The measurement of the synoptic scale wind over the ocean

Abstract

Mesoscale and microscale features of the turbulent winds over the ocean are related to the synoptic xscale winds in terms of published spectral forms for the microscale, a mesoscale valley, and published values of u*, Var u′, Var v′, and z/L, as defined in the text and as obtained for moderate to gale force winds. The frequencies involved correspond to periods longer than 1 hour and extend to the microscale, which starts at a period near 2 min, or so, and continues to the Kolmogorov inertial range. Nondimensional spectra that span both the mesoscale and the microscale are derived as a function of u*,f/(=nz/ū), and z/L, where z is 10 m, L is the Monin‐Obukhov stability length, and ū is evaluated at 10 m. For the same ū, different values of z/L produce a range of values of u* which in turn result in variations of the eddy structure of the mesoscale and microscale spectra. Both conventional anemometer averages and remotely sensed winds contain a random component of the mesoscale wind in their values. These components are differences and not ‘errors’ when winds are compared, and quantitative values for these differences are given. Ways to improve the measurement of the synoptic scale wind by transient ships, data buoys, and scatterometers on future spacecraft are described. These ways are longer averaging times for ships and data buoys, depending on the synoptic conditions, and pooling spacecraft data to form superobservations. Design considerations for future remote sensing systems are given.

Synoptic Scale Wind Field Properties From The SEASAT SASS

“…_¢aenlarge volumes of data are processed such as for the SASS, and its successors, an average of one out of ten variance estimateswill be outside the 90% Chi Square confidenceinterval. Pierson (1983a)has treated aspects of this part of the analysis of superobservations (See page 1703) with the further assumptionthat the random variabilityis normally distributed.…”

“…If it is wrong and agrees with the update, the update is, nevertheless, still wrong. Given the sources of error introduced by Beaufort reports and the conventional two minute averages for transient ships discussed by Pierson (1983a), one can hardly dare to hope that a given planetary boundary layer initial value specification is correct to within the limits described by . With an incorrect wind field, an incorrect specification of the sea surface atmospheric pressure field, and probably the thermal wind for the first few kilometers, the correct specification of the entire atmospheric column, even with remote soundings from spacecraft, becomes doubtful.…”

“…A single isolated ship report in a oceanic area five degrees of longitude by five degrees of latitude with all of its inherent sources of error (Pierson (1983a)) can influence a very large area of the ocean, even outside of the five degree by five degree area. If the report is so much in disagreement with the first guess initial value specification (a forecast from a past initial value update) that it is disregarded, it does no good (or harm) at all.…”

“…The terms AUm, AUc, AVm and AVc have a new meaning and must be interpreted in terms of the analysis given by Pierson (1983a) The terms, AUm and AVm, represent the contribution to a particular SASS wind of the mesoscale variability of the wind for each SASS cell. The terms, AU c and AVe, represent the communication noise error in the measurement made by SASS for each pair of cells.…”

“…Pierson (1982aPierson ( ,1983b,assumed that gradients of the synoptic scale wind could be neglected in the study of superobservations. For some parts of the wind fields obtained from the SASS, the assumption is not justified, and this correction needs to be made.…”