Dear Pengwang,
are you sure this is a spectral problem? The ARTS LUT does not do any spectral
interpolation, only interpolation in pressure, temperature, and water vapor
concentration. It could be these other interpolations that lead to negative
values, if the real absorption is close to zero.
Thanks, Oliver. I tried your recipe and it works fine to me.
Another small thing (or bug) is that arts LUT function sometimes generates
small negative values for absorption coefficient, when the spectral regions is
slightly outside an absorption band. If the developers are interested in fixing