1. 1 22 June 2009 Arctic System Reanalysis Land Component Update Michael Barlage and Fei Chen Research Applications Laboratory (RAL) National Center for Atmospheric Research

2. 2 Alaska Measuring Stations 1km 2 measurement grid with 121 points 100m apart

3. 3 HRLDAS Simulation Specifics 27-year (1980-2006) point simulations over CALM measurement sites Forcing data: ERA-40 (1980-1999); JRA-25 (2000-2006) layers_control = (/0.05,0.25,0.70,1.5/) layers_zeroflux = layers_control layers_stagger = (/0.05,0.15,0.25,0.40,0.65,1.05,1.70,2.75,4.45,7.20, \ 11.65,18.85/) layers_constant = (/0.05,0.25,0.70,1.5,2.5,3.5,4.5,5.5,6.5,7.5, \ 8.5,9.5,10.5,11.5,12.5,13.5,14.5,15.5,16.5,17.5/) layers_highres = (/0.01,0.03,0.05,0.07,0.09,0.11,0.13,0.15,0.17,0.19, \ 0.21,0.23,0.25,0.27,0.29,0.31,0.33,0.35,0.37,0.39, \ 0.425,0.475,0.525,0.575,0.625,0.675,0.75,0.85,0.95,1.1, \ 1.3,1.5,1.7,1.9,2.25,2.75,3.25,3.75,4.25,4.75/) layers_organic = layers_highres Organic layer (peat) in the top 12cm

4. 4 Point Simulation: Active Layer Thickness Black: control Blue: zeroflux Red: stagger Green: constant Orange: highres Brown-ish: organic 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Active Layer Thickness[cm] Computational Cost Control: 4 layers 12 layers1.56x 20 layers2.27x 40 layers3.51x

5. 5 Point Simulation: Active Layer Thickness Black: control Blue: zeroflux Red: stagger Green: constant Orange: highres Brown-ish: organic control_bias = 57.0 zeroflux_bias = 41.6 stagger_bias = 49.0 constant_bias = 50.2 highres_bias = 40.7 organic_bias = -0.9 Active Layer Thickness[cm] 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

6. 6 Point Simulation: Temperature Profiles Black: control Blue: zeroflux Red: stagger Green: constant Orange: highres Brown-ish: organic July January

7. 7 Point Simulation: Temperature Profiles Black: control Blue: zeroflux Red: stagger Green: constant Orange: highres Brown-ish: organic July January

8. 8 Point Simulation: Snow Depth Black: control Blue: zeroflux Red: stagger Green: constant Orange: highres Brown-ish: organic Snow depth much too low Snow Depth [cm] 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

9. 9 Point Simulation: Temperature Profiles July January Black: highres Blue: organic Red: organic_2x Green: organic_4x Artificially add precipitation to get deeper snow

10. 10 Point Simulation: Snow Depth Black: highres Blue: organic Red: snow z 0 Green: herb. tundra Changing z o over snow covered tundra brings model in line with observations Snow Depth [cm] 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

11. 11 Point Simulation: Active Layer Thickness highres_bias = 40.7 organic_bias = -0.9 snowz0_bias = 7.1 herb_tundra_bias = 1.8 Black: highres Blue: organic Red: snow z 0 Green: herb. tundra Active Layer Thickness[cm] 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

12. 12 Point Simulation: Temperature Profiles July January Black: organic Blue: snow z o Red: herb. tundra

13. 13 “Good” Sites: Coupled WRF Results WRF Precip WRF SWE Obs SWE Sublimation Melt

14. 14 “Bad” Sites: Coupled WRF Results

15. 15 Major reasons for early snowmelt in Noah at “bad” sites Higher downward solar radiation in WRF after adjusted to sub-grid terrain slope High melt amount in early spring –Raise maximum snow albedo to 0.85 Sublimation in winter –Reduce surface exchange coefficient (Ch) for nocturnal stable regime –Reduce surface roughness length (Zo) based on snow depth and forest height

16. 16 Effects of increasing maximum snow albedo Bad siteGood site Dash lines: Simulation with new max snow albedo Solid line: Default simulation Legend legend LV: Livneh albedo ML: model level forcing 85: max albedo set to 0.85

17. 17 Effects of reducing snow-surface Zo and Ch for stable regime Bad siteGood site Dash lines: Simulation with reduced Zo and Ch Solid line: Default simulation Legend legend LV: Livneh albedo ML: model level forcing ZE: Zo based on exposed veg SL: Slater stability

18. 18 New adjustment (in red) reduce Ch for stable regime Bad siteGood site Lower Ch in nighttime stable BL BL is stable through most of days 2 - 5 Legend legend LV: Livneh albedo ML: model level forcing CH: WRF MYJ stability

19. 19 Combined effects on long-term SWE Good sites Dash lines: Simulation with changes in Zo and albedo Solid line: Default simulation(contains new WRF Ch) By reducing sublimation and early spring melt, these mods delay snow melt Legend legend LV: Livneh albedo ML: model level forcing CH: WRF MYJ stability ZE: Zo = f(exposed veg) 85: Max albedo = 0.85

20. 20 Combined effects on long-term SWE bad sites Dash lines: Simulation with changes in Zo and albedo Solid line: Default simulation(contains new WRF Ch) By reducing sublimation and early spring melt, these mods improves timing of snow melt season Legend legend LV: Livneh albedo ML: model level forcing CH: WRF MYJ stability ZE: Zo = f(exposed veg) 85: Max albedo = 0.85

21. 21 Full grid effects of stability adjustment and all changes Left: Control Middle: Control + Ch Right: Control + Ch + albedo + Zo Effects are domain- wide but increased time to melt does not propagate out of mountains Legend legend LV: Livneh albedo ML: model level forcing CH: WRF MYJ stability ZE: Zo = f(exposed veg) 85: Max albedo = 0.85 SW: GOES SW forcing

22. 22 Comparison of HRLDAS Initial Soil Moisture 10 - 40 cm Nearly 8 year spin up of land states using HRLDAS and JRA25 forcing

23. 23 Comparison of HRLDAS Initial Soil Temperature 10 - 40 cm 40 - 100 cm

24. 24 Global MODIS Climatology 1km fPAR and LAI climatology on July 20 This is a single 10deg x 10deg MODIS tile in sinusoidal projection Hudson’s Bay

25. 25 Global MODIS Climatology A single point in the center of the previous MODIS tile Each dot represents a MODIS 8-day product value for one year Mean is shown in black; median in blue Some unresolved issues: filling in where data are missing; scaling fPAR to be equivalent to green vegetation fraction

26. 26 Slope-Aspect Adjustment for ASR Domain Tested slope and aspect adjustment based on terrain Bin results based on cardinal directions: North (-45°- 45°), etc. Results are consistent with terrain structures in domain

27. 27 Slope-Aspect Adjustment for ASR Domain However, if slopes < 1° are masked, the resulting locations where slope-aspect adjustment would make a difference are minimal 15km grid is too coarse to necessitate adjustment

28. 28 Short-term Plans Finish aggregation and quality control of CALM data with the ultimate goal to complete a short paper on the improved simulation of active layer depth Finalize MODIS products that will be used in ASR simulations and determine how they will be incorporated Complete tests of three-layer snow model Test run with and without HRLDAS initial conditions and run-time use of HRLDAS output and observed parameters Test run with new physics parameters and parameterizations to determine which should be used in ASR runs.

29. 29 Snow Roughness Length SUBROUTINE SNOWZ0 (SNCOVR,Z0, Z0BRD) ! ---------------------------------------------------------------------- ! SUBROUTINE SNOWZ0 ! ---------------------------------------------------------------------- ! CALCULATE TOTAL ROUGHNESS LENGTH OVER SNOW ! SNCOVR FRACTIONAL SNOW COVER ! Z0 ROUGHNESS LENGTH (m) ! Z0S SNOW ROUGHNESS LENGTH:=0.001 (m) ! ---------------------------------------------------------------------- IMPLICIT NONE REAL, INTENT(IN) :: SNCOVR, Z0BRD REAL, INTENT(OUT) :: Z0 REAL, PARAMETER :: Z0S=0.001 !m Z0 = (1.- SNCOVR)* Z0BRD + SNCOVR * Z0S Z0 = Z0BRD ! ---------------------------------------------------------------------- END SUBROUTINE SNOWZ0 ! ----------------------------------------------------------------------