search

icecube / pisa

Monte Carlo-based data analysis
http://icecube.github.io/pisa/
Apache License 2.0
18 stars 47 forks source link

Trying to reproduce pisa results #724

Open rruizdeaustri opened 1 year ago

rruizdeaustri commented 1 year ago

Hi,

I have tried to reproduce pisa results with my own code using vanilla prob3++ and just including detector systematics. I attach the 90% CL contours.

After inspecting the spectra for a few benchmarks I see that it agrees quite well (order 1%) excepting in low energy bins with large zenit angles in which can differ up 5%. If I use squids (matter effects for neutrinos passing through the core of Earth) instead of prob3++ the errors in those bins are reduced. So my bet is that a non proper propagation of neutrinos crossing the center of Earth is responsable of the shift in the contours. Does this make sense ?

On the other hand, what I don't understand is why pisa results differ with the ones in "Measurement of Atmospheric Neutrino Oscillations at 6-56 GeV with IceCube DeepCore" paper. Specially in the position of the bf in delta m^2_32.

Systematics.pdf

Thanks Roberto

philippeller commented 1 year ago

Hi, are you using the public dataset? If yes, then please note that the systematics treatment is a simplified version of what is used in the paper you cite, and the dataset is also slightly newer. So we do not expect 100% identical results. Your plot suggests that the results are very well compatible i think.

ggarg07 commented 1 month ago

so, can we mark this issue as resolved and close it?