Intercomparison Of Air-sea Latent Heat Flux And Sensible Heat Flux Products

Ocean and atmosphere are two main components of the earth climate system.The ocean influences atmosphere by transferring heat energy, while atmospherechanges the ocean currents and redistribute ocean heat through wind stress. The latentheat flux (LHF) and sensible hest flux (SHF) at the air-sea interface have ansignificant impact on global energy and water cycle balance. Five monthly globallatent heat flux and sensible heat flux products during1996~2005are compared anddiscussed in this paper.Firstly, monthly LHF of the Goddard Satellite-Based Surface Turbulent Fluxes,version3(GSSTF3), the Hamburg Ocean Atmosphere Parameters and Fluxes fromSatellite Dataset version3(HOAPS3), the Japanese Ocean Flux Datasets with UsedRemote Sensing Observations version2(J-OFURO2), the National Centre forEnvironmental Prediction and the National Center for Atmospheric ResearchReanalysis dataset (NCEP reanalysis) and the Objectively Analyzed air-sea Fluxes(OAFLUX) are compared and analyzed. The products show a similar patterns ofglobal and zonal distribution, larger differences mainly locate in the trade winds zoneand high latitude area, the mean LHF of three satellite-derived products, i.e. GSSTF3,HOAPS3, J-OFURO2are closer with root mean square error(RMSE) between eachgroup of the two products less than20W/m2and correlations above0.90. OAFLUXand NCEP reanalysis are closer, root mean square error(RMSE) of them is also lessthan20W/m2, correlations is higher than0.90. The lowest RMSE is between GSSTF3and HOAPS3is14.08W/m2, the highest RMSE is between NCEP and GSSTF3withthe value of25.98W/m2. The lowest correlation is between NCEP and HOAPS3aswell as GSSTF3and OAFLUX with the value of0.87. The highest correlation isbetween GSSTF3and HOAPS3with the value of0.97.Secondly, monthly SHF of five products are compared and analyzed, the resultshows that the mean SHF of HOAPS3is larger than other four products in the high latitude area of Southern hemisphere. Zonal distribution pattern of all products in areaof40°S~60°N are similar, large differences mainly located in the high latitude of bothhemispheres. RMSE between HOPAS3and other four products are higher than10W/m2, and the correlations are lower than0.70. The lowest RMSE is betweenJ-OFURO2and OAFLUX with the value of6.29W/m2, the highest RMSE is betweenJ-OFURO2and HOAPS3with the value of15.92W/m2. The lowest mean correlationis between J-OFURO2and HOAPS3with the value of0.45, the highest meancorrelation is between GSSTF3and NCEP reanalysis with the value of0.93.The Latent heat flux and sensible heat flux calculated based on Dong Fanghong2is compared with OAFLUX, the rmse,bias and stdv of latent heat flux betweenthem are28.80W/m2,-19.95W/m2and20.93W/m2, the rmse,bias and stdv of latentheat flux are6.16W/m2,-2.39W/m2and5.72W/m2.The interconparison of the monthly LHF and SHF data of the five productsshows that, all products have a similar pattern of global and zonal distribution. Thelarger difference of LHF is located in the trade winds zone and high latitude area. Thelarger difference of SHF is located in high latitude area. The LHF data of threesatellite-derived products are quite different from NCEP reanalysis and OAFLUX.The SHF of HOAPS3has larger difference from the other four products.