), the
following four VIC model runs were made
Mackenzie 37 mm
Mackenzie 43 mm
Lena 3 mm
Lena 31 mm
Ob 27 mm
Ob 12 mm
Description of Figures

• Things to explore using these figures
• For each season and basin, are streamflow and
  albedo trends more sensitive to trends in
  precipitation or temperature?
• Do runoff trends match precipitation trends and
  why or why not?
• Do simulated runoff trends match observed runoff
  trends for each season?
• Note Albedo is basin average (uses aging
  algorithm)
• Things learned and new questions
• Precipitation Albedo , Temperature
  Albedo
• Precipitation Runoff ? , Temperature
  Runoff ?
• Ob Runoff sensitive to precip, mostly
  insensitive to temp. Spring albedo sensitive to
  temp.
• Lena Runoff sensitive to both precip and temp,
  e.g. winter temp controls, spring precip
  controls.
• Summer and Fall albedo sensitive to temp.
• Ob Simulated runoff similar to observed runoff.
• Lena Simulated runoff different from observed
  runoff. What is missing from the model? or
  Inaccurate forcings?

Temperature 1930-1989
Before
After
Mackenzie 0.25 C
Mackenzie -0.39 C
Lena -0.33 C
Lena -0.08 C
Example Ob Spring Exp. 2 Precip
Runoff ?Storage why?... future work
Precipitation
Evaporation
Runoff
Storage Release
Ob 0.59 C
Ob 0.83 C
1
Modeling Framework

• Features Specific to Cold-Land Processes
• Two-layer energy balance snow model (Storck et
  al. 1999)
• Frozen soil/permafrost algorithm (Cherkauer et
  al. 1999, 2003)
• Lakes and wetlands model (Bowling et al. 2004)
• Blowing snow algorithm (Bowling et al. 2004)
• Calibration (Su et al. 2005)
• Eleven Regions were calibrated separately (not
  including Greenland
• Calibration was focused on matching the shape of
  the monthly hydrograph.
• Parameter transfer to un-gauged basins was based
  on the hydro-climatology of the region.

0.29 mm year-1
Depth, mm
-0.01 mm year-1
0.16 mm year-1
-0.14 mm year-1
4
Effects of Permafrost Extent Changes on Runoff
Variability
Run 5 Tdamp from 1931-1939 period
Run 6 Tdamp from 1980-1989 period
Adjusting the spurious trends in precipitation
and temperature caused the simulated streamflow
long-term trend to match observed for the Ob, but
did not improve the trend over the Lena.
Decrease in Permafrost Extent
Ob
Observed
Before Trend Adjustment
After Trend Adjustment
Ob
Over the south-eastern part of Ob where the
greatest changes in ground temperatures occurred,
Run 6 spring runoff is greater and summer runoff
is less than Run 5, indicating an earlier melt.
Because air temperatures are the same for both
runs, warmer ground conditions are responsible.

• Validation (Su et al. 2005)
• Snow Cover Extent via comparison to NOAA-NESDIS
  weekly snow charts
• Permafrost active layer depth via comparison to
  CALM network observations
• Lake algorithm validation via comparison of lake
  freeze and thaw dates to observed
• Domain
• Pan-Arctic Domain per ArcticRIMS
• 100 km by 100km EASE
• Period 1930-1989, 1 year spin-up

Lena
Observed
Before Trend Adjustment
After Trend Adjustment

• CONCLUDING REMARKS
• Adjustment of spurious trends in precipitation
  and temperature (minimum and maximum daily)
  improves streamflow trends and variability in
  some basins, but not all. Is there something
  missing from the model, or is there still
  something wrong with the inputs? .. Future
  work.
• The model shows that runoff trends are not
  always consistent with precipitation trends our
  example shows increasing spring storage releases
  over the Ob. What is causing this increase? Is
  this realistic?
• Warming ground temperatures are shown to have a
  direct effect on streamflow response this will
  also be explored further.
• The effects of changing land cover
  classification on streamflow and albedo
  variability will be explored, e.g. northward
  moving tree-line.
• Anthropogenic effects on streamflow and albedo
  variability will also be explored, e.g.
  agriculture and reservoir storage.

Change in streamflow seasonality peak flows
during the first three decades are improved for
the Lena there is no apparent improvement for
the Ob.
2810 cells routed to 643 outlets Contributing
Area 25 million km2