Progress in hydrological modeling over high latitudes

--under Arctic Climate System Study (ACSYS)

Dennis P. Lettenmaier and Fengge Su

Arctic Climate System Study (ACSYS) Mission

• What are the global consequences of

natural or human-induced change in the

Arctic climate system?

• Is the Arctic climate system as sensitive to

increased greenhouse gas concentrations

as climate models suggest?

ACSYS is a core project of the World Climate Research Programme (WCRP). It started from January 1994, and ended in December 2003.

Components of ACSYS

1. Arctic ocean circulation programme

2. Arctic sea-ice programme

3. Arctic atmosphere programme

4. Hydrological cycle in the Arctic region

5. ACSYS modelling programme

6. Data Management and Information

Objectives of ACSYS hydrological programme

• Developing mathematical models of the hydrological cycle under

specific Arctic climate conditions suitable for inclusion in

coupled climate models;

• Determining the elements of the fresh water cycle in the Arctic

region and their time and space variability;

• Quantifying the role of atmospheric, hydrological and land surface processes in the exchanges between different elements of the hydrological cycle;

• Providing an observational basin for the assessment of possible long-term trends of the components of the fresh water balance in the Arctic region under changing climate.

Major components of the ACSYS hydrological programme

• Development regional data bases for the

main components of the fresh water

balance of the Arctic region;

• Development of hydrological models of

selected Arctic river basins and their

validation using appropriate observational

data sets.

Projects contributing to ACSYS

1. The project for the Intercomparison of Land-Surface Parameterization Schemes Phase 2(e) – PILPS 2(e)

2. GEWEX Continental Scale Experiments (CSEs)

. MAGS . BALTEX . GAME-Siberia . GCIP/GAPP

Hydrological modelling activities in ACSYS were performed in close collaboration with WCRP Global Water and Energy Experiment (GWEX) project.

The Project for the Intercomparison of Land-Surface Parameterization Schemes (PILPS) 2(e)

A small working group was established

by ACSYS in August 1998 for the

purposes of planning an Arctic hydrology

model intercomparison project. The

resulting project, PILPS phase 2(e), was a

joint experiment of ACSYS and GEWEX,

and was designed to evaluate the

performance of land surface models in

high latitudes.

Special Issue on PILPS 2(e): Global and Planetary, 38(1-2), 2003

List of participating models of PILPS Phase 2(e)

Torne– Kalix River basin and the 218 computational ¼ ° latitude/longitude grid cells.

Bowling et al., 2003

Bal

tic S

ea

Mean monthly observed (dots) and simulated (lines) discharge for the Kalix and Torne river basins.

Nijssen et al., 2003

Observed (dots) and simulated (lines) snow water equivalent for five locations

Number of days with snow cover for models participating in PILPS 2(e). Lowest right panel shows a satellite derived estimate

Mean annual latent heat flux for the model reruns. The last panel shows annual latent heat flux estimated from basin water balance.

after Nijssen et al, 2003Bowling et al., 2003

MAGS

BALTEX

GAME-Siberia

Mackenzie GEWEX Study (MAGS)

Models• CLASS• WATFLOOD• WATCLASS• CRCM/CLASS

WATFLOOD representation of Mackenzie River drainage basin. Each linear segment represents a 50 kilometre river reach.

Snelgrove et al, 2005

Canadian Land Surface Scheme (CLASS)

Soulis et al., 2005

Framework for hydrological modelling

in MAGS

Snelgrove et al, 2005

WATFLOOD (Level 0) WATCLASS (Level 2)

Snelgrove et al, 2005

Baltic Sea Experiment (BALTEX)

Models:

HBV- conceptual hydrological model

SEWAB - land-surface scheme

49° N - 69 ° N

Baltic Sea Drainage Basin

Monthly averages of freshwater flow into the major subbasins of the Baltic Sea, calculated with the HBV mode

Raschke et al., 2001

Baltic Sea drainage area: 1.6 ×106 km2

Modeled seasonal river discharge to the Baltic Sea from HBV-Baltic for present-day conditions (shaded) and four climate

change scenarios.

Graham, 2004

Bal

tic

Sea

Hydrological components of the land surface scheme

SEWAB (Surface Energy and Water Balance)

• VIC approach (Warrach et al.,

1999)

• TOPMODEL approach

( Stieglitz et al., 1997)

• The concept of ponding at the

surface (Mengelkamp et al.,

2001)

• The processes of soil freezing

and thawing and the seasonal

snow cover (Warrach et al.,

2001)

Various optional versions of SEWAB

49° N

54° N

The Odra drainage basin (119,000 km2 ) with 18 km mesh size

Daily streamflow simulated with SEWAB and Lohmann et al. (1996) routing scheme

Mengelkamp et al., 2001

GEWEX Asian Monsoon Experiment GAME-Siberia

GAME-Siberia project

concentrates on observation

and modeling of land

surface processes, and

regional analysis of energy

and water cycle in

permafrost region of eastern

Siberia.

Lena River Basin

Ma et al., 2000

Six hydrological stations within the Lena River Basin.

Simulated runoff at the six hydrological stations from October 1986to September 1987, using a combined model which is composed of a SVAT model, runoff model, and river routing model .

Annual effective rainfall (P-E), Oct 1986- Sep 1987

Simulated evapotranspiration (E), Oct 1986-Sep 1987

Precipitation, Oct 1986- Sep 1987

Ma et al., 2000

A number of changes have

been made to improve the VIC

model's representation of cold

season processes, in

conjunction with the GEWEX

Continental-scale International

Project (GCIP) activities in the

upper Mississippi River basin,

and the PILPS Phase 2(e)

experiment in the Torne-Kalix

River basin.

The VIC model

Cold

land

processes in

VIC

A two-layer energy balance snow model (Storck et al, 1999; Cherkauer et al, 2003, JGR)

A frozen soil/permafrost algorithm (Cherkauer et al, 1999; 2003, JGR)

A lake and wetlands model (Bowling et al, 2003,WRR)

A blowing snow model (Bowling et al, 2004, J. Hydromet)

Cold land processesCold land processes

The VIC model was applied to

the Mackenzie and Ob River

basins at 2° spatial resolution

and daily temporal resolution to

examine the space-time

structure of the predicted

hydrologic variables (i.e., runoff,

evaporation, soil moisture, and

snow water equivalent)

Bowling et al., 2000Routing networks for the Mackenzie (A) and the Ob (B)

A

B

Ob Mackenzie

Bowling et al., 2000

The goal of ACSYS hydrological

programme is to determine the space-

time variability of the Arctic hydrological

cycle and the fluxes of freshwater to the

Arctic Ocean.

Arctic river network over 100km grid system

Estimates of Annual Continental Freshwater into the Arctic Ocean

Basin Definition

Contribution Area (×1000 km2)

Volume (km3/yr)

Periods

“Arctic Ocean River Basin” in Prowse et al.[2000]a 11045/15504 2338/3299 1975-1984

“All Arctic Regions” in Shiklomanov et al.[2000] 18875 4300 1921-1996

“Arctic Ocean Basin” in Shiklomanov er al.[2000] 23732 5250 1921-1996

“Arctic Climate System” in Grabs et al.[2000]a 12868/18147 2603/3671

AORB - Northern Greenland + Arctic Archipelago in [Lammers et al.2001]

16192 3302 1960-1989

The largest Arctic Rivers in Dai et al.[2002] 16850 3658

AORB - Northern Greenland, VIC1 15017 3354 1979-1999

AORB - Northern Greenland + Arctic Archipelago, VIC2

16397 3596 1979-1999

Basin area-annual flow volume relationship for different estimations

y = 217.81x

R2 = 0.9756

2000

3000

4000

5000

6000

10 12 14 16 18 20 22 24 26

Contribution Area (106km3)

Flo

w V

olu

m (

km3 /y

r)

VIC1VIC2VIC1 VIC2

Simulated mean monthly streamflow discharge into the Arctic Ocean (1979-1999)

Monthly mean discharge to the Arctic Ocean(1979-1999)

0

200

400

600

800

1000

1200

1 2 3 4 5 6 7 8 9 10 11 12

Month

Dis

cha

rge

(km

3 /yr)

Total = 3354km3/yr

Area = 1.5×106km2Total =3354km3/yrArea = 1.5×106km2

What did the ACSYS achieve? 1. Several land surface models had been improved in representing snow accumulation

and ablation, soil freeze/thaw and permafrost, and runoff generation motivated by the

PILPS 2(e) experiment in Torne-Kalix River basin and other projects related to ACSYS.

2. Intensive field measurement under MAGS and GAME-Siberia promoted the

development of process algorithms of snow accumulation, redistribution, and ablation,

and water infiltration into frozen soil, and the development of one-dimensional land

surface models for cold region.

3. The VIC model and the macroscale hydrological models developed under the MAGS,

BELTEX, and GAME-Siberia have been used to simulate the surface water and energy

balance of high-latitude river basins.

4. The Arctic river runoff in both gauged and ungauged basins and the freshwater river

inflow to the Arctic Ocean were estimated and analyzed by using a macroscale

hydrological model.

5. The water balance terms of the land surface water in the Arctic region and their time

and space variability were determined and evaluated by using a land surface model and

the ERA-40 reanalysis.

What remains to be done ?

• The roles of frozen soil moisture and blowing snow

parameterizations in the large-scale simulation of runoff,

temperature, and evaporation are not completely clear.

• Existing wetlands and lake models in land surface models need to

be further improved and validated.

• Many results from process investigations of snow and frost-related

hydrological processes remain to be incorporated into large-scale

hydrological models.

• Continued development of hydrological models and linkages

between atmospheric and hydrological models are needed in

scientific studies of the interactions between climate, snow and

frost hydrology.

• Most of the key issues have also been addressed in the Science

Plan of Climate and Cryosphere (CliC), which is the successor of

ACSYS.

Thank You!