Update on SERENDIP/Nebula:
- Using the 2017 data from SERENDIP/Arecibo/ALFA, I scored 100K pixels on the Atlas cluster, and copied the results to our servers.
- I modified all the web features (multiplet and pixel lists, waterfall plots, custom RFI removal, and scoring details) to work for SERENDIP as well as SETI@home.
The results are visible here. You'll notice that the only signal type is spikes (that's all that SERENDIP detects), and that there are no "adjusted" scores (there's no need to normalize across signal types).
There are still various glitches; e.g. the scores are astronomically high because there was a normalization that I didn't know about. However, the preliminary results are already very useful. You can see that the top-scoring multiplets consist mostly of two types of hits:
- Ones near, but not actually in, a dense strip of frequency-stable RFI. We're trying to decide if these should be considered RFI also.
- Ones in bands of drifting RFI that are only partly removed by the drifting algorithm. We may need to tweak the algorithm.
This is a big step for both SERENDIP and Nebula. Having generalized Nebula for this data source, it will be easy to extend it to other data sources, such as SERENDIP on Parkes and GBT. A big thanks to Bruce Allen and AEI in Hannover for letting us use Atlas.