17 of 34
We conclude
?
MBH98 underestimates low-frequency
variability.
?
Moberg
does a reasonable job (maybe
fortuitously).
?
Borehole temperatures reconstructions
provide reasonable estimates, albeit
heavily smoothed.