Great Circle Distance

So various legacy web sources give the SEA to CDG airport distance as 8073 km (roughly 1605225 rods), though I cannot figure out where that number comes from, as it does not line up with any of the Earth radius estimates for kluging a great circle distance onto a not-a-sphere, or I've made a math error somewhere.

    (defstruct point
      (lat 0.0d0 :type long-float)
      (long 0.0d0 :type long-float))

    (defun deg2rad (d) (/ (* d pi) 180.0d0))

    (defun make-point-degrees (&key (latitude 0.0d0) (longitude 0.0d0))
      (make-point :lat (deg2rad latitude) :long (deg2rad longitude)))

    (defun central-angle (p1 p2)
      (let ((deltal (abs (- (point-long p2) (point-long p1)))))
        (acos (+ (* (sin (point-lat p1)) (sin (point-lat p2)))
                 (* (cos (point-lat p1)) (cos (point-lat p2))
                    (cos deltal))))))

    (defun distance (cangle radius) (* cangle radius))

    (defparameter SEA
      (make-point-degrees :latitude 47.448889d0 :longitude -122.309444d0))
    (defparameter CDG
      (make-point-degrees :latitude 49.009722d0 :longitude 2.547778d0))

    (dolist (earth-radius-estimate '(6335.439 6399.594 6371.009))
      (format t "~&~$~&"
              (distance (central-angle SEA CDG) earth-radius-estimate)))

The closest number (8085 km) is for the "spheroid at the poles" radius (6400 km) which is likely the most appropriate: if you're fleeing from America to Paris the trip would typically go closer to a pole than to the equator. Well, maybe not in a few hundred million years, but you get the idea.

https://en.wikipedia.org/wiki/Great-circle_distance

P.S. maybe make an offline copy of wikipedia before it goes in for some creepy corporate age verification thing, or the interface gets any worse?