Doubt in FSLMotionOutliers

Hello everyone, I am using the FSL function- FSLMotionOutliers -
https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FSLMotionOutliers

I am passing the motion corrected file to the function.

motionOutliers = Node(MotionOutliers(no_motion_correction=True, out_metric_plot = 'refrms_plot.png',
                                     out_metric_values='refrms_raw.txt'),name='motionOutliers')

I am getting a peak at the 99th volume (Middle) in all the subjects consistently.

refrms_plot

refrms: RMS intensity difference of volume N to the reference volume

Q1. What is the reference volume taken by default?
Q2. Is the peak at 99th volume in all the subjects correct? Why are all the middle volumes most errornous?

We saw similar behavior when looking at DVARS after MCFLIRT motion correction in MRIQC: https://github.com/poldracklab/mriqc/issues/221