Satellite surface salinity refers to measurements of surface salinity made by remote sensing satellites. The radiative properties of the ocean surface are exploited in order to estimate the salinity of the water's surface layer.
The depth of the water column that a satellite surface salinity measurement is sensitive to depends on the frequency (or wavelength) of the radiance that is being measured. For instance, the optical depth for seawater at the 1.413 GHz microwave frequency, used for the Aquarius mission, is about 1–2 cm.
As with many passive remote sensing satellite products, satellites measure surface salinity by initially taking radiance measurements emitted by the Earth's atmosphere and ocean. If the object emitting the measured radiance is considered to be a black body, then the relationship between the object's temperature and the measured radiance can be related, at a given frequency, through the Planck function (or Planck's law).
I\nu=
| 2h\nu3 | |
| c2 |
| 1 | |||||||||
|
I\nu
\nu
\nu+d\nu
T
h
\nu
c
k
For ideal black bodies, the brightness temperature is also the directly measurable temperature. For objects in nature, often called Gray Bodies, the actual temperature is only a fraction of the brightness temperature. The fraction of brightness temperature to actual temperature is defined as the emissivity. The relationship between brightness temperature and temperature can be written as:
Tb=eT
Studies have shown that measurements of seawater brightness temperature at the 1.413 GHz (L-band) are sufficient to make reasonably accurate measurements of seawater surface salinity.[2] [3] The emissivity of seawater can be described in terms of its polarized components of emissivity as:
eH=1-\left[
| ||||||||||||
|
\right]2
eV=1-\left[
| ||||||||||||
|
\right]2
The above equations are governed by the Fresnel equations, the instrument viewing angle from nadir θ, and the dielectric coefficient ε. Microwave radiometers can be further equipped to measure the vertical and horizontal components of the surface seawater's brightness temperature, which relates to the horizontal and vertical components of the emissivity as:
TbH=eHT
TbV=eVT
Tb
T
Several models have been proposed to estimate the dielectric constant of sea water given its salinity and temperature.[4] The "Klein and Swift" dielectric model function is a common and well-tested model used to compute the dielectric coefficient of seawater at a given salinity, temperature, and frequency. The Klein and Swift model is based on the Debye equation and fitted with laboratory measurements of the dielectric coefficient.[5] Using this model, if the temperature of the seawater is known from external sources, then measurements of the brightness temperature can be used to compute the salinity of surface seawater directly. Figure 1 shows an example of the brightness temperature curves associated with sea surface salinity, as a function of sea surface temperature.
When looking at the polarized components of the brightness temperature, the spread of the brightness temperature curves will be different depending on the component. The vertical component of the brightness temperature shows a greater spread in constant salinity curves than the horizontal component. This implies a greater sensitivity to salinity in the vertical component of brightness temperature than in the horizontal.
There are many sources of error associated with measurements of sea surface salinity:
Most of the error sources on the previous list stem from either standard instrument errors (Antenna, System Pointing, etc.) or noise from external sources measurement signal (Solar, Galactic, etc.). However, the largest error source comes from the effect of ocean surface roughness. A rough ocean surface tends to cause an increase in the measured brightness temperature [6] as a result of multiple scattering and shadowing effects.[7] Quantifying the influence of ocean roughness to the measured temperature brightness is crucial to make an accurate measurement. Some instruments use radar scatterometers to measure the surface roughness to account for this source of error.[8]