1. The ocean and the global hydrologic cycle Jim Carton (University of Maryland) Paulo Nobre (INPE) São Paulo Summer School on Global Climate Modeling October, 2011 Guiding Question: Does the oceanic component of the hydrologic cycle vary, and if so what are the consequences for climate?

2. Outline Overview: –mean global salt and freshwater budgets Measuring the oceanic hydrologic cycle Observed and modeled trends Connection between salinity trends and climate Introduction to ocean modeling Carton

3. The global hydrologic cycle Dai and Trenberth Carton

4. Some Numbers Net evaporation from the ocean: 1.2 m/yr = 3.8x10 -8 m/s Area of the ocean: 3.6x10 14 m 2 Surface volume flux due to evap: 13.7x10 6 m 3 /s or 13.7Sv (similar to the rate of formation of North Atlantic Deep Water!) Amazon River discharge: 0.2x10 6 m 3 /s (so think of evaporation as: 70 Amazons! Carton

5. How does this freshwater budget relate to the salinity budget? First a little math Mass is conserved, so storage and advection must balance net surface flux. The mass budget is: Salinity must be conserved, but for salinity there are no sources or sinks! (except a very weak river source) (2 (2) Can be rewritten using (1) to give us the salt budget: (4 (1 Storage + advection = effective net surface flux Carton

6. Climatological sea surface salinity In warm water: So, change in density due to 1psu is equivalent to 5C Stephens et al., NOAA, 2002 Carton

7. Stephens et al., NOAA, 2002 Observed Surface Temperature

8. Observed E-P Carton

9. Surface meteorology Ocean Circulation and Climate

10. Annual Range of Salinity Carton

11. Observing systems Carton

12. Aquarius 7dy Argo Present and future salinity sampling PIRATA Carton

13. Current Argo distribution Carton

14. Salinity from the PIRATA mooring at 21N, 23W: evidence of eddies! Carton

15. Time series at the 21N, 23W PIRATA mooring Carton EVAP

16. Carton Sea surface salinity Lagerloef et al., 2007: The Aquarius/SAC-D Mission: Designed to Meet the Salinity Remote-Sensing Challenge, Oceanography Magazine. Microwave brightness temperature varies with salinity. Panel to the right shows the variations of radiation expected from a flat surface (no waves). Note that the dependence is highest at higher temperatures. Aquarius exploits this dependence to obtain an SSS measurement with an expected ~0.2PSU accuracy at monthly timescales.

17. Carton WOW!

18. Trends Carton

19. (psu/50yr) surface salinity linear trend Durack and Wijffels (2010) Observed 50yr drying trend over Africa Held et. al., 2005 psu/50yr Carton

20. Vertical structure of the salinity change in the Atlantic (zonally averaged) Carton

21. Freshening trends in the deep North Atlantic Dickson, et. al., Nature, 2002

22. Surface salinity change in an atmosphere-ocean coupled GCM (CM2.1) in response to elevated CO 2 Stouffer et al. (2006) Carton

23. Salinity and climate Carton Heat and freshwater cycles are linked directly through the relationship between latent heat loss and evaporation: –0.6 Sv Freshwater loss corresponds to 1.5 petawatts of latent heat flux They are linked indirectly through the impact of salinity changes on density.

24. How the equation of state depends on salinity Carton

25. The meridional overturning circulation Carton

26. Heat transport by the oceans Houghton et al., (1996: 212) Carton

27. Some Numbers Assume 15Sv = 1.5x10 6 m 3 /s northwards at the surface and southwards below 2000m depth Assume a 15 o C temperature difference between the two flows The net heat transport is: 1.5x10 7 x 15 x 4x10 6 = 0.9x10 15 W! or nearly the total amount of heat being transported northwards in the North Atlantic. Carton

28. Freshwater transport Jourdan et al., J. Phys. Oceanogr., 27, 457-467, 1997 Carton

29. Will the Atlantic MOC change in response to increasing greenhouse gasses? IPCC 4 th Assessment Carton

30. Ocean general circulation modeling Carton

31. Basic Equations Carton

32. Numerics Arakawa-B grid in horizontal (2 nd order) Upstream advection Leap frog time differencing Rigid lid (w(z=0)=0) Separate internal and extenral modes SHMEM, MPI, shared memory, multi-threaded version u,v T,S Arakawa-B u,v Carton

33. Our model General circulation ocean model using POP2 numerics Global grid 0.1 o x0.1 o x42Levels (3x10 8 ) ‘Normal year’ forcing. 64yr spinup Compute/save full salt budgets (1yr so far) 1 year requires 12K PE hrs on an IBM Power6 10% of actual resolution Carton

34. Observations Model Simulation Carton

37. What terms balance E-P in the salt budget of the mixed layer? ? ? Do eddies contribute? Carton

38. Mean salt balance 0-100m average Cool color means exporting salt Carton

39. What we’ve learned Oceanic hydrologic cycle overview –The oceans play a central role in the global hydrologic cycle –Patterns of surface salinity mainly reflect patterns of E-P Observing systems: rapidly improving –ARGO born 2001 –Aquarius born 2011 Salinity trends: –In the past 50 years salty places are getting saltier, fresh places are getting fresher –In particular, the subpolar North Atlantic has been getting fresher –These results seem to be consistent with CO 2 effects based on coupled models –The implications for the meridional overturning circulation (AMOC) are still not clear Ocean General Circulation Modeling –We’ll use this tool to look at the salt budget of the upper 100m. High salinity ‘ocean deserts’ (source waters for the tropical thermocline) are maintained by: 1) surface evaporation, 2) poleward wind-driven transport of freshwater, and 3) horizontal eddy exchanges. How will they change? The ‘dry’ parts of the ocean have been getting saltier possibly reflecting an intensification of the hydrologic cycle. What will this mean for Saudi Arabia? – What changes have occurred historically? – How can we improve our guesses about future conditions? –What are the impacts of these physical changes on marine biogeochemical processes? Carton

40. Does the oceanic component of the hydrologic cycle vary, and if so what are the consequences for climate? The answer to the first part is clearly yes. But what about the second?