Kaiyuan Y. Li and Dennis P. Lettenmaier University of Washington

1. Upgrading Community Land Model (CLM) HydrologyIncorporation of the VIC Surface Runoff and Baseflow Schemes Kaiyuan Y. Li and Dennis P. Lettenmaier University of Washington

2. Outline • What is CLM? • What are the problems with CLM? • How is VIC fitted to CLM? • How does the upgraded CLM work?

3. What is CLM? • Stands for Community Land Model (used to be Common Land Model); • Community-developed land surface model: lead by NCAR in collaboration with universities; • Intended to be coupled into CAM or CCSM; • A framework into which other land surface models can be feasibly incorporated; • Professional software written in FORTRAN90 (similar to C with pointer and dynamic memory); • Optimum performance on both cache-dependent and vector-based computational platforms (parallel computing).

4. CLM: data structure

5. CLM: functionalities • Vegetation composition, structure, and phenology. • Absorption, reflection, and transmittance of solar radiation. • Absorption and emission of longwave radiation. • Momentum, sensible heat (ground and canopy), and latent heat (ground evaporation, canopy evaporation, transpiration) fluxes. • Heat transfer in soil and snow including phase change. • Canopy hydrology (interception, throughfall, and drip). • Snow hydrology (snow accumulation and melt, compaction, water transfer between snow layers) • Soil hydrology (surface runoff, infiltration, subsurface drainage, redistribution of water within the column). • Stomatal physiology and photosysnthesis. • Lake temperatures and fluxes. • Routing of runoff from rivers to ocean. • Volatile organic compounds. • Vegetation dynamics and carbon cycle -- coming soon.

6. CLM Hydrology • Canopy hydrology • Interception; • Throughfall; • Drip. • Soil Hydrology • Surface runoff (Based on TOPMODEL); • Baseflow; • Soil water (based on Richard’s flow equation). • Snow Hydrology • Based on Jordan (1991).

7. CLM EvaluationThe FIFE Prairie Site (39.0ºN 96.5ºW) • CLM overestimates surface runoff; • CLM underestimates latent heat and soil moisture contents.

8. CLM EvaluationThe ABRACOS Forest Site (10.1ºS 61.9ºW) • CLM overestimated surface runoff; • CLM underestimated latent heat and overestimated sensible heat; • CLM poorly simulated soil moisture and evapotranspiration.

9. CLM EvaluationThe Valdai Grassland Site (57.6ºN 33.1ºE) • CLM poorly simulated soil moisture content.

10. CLM EvaluationThe HAPEX-MOBILHY Soybean Site (43.7ºN 0.1ºW)

11. CLM EvaluationThe HAPEX-MOBILHY Soybean Site (43.7ºN 0.1ºW) – Cont. • CLM overestimated runoff; • CLM undersimulated latent heat and overestimated sensible heat; • CLM poorly simulated soil moisture content.

12. CLM EvaluationTorne-Kalix Basin • CLM tends to overestimate runoff peak for most basins; • CLM tends to melt snow earlier than observed.

13. CLM EvaluationArkansas and Red river basin

14. CLM EvaluationArkansas and Red river basin – Cont.

15. CLM EvaluationArkansas and Red river basin – Cont. • CLM tend to overestimate runoff

16. CLM EvaluationThe Colorado river basin

17. CLM EvaluationThe Colorado river basin – Cont.

18. CLM EvaluationThe Colorado river basin – Cont. • CLM overestimated runoff; • CLM melt snow one month earlier;

19. CLM EvaluationConclusions • CLM tends to overestimate runoff; • CLM tends to melt snow earlier; • CLM poorly simulates soil moisture contents; • In general, VIC performs better than CLM in terms of hydrologic predictions; • Improvements of CLM are expected by incorporating some aspects of VIC hydrologic parameterizations into CLM.

20. 10-Layer CLM 3-Layer VIC VIC Upper Layer 3.43 m VIC Lower Layer Dynamic Depth Fixed Depth Diagram: Matching CLM layer scheme to VIC Layer scheme Upgrading CLMIncorporation of the VIC surface runoff and baseflow schemes

21. Upgrading CLMIncorporation of the VIC surface runoff and baseflow schemes – Cont. • The Upgraded CLM retains: • Data structure; • Input and output format; • Model structure. • The Upgraded CLM requires 5 VIC parameters: • Upper layer depth (first plus second layer in VIC-3L); • b: Infiltration parameter; • Ws: Fraction of maximum soil moisture content when baseflow occurs; • Dsmax: Maximum velocity of baseflow; • Ds: Fraction of Dsmax where non-linear baseflow occurs.

22. Performance testing of Upgraded CLMThe Arkansas-Red basin

23. Performance testing of Upgraded CLMThe Arkansas-Red basin – Cont.

24. Performance testing of Upgraded CLMThe Colorado basin

25. Performance testing of Upgraded CLMThe Colorado basin – Cont.

26. Performance testing of Upgraded CLMThe FIFE Prairie Site (39.0ºN 96.5ºW)

27. Performance testing of Upgraded CLMThe FIFE Prairie Site (39.0ºN 96.5ºW) – Cont.

28. Performance testing of Upgraded CLMThe ABRACOS Forest Site (10.1ºS 61.9ºW)

29. Performance testing of Upgraded CLMThe ABRACOS Forest Site (10.1ºS 61.9ºW) – Cont.

30. Performance testing of Upgraded CLMConclusions • The Upgraded CLM, into which VIC runoff parameterization is incorporated, performs significantly better than original CLM; • The Upgraded CLM requires only 5 VIC parameters, which are transferable to CLM without massive calibration although some systematical adjustment may be required for some basins; • The enhancement of the CLM snow model remains to be done.

31. Acknowledgement Ted Bohn Fengge Su Jenny Adam Joanna Gaski Chunmei Zhu Niklas Christensen Mariza Costa-Cbral Nathalie Voisin Kostas Andreadis Alan Hamlet Lan Cuo