A New Air–Sea Interaction Gridded Dataset from ICOADS With Uncertainty Estimates

Abstract

A new method of analyzing weather reports from merchant ships leads to improved in situ air-sea heat flux datasets and information about the uncertainty of the flux estimates.T he ocean surface heat budget comprises radiative and turbulent components: the radiative components are solar shortwave and thermal longwave, both of which are strongly affected by cloud; the turbulent heat flux components are the direct transfer of sensible heat and the evaporative transfer of latent heat. Although direct measurements of the turbulent fluxes are possible, they remain a research activity due to the high cost of the measurements. Measurement of the radiative components is easier but still challenging on long-term deployments. As a result, global marine flux datasets are typically constructed using bulk estimates of the mean meteorological parameters and flux parameterizations, known as "bulk formulas."Various sources exist for the bulk meteorological parameters, ranging from model output and satellite observations to •

A comparison of nine monthly air–sea flux products

“…This is believed to be due to systematically larger values of q air in the FSU3, reducing the air-sea moisture gradient. This bias in humidity is largely consistent with the adjustments made in a new NOC product (L. Kent, personal communication, 2008;Berry and Kent, 2009). The JRA product also has relatively large values of q air , but retains large LHFs because the value of near-surface humidity is overestimated.…”

“…In addition, previous comparisons (Chou et al, 2003;Kubota et al, 2003) have shown Da Silva et al (1994) to have severe limitations related to sampling patterns (particularly in the southern ocean (SO) where the product has very little variability), and some versions of the product have biases related to assumptions regarding ocean budgets (Da Silva et al, 1994). Also excluded is the 2009 update of the NOC-Southampton in situ analysis (Berry and Kent, 2009 Bentamy (personal communication, 2007), the IFREMER surface turbulent fluxes are created using physical properties of active and passive satellite instrument measurements, empirical and Table I. The nine flux products compared, including the product type, original spatial grid, and temporal period (month/year) when the monthly fluxes are available.…”

ICOADS Release 2.5: extensions and enhancements to the surface marine meteorological archive

“…In these circumstances, instrument type and placement are critical metadata that can lead to improved accuracy. Many global ocean flux datasets have been derived from ICOADS in the past two decades (da Silva et al, 1994;Josey et al, 1999;Smith et al, 2004;Bourassa et al, 2005;Berry and Kent, 2009).…”

A composite look at short‐time‐scale sea‐surface temperature changes in the western North Pacific based on ships and buoys

“…Temporal and spatial (error) covariance, perhaps oriented along an SST gradient (Ishii et al, 2003), would also need to be postulated. In turn, an explicit measure of analysis error could then be obtained (Berry and Kent, 2009). It should also be possible to compute the corresponding contributions of each measurement method (as in Figure 8), although experiments to remove such a contribution would require slightly more effort, as analyses would depend on observations prior to a 3 d window.…”

“…In any case, future work on observed cyclone impacts will need to accommodate interesting challenges of interpretation from all platforms and methods of measurement. This should be facilitated both by improvements in the analysis approach (Brasnett, 1997;Ishii et al, 2005;Berry and Kent, 2009) and by a growing familiarity with the use of ancillary information (Kent et al, 2007a). Table I.…”