| Abstract: |
MIRAS is an L-band Y-shape 2-D interferometric radiometer installed on board the ESA satellite SMOS, wich was launched in November 2009 with the scientific objective of measuring global Soil Moisture and Ocean Salinity. Since mid 2010 MIRAS has been providing well calibrated full-polarimetric brightness temperature maps of the Earth surface at a rate of one complete image of the four Stokes parameters every 2.4 seconds.
The SMOS level 1 operational processor retrieves the zero-spacing visibility samples from the overall antenna temperature measured by three stable Noise injection radiometers (NIR) installed at the instrument Hub. Since the second mission reprocessing, only one of the three NIRs is used due to relatively large temporal instabilities shown by the other two. But, even though the remaining NIR is more stable than the other two, it still seems to exhibit seasonal oscillations high enough to limit the quality of the scientific data, especially the ocean salinity long-term trends.
The Noise Injection Radiometers are calibrated using data acquired in "external maneuvers", with the instrument tilted to point the cold sky, which is performed every roughly two weeks. Using long series of observations over the ocean, it has been observed that temporal variations of NIR calibration parameters are translated into science measurement variations instead of stabilizing them. This fact suggests that the actual NIR behavior is much more stable than the calibration data suggests. The fact that the NIR is calibrated with the instrument looking to the sky, with different thermal conditions with respect to normal attitude, may contribute to this discrepancy.
An alternative method proposed by the authors to calibrate the NIR by using internal calibration events (every roughly two months) predicts a more stable calibration parameters, but still not enough to cancel long-term instabilities over the ocean. A possible solution seems to be using constant NIR calibration parameters. But this leads to the problem of estimating the most correct value for these parameters.
On the other hand, the zero-spacing visibility can also be retrieved from the average of antenna temperatures measured by all (or some) individual receivers forming the instrument, each one acting as a total power radiometer. This technique, referred to as "all-LICEF" has been demonstrated to provide high stability data [1], although this technique is very sensitive to random jumps observed quite often in diode-detected output voltages, probably caused by conducted EMC in DC circuitry.
In this paper, the absolute NIR calibration parameters are adjusted by performing an inter-calibration between all-LICEF and NIR antenna temperatures. One of the main advantages of this method is that NIR is calibrated with the instrument in nominal attitude. Also, this method allows to reduce the discrepancy between the zero-baseline amplitude and that of the other baselines, which has been shown to reduce the land-sea contamination effect in the images[2]
[1] Corbella, I., Gonzalez-Gambau, V., Torres, F., Duffo, N., Duran, I., Martin-Neira, M. “The MIRAS "All-LICEF" calibration mode” (2016) International Geoscience and Remote Sensing Symposium (IGARSS), 2016-July, pp. 2013-2016.
[2] Duran, I., Corbella, I., Torres, F., Duffo, N., Oliva, R., Martin-Neira, M.” Impact of amplitude calibration errors on SMOS global images” (2016) 14th Specialist Meeting on Microwave Radiometry and Remote Sensing of the Environment, MicroRad 2016 |
