Why is Rosetta not going to hover or descend multiple times before crashing into 67P/Churyumov–Gerasimenko?

Question

Currently Rosetta is on a free fall collision course with 67P/Churyumov–Gerasimenko. If I understand correctly the probe will send back data and pictures during the descent, and once it has crashed onto the comet no communication will be possible anymore.

This raises the question, why not use some of the (presumably still available) fuel to have the space probe hover close to the comet so that there is enough time for measurements taken very close to the surface to be transmitted back? Or alternatively, if hovering is too complicated a maneuver while taking measurements, why not have multiple descents that get iteratively closer to the surface?

The reason why I am asking is that the impact speed that I found mentioned was 1m/s, and the current data transmission rate is of the order of dial-up speed [45kbps], which doesn't allow for much data to be sent back in the very last stages closest to the surface, which presumably is a region of great interest.

As far as the camera is concerned I understand that it can't focus on anything closer than 1km, but presumably it could be pointed sideways at surface features during the decent? And even if not, the other instruments might generate interesting data very close to the surface.

So as far as I can see there might be two answers to this:

1. The above maneuvers cannot be performed or
2. The data close to the comet wouldn't actually be all that interesting [e.g. due to the instruments not being built/optimized for that region].

Answer 1

This question has been asked several times in different places and I would think, that it was also carefully considered by ESA. The pros of such a strategy seems obvious. The probe will be abandoned anyway, so there seems to be not much of a risk.

Hovering is not an option, the probe would not be able to control and maintain this state. If there is position/attitude uncertainty (after a thruster burn or after passing an inhomogeneous gravity field), it takes a relatively long time to assert the state vector again. This is by no means possible in near real time.

A descent with an abort maneuver was done before by Rosetta, when it set Philae on its impact trajectory. Technically speaking, the descent is an orbit that intersects the comets surface. This would enable the probe to reach perspectives that are not available with flybys. The abort is time critical and probably needs to be done without prior state determination, while a flyby gives you much more time to reassert the state vector. ESAs flight dynamics team has some first hand experience with close flybys and the pointing errors that result from going so close. The famous lander image f.e. was planned to have the suspected landing spot "Abydos" centered. In the end, they were lucky to have it covered at all.

How is a descent without abort different from that and why should it be better? The scientific phase could last a little bit longer, but at the same time, you are less confident to capture anything of interest because of pointing error. The time to transfer the data will be hard limited, in fact data from the nearest approach will never reach us for sure. When you look at this with exploration on your mind, it is certainly worse.

ESA also wants to make history with a success story. An ambitious maneuver may leave a strange taste if it does not work out as expected. And finally, there are also expenses that need to be planned. I think, the scientific return from the finale will not add too much to the overall achievements. It is just an emotional thing.

Paolo Ferri on ESA-TV: "We wanted a clean end."