Test Observations
Tests we periodically perform to analyze sensor performance.
Last updated
Tests we periodically perform to analyze sensor performance.
Last updated
The sun emits microwave radiation in the L-band, which can be detected with our sensor. The idea is to let the sun drift through the main beam of the sensor without moving the sensor itself. Ideally, this should be done during an entire day, from sunrise until sunset, since it already takes two hours for the sun to travel through the antenna's main beam. The sun will raise the measured brightness temperature by several Kelvin as it drifts through the antenna beam.
The brightness temperature of the solar emission depends strongly on the level of solar activity as expressed in the so-called solar flux unit (sfu).
Daily sfu values are measured at several locations worldwide. The Canadian observations can be obtained here, for example.
Without going into the details, the approximate increase in brightness temperature that can be expected is:
EQUATION HERE
For this test to be succesful, it is important to set up the sensor's azimuth and elevation angle correctly to ensure the main beam will point at the sun around noon.
The user may ask, why not speed up this test and rotate the antenna in azimuth? The answer is that keeping the sensor fixed will result in a much more stable baseline. It would be difficult to distinguish changes in the baseline level caused by the movement of the sensor and the emission of the sun.
The increase in brightness temperature caused by the sun is comparable to the change in soil's brightness temperature caused by changes in moisture content. The point of this test is to confirm that the sensor has enough sensitivity to detect changes in the soil's brightness temperature due to changes in soil moisture.
Below is the result of a typical solar drift scan. The sun traversed through the antenna's main beam at around 13:30 local time, resulting in a 30 Kelvin increase in receiver system temperature. The blue points show the total power signal, and the red line shows the unpolarized component only. Assuming that any polarized emission is due to radio interference, the red line is a "filtered" version of the data.
With this test, we want to look at the difference in signal level between a "hot-load" and a "cold-load". This test makes it possible to a) calibrate the sensor output in Kelvin and b) determine the receiver noise temperature.
As cold-load, we can use the sky. The brightness temperature of the sky is 3K. If we point the antenna straight up, we should theoretically get a signal mainly from the receiver itself (i.e. the receiver components that contribute to its total noise figure.) The problem here is that we will always have contributions from emissions received through the sidelobes or the back lobe of the antenna. The actual cold-load temperature will more likely be around 20 to 30K.
As hot-load, we use a large piece of microwave-absorbing foam. If the absorber is placed directly in front of the antenna, it acts as a resistor. Its brightness temperature is equal to its physical temperature (around 300K). The absorber should ideally be placed at least 50cm away from the antenna while filling the entire main beam.
Alternatively, if no absorber is available, the antenna can simply be turned around to look at the ground. It is still necessary to keep a distance of at least 50cm or so between ground and antenna. However, the brightness temperature of the ground cannot be assumed to be 300K because it will depend on the material and moisture content of the ground.
The point of this test is to determine system parameters such as receiver noise and calibration experimentally.
