Latent and Sensible Heat flux comparisons

Turbulent heat fluxes were computed using CLIMODE mooring data and the Coupled Ocean Atmospheric Response Experiment (COARE) bulk air-sea flux algorithm Version 3.0. The data collected at the mooring included air temperature and humidity, sea surface temperature and surface wind speed. 

We compare the CLIMODE fluxes with several turbulent flux products listed below using a Taylor diagram. These analyses were conducted in 2009 using products available at the time.

  • ECMWF
  • ECMWF geophysical variables run through the COARE algorithm
  • NCEP2
  • NCEP2 geophysical variables run through the COARE algorithm
  • OAFlux produced at Woods Hole Oceanographic Instituion (WHOI)
  • COARE computed flux using QuikSCAT for wind speed, AMSR SST from Remote Sensing Systems and ECMWF for remainder of input variables
  • COARE computed flux using QuikSCAT for wind speed, OISST from NOAA using both microwave and infrared SST and ECMWF for remainder of input variable
 CLIMODE mooring location
Location of CLIMODE mooring

latent heat flux comparison

Latent Heat Comparison

The ECMWF latent heat product (green dot) has a correlation of about 0.83 and the same relative magnitude as the CLIMODE flux. When ECMWF geophysical variables are run through the COARE algorithm (black circle), the correlation is increased, the energy level is reduced, and the normalized error is improved. The NCEP2 latent heat product (black dot) has a very high energy level with a correlation of about 0.8. When NCEP2 variables are put through the COARE algorithm (yellow dot rimmed with red), the correlation is increased and the energy level is lower. The OAFlux product (magenta) shows a correlation of about 0.85 with an energy level comparable to the CLIMODE flux. Using the COARE algorithm, two latent heat products using satellite data were produced, using QuikSCAT wind speed and ECMWF air temperature and humidity. The results are comparable using either AMSR microwave SST (cyan dot) or NOAA OISST (blue dot), which contains both microwave and infrared data, with correlations above 0.9, an energy level a bit higher than the observed, and low normalized errors.
Sensible heat Taylor diagram
Sensible Heat Comparison

The sensible heat comparison results are similar to that of the latent heat in that the use of the COARE algorithm for ECMWF and NCEP2 reduces the normalized error. The satellite products again have higher correlations and lower errors than the other products. A difference in the sensible heat comparison is that all of the products have higher energy than the CLIMODE mooring flux.