extend the range up to eta=100,000, well above where the model will break down
regrid the photon distribution table on a more consistent, continuous 2D grid
We were starting to run simulations which reached eta > 10 so needed to update the tables to get the correct photon production rates and energy distributions. I regridded the chimin, h_sokolov, ksi_sokolov tables to run from eta = 10^-5 to eta=10^5 in 201 steps.
I also regridded chi such that it runs from chimin = 10^-12 eta^2 until chimax=eta/2 up until eta=1 and then from chimin = 10^-12 eta until chimax=eta/2 above that. The chi grid is now continuous and reflects how the average chi is proportional to eta^2 for low eta and proportional to eta for high eta.
I calculated the integrals in Matlab using the besselk functions. The important things were to have enough resolution on the integral (I used 100,001 points) and to extend it high enough in the variable y=4chi/(3eta[eta-2chi]) (I used 100*max[y]).
I've updated the tables for photon emission to:
We were starting to run simulations which reached eta > 10 so needed to update the tables to get the correct photon production rates and energy distributions. I regridded the chimin, h_sokolov, ksi_sokolov tables to run from eta = 10^-5 to eta=10^5 in 201 steps.
I also regridded chi such that it runs from chimin = 10^-12 eta^2 until chimax=eta/2 up until eta=1 and then from chimin = 10^-12 eta until chimax=eta/2 above that. The chi grid is now continuous and reflects how the average chi is proportional to eta^2 for low eta and proportional to eta for high eta.
I calculated the integrals in Matlab using the besselk functions. The important things were to have enough resolution on the integral (I used 100,001 points) and to extend it high enough in the variable y=4chi/(3eta[eta-2chi]) (I used 100*max[y]).