The 30m and the CloudSat

After several unsuccessful attempts the Earth exploration (active) satellite CloudSat is expected to be launched on 28 April 2006. The CloudSat operates in the frequency band 94 - 94.1 GHz that has been primary allocated for the Earth exploration (active) satellites according to the ITU footnote 5.562, sharing a secondary allocation for the radio astronomy service. The contiguous bands 92 - 94 GHz and 94.1 - 95 GHz continue primary allocated for radio astronomy service, although there is a potential risk for harmful interferences.

The CloudSat has a nearly circular orbit at 705 km altitude with a 99 min. period that repeats every 233 orbital revolutions. The orbital trajectories describe a reticular grid around the Earth that repeats every 16 days. The transmitted signal has the form of short pulses of 1.8 kW peak, with approx. 3.3 $\mu$s duration and 250 $\mu$s repetition rate. The antenna gain is 64 dBi giving a peak surface power flux density over all emitted frequencies of -31 dB(Wm$^{-2}$). The emission is to the nadir with a maximum deviation of $0 \fdg 1$.

The first effect to consider with respect to the CloudSat emission is the possible damage of the receivers due to the direct impact when the antenna points to the zenith and the satellite crosses just above. Considering the effective collecting area of the 30m telescope of 700 m$^2$ and the flux density of -31 dB(Wm$^{-2}$) the power concentrated into the mixer reaches 556 mW. This power is in the order of millions times more than the standard required LO power for normal operation and enough to permanently damage the junction.

We have protected the receivers against this situation by installing an automatic closing mechanism on the vertex window for antenna elevations higher than $88 \fdg 5$. In order to minimize a possible failure on the system, two redundant processes are in charge. The first process is implemented at the level of hardware by means of the PLC (programmable logic controller) of the servo system. When the antenna prelimit-up in elevation is reached the electrical power that keeps the vertex open is switched off, and the vertex is then closed. The second process, implemented by software, is permanently running in a VME machine, monitoring the antenna parameters: antenna elevation bigger than $88 \fdg 5$, elevation prelimit-up and working limit-up, and if any of these three conditions is reached the vertex is commanded to close. Any failure of the previous operation is notified to the operator by the acoustic alarm in the control desk alarm system.

Making pointing scans on Venus at 94.05 GHz we have measured the total power collected by the receiver with the vertex open and closed. We conclude that the attenuation introduced by the vertex closed is, at least, of 30 dB, enough to protect the receivers.

According to the transmit signal characteristics of the CloudSat (http://www.iucaf.org/CloudSat/CloudSat-TechDetailsV3.pdf) the satellite antenna gain at 1from the peak emission drops by more than 40 dB and the expected power reaching the receiver seems not dangerous. With the protection implemented, the vertex is closed within $1\fdg 5$ from the zenith, outside that cone the vertex can remain open since the level of power flux density from the CloudSat, even if the 30m is pointing to it, is not harmful for the receivers.

In the foreseen CloudSat ephemeris data (http://www.iucaf.org/CloudSat/CloudSat-Approaches-WebPage.htm) the closest approach estimate to the zenith of the 30m every 16 days cycle is very favourable, with no orbits or tracks in a 30 km radius from the zenith ($2\fdg 4$). The situation is less favourable at Plateau de Bure site with two orbits at less than 15 km in a similar time period. The final CloudSat tracks will be similar, but not identical.

The second effect to consider with the CloudSat is the harmful interference in the adjacent bands 92 - 94 GHz and 94.1 - 95 GHz. According to the ITU Recommendation 769 and CCIR 224-7 the harmful interference level for Radioastronomy Continuum Observations at 89 GHz with 2000 seconds of integration time is a power flux density of -125 dB(Wm$^{-2}$).

The signal transmitted by the CloudSat has a minimum attenuation outside the window 94 - 94.1 GHz of 50 dB. On the other side, far side lobes more than 10away from the peak have a minimum attenuation respect to the peak of 75 dB. Then, the maximum power flux density reaching the observatory when observing outside the band 94 - 94.1 GHz and farther than 10from the satellite will be $-31-50-75 = -156$ dB(Wm$^{-2}$), well below the threshold recommended by the ITU. In addition, if this maximum value is averaged over 2000 second (or 250 $\mu$s) the averaged power flux density is even 18.4 dB smaller or $-156 - 18.4 = -174.4$ dB(Wm$^{-2}$).

Nevertheless, if the observation is done in the window $94 \-- 94.1$ GHz or with the antenna pointing closer than 10from the satellite, the harmful threshold of -125 dB(Wm$^{-2}$) at the observatory could be exceeded. In the first case with $-31-75-18.4 = -124.4$ dB(Wm$^{-2}$) and in the second case with $-31-50-18.4 = -99.4$ dB(Wm$^{-2}$).

A good criteria to guarantee that the running observation has not harmful interference due to the CloudSat would be monitoring (and notifying in the control room) the short periods when the CloudSat is above the horizon, informing of the instantaneous azimuth and elevation of the satellite. If the observation is outside the band 94 - 94.1 GHz and the antenna is pointing farther than 10away from the satellite no interference must occur. On the other hand if this is not the case the situation must be analysed carefully or the observation be even rejected.

J. Peñalver and S. Navarro