Error sources and feasibility for microwave remote sensing of ocean surface salinity

A set of geophysical error sources for the microwave remote sensing of ocean surface salinity have been examined. The error sources include the sea surface temperature, sea surface roughness, atmospheric gases, ionospheric Faraday rotation, and solar and galactic emission sources. I t is shown that the brightness temperature errors of a few Kelvin can be expected for most of these error sources. The key correction requirements for accurate salinity measurements are the knowledge accuracy of 0.5"C for the sea surface temperature (SST), 10°C for the surface air temperature, 0.2" accuracy for the Faraday rotation, and surface roughness equivalent to 0.3 mas-' for the surface wind speed. We suggest the use of data products from AMSR-type instruments for the corrections of SST and liquid cloud water, the numerical weather analysis for the air temperature, coincidental radar observations with 0.2 dB precision for surface roughness and on-board polarimetric radiometer channel for Faraday rotation. The most significant sky radiation is from the sun. A careful design of the antenna is necessary to minimize the leakage of solar radiation or reflection into the antenna sidelobe. The narrow-band radiation from galactic hydrogen clouds with a bandwidth of less than 1 Mhz is also significant, but can be corrected with an accurate radio sky survey or minimized with a notched (band-rejection) filter centered at 1.420 GHz in the receiver. The other planetary and galactic radio sources can be flagged with negligible data loss. We have performed a sampling analysis for a polar-orbiting satellite with 900 km swath width to determine the number of satellite observations over a given surface grid cell during an extended period. Under the assumption that the observations from different satellite passes are independent, it is suggested that an accuracy of 0.1 psu is achievable for global monthly 1-degree latitude and 1-degree longitude gridded products.