Marine Stratus and Its Relationship to

1. Marine Stratus and Its Relationship to Regional and Large-Scale Circulations:An Examination with the NCEP CFS SimulationsP. Xie1), W. Wang1), W. Higgins1), and P.A. Arkin2)1) NOAA Climate Prediction Center2) ESSIC, Univ. of Maryland

2. Objectives: • To examine the CFS model bias associated with the insufficiently simulated marine stratus clouds over SE Pacific and SE Atlantic; • To investigate the atmospheric circulation involving the formation and variations of the marine stratus clouds;

3. Two Sets of CFS Simulations Examined • CFS AMIP Simulations [GFS] • The atmospheric component of the CFS model (GFS03) is forced by observed oceanic condition; • 1979 – 2003; • CFS CMIP Simulations [CFS] • The GFS03 atmospheric model is fully coupled with an OGCM (MOM3) ; • 4 sets of CMIP simulations for a 32-year period each

4. Annual Mean Precipitation • Large-scale precipitation patterns reproduced reasonably well; • Differences exist in the magnitude of precipitation and in the latitudinal position of the ITCZ;

5. Annual Cycle of the Atlantic ITCZ • GFS simulates annual cycle pretty well, with excessive precipitation; • The Atlantic ITCZ located too south during DJF and MAM in the CFS simulation;

6. Annual Cycle of E. Pacific ITCZ • Relatively good reproduction of precipitation fields by GFS; • ITCZ too strong and located a little bit too south during DJF and MAM in the CFS;

7. Annual MeanSST / Surf. Wind • Warm SST bias in the CFS over the southeast Atlantic and southeast Pacific; • Surface wind bias over the regions of warm SST bias.

8. Possible Causes for the Warm SST Bias over the Regions 1. Insufficient vertical transportation  oceanic observations (will not check this time) 2. Enhanced heat transport through advection  surface wind speed 3. Reduced evaporation  Surface wind speed 4. Excessive incoming solar radiation  cloudiness

9. Annual MeanTotal Cloud (%) • In general, CFS cloud amount is smaller than that of observations almost everywhere; • Insufficient amount of clouds simulated by the CFS over the regions with warm SST bias;

10. Cloudiness in other NCEP Products • Only CDAS1 reproduced the cloudiness (mostly low clouds) reasonably well; • CDAS2, GFS and CFS failed to generate cloud amounts over the regions;

11. Annual MeanSW Radiation • Excessive incoming solar radiation over both the SE Atlantic and SE Pacific; • Differences of over 50W/m2 over SE Pacific;

12. Comparison with EPIC Buoy Data at [20oS,85oW] • Close agreements between buoy and satellite observations; • Warm SST bias of ~2oC and positive SW radiation of ~50W/m2 in CFS compared to the EPIC buoy observations;

13. Cloud Picture over SE Pacific [from fig.3 of Bretherton et al. (2004)] • Scattered stratus clouds over the regions;

14. Vertical Profiles of Boundary Layer [from fig.10 of Bretherton et al. (2004)] • Well mixed boundary layer of 1-1.5 km capped by inversion; • Stratus cloud layer of ~500m atop the boundary layer; Liquid Water Water Vapor Temperature

15. Diurnal Cycle in ISCCP cloudiness • 24-hour mean cloudiness (top) shows a bi-polar structure over the dry zone and nearby continent, suggesting a diurnal cycle cell caused by sea breeze; • 3-hourly ISCCP cloud data (bottom) presents distinct diurnal cycles of different phases over the dry zone and the land area;

16. Diurnal Cycle in CDAS1 Cloudiness

17. Diurnal Cycle in circulation from CDAS1 06Z • 24-hourly mean removed to examine the diurnal cycle; • Cells of regional circulations involving land-sea contrasts;

18. Summary • Overall good performance of CFS model in reproducing large-scale precipitation patterns; • Merdional shifts of ITCZ over Eastern Pacific and Atlantic sectors; • The displacements of the ITCZ closely related to the warm SST bias in the SE Atlantic and SE Pacific stratus deck regions; • The warm SST bias caused largely by insufficiently simulated stratus clouds; and • Strong diurnal cycle in the stratus clouds generated by regional circulation caused by land-sea contrasts between the oceanic regions and their adjacent continents.

19. Annual MeanSurface Wind • Small differences in surface wind speed between the CFS model and observations (QuikScatter) Over the SE Atlantic and SE Pacific;