Seasonal-to-Interannual Variability Of Latent And Sensible Heat Fluxes In The Tropical Atlantic Ocean

The knowledge of the surface air-sea heat fluxes variability on different space-time scales is vital to understanding the earth's climate change and the balance of global energy and fresh water budget. The studies of heat flux space-time variability are limited because of the poor qualities of heat fluxes datasets. Based on the importance of heat flux variability to the earth's climate change and the dependence of the study about heat flux variability on the accuracy of heat flux data, this study will focus on the quality evaluation of heat flux datasets available and seasonal and interannual variability of latent and sensible heat fluxes in the tropical Atlantic Ocean.The turbulent heat fluxes in WHOI (Woods Hole Ocean Oceanography Institute) synthesized fluxes product, NCEP1 and NCEP2 reanalysis fluxes product are evaluated using high-quality SOC flux climatology (Southampton Oceanography Center) which is constructed from ship observations and moored buoy observations PIRATA (Pilot Research Moored Array in the Tropical Atlantic) as the references. It is found that the biases in NCEP1 and NCEP2 turbulent heat fluxes from the biases in the algorithm used in flux calculation in trade wind regions are greater than equatorial regions. In trade wind regions, the biases in fluxes from the biases in algorithm are about 3-4 times of the biases in the fundamental flux-related variables and 1-2 times in equatorial regions. In the most area of tropical Atlantic, the evaluation of magnitude and time series trends of flux in WHOI are more accurate than NCEP1 and NCEP2. In equatorial regions, the air-sea humidity difference is overestimated in WHOI relative to PIRATA, which leads to the biases in latent heat flux. Generally, WHOI flux product is the most reliable and suitable for use.Seasonal variability of latent and sensible in the trade wind regions of tropical Atlantic is examined using WHOI flux product. It is found that the climatological mean of wind speed and air-sea humidity difference are both large, the variation of wind speed are almost in phase with air-sea humidity difference, yielding much larger or smaller latent heat flux. So the ocean release the most latent heat in its own winter of the two hemispheres when both wind speed and air-sea humidity difference are large. In equatorial region, the mean air-sea humidity difference is much smaller, so the variations of air-sea humidity difference under large mean wind speed dominate the variations of latent heat flux. The ocean release the most heat during the period from March to May, when SST is highest in the whole year because SST dominates the variation of air-sea humidity difference. In Climatological ITCZ, the mean air-sea humidity difference is greater relatively so that the variation of wind speed under large mean air-sea humidity difference is the main factor influencing the variation of latent heat flux. The latent heat loss is greatest from June to August when the windspeed is largest in the region. In the region north of Brazil, although both wind speed and air-sea humidity difference have dramatic variations, the variation of latent heat flux is just ordinary in magnitude. The reason is that the variations of wind speed and air-sea humidity difference are out of phase with each other so that they decrease each other greatly. The variation of wind speed dominates the variation of latent heat flux eventually. The variation of sensible heat flux is dominated by air-sea temperature difference.Interannual variability of turbulent heat fluxes is examined using SVD, EOF and MTM methods. It is found that two main EOF modes of turbulent heat fluxes mainly describe the flux variations in two trade wind regions. The MTM spectrum of the first time series of expansion coefficient shows a periodic signal at about two-to-three months and two-to-three years quasi-periodic signal (at the 95% level). The second mode shows a peak at 2-3 years. In equatorial cold tongue region, SST anomalies have great influence on the variation of turbulent heat fluxes. W...