Thanks Cyclops. Please make it appear on the WCG mainpage also, and not "hidden" under "Show all news →"

Hm, the article says Prof. "Nick van de Geisen", but in reality the name of the Prof. is:

Nick van de Giesen

Mike - nearly spot on!

First up, on a quick search I found the following from Keith Uplinger, in response to a question in an earlier post (both quoted below...)

And I've had a look at some current namelist.input files which have

  dx=9000,3000,1000
  dy=9000,3000,1000

which confirms the above if the scaling unit is 1 (metre).

Interestingly, there is a [sort of] 50 factor involved -- the notional grid dimension (West->East or South->North) is specified via

  e_we=52,52,52
  e_sn=52,52,52

There is a requirement that those values are one more than a multiple of the parent_grid_ratio for each nest (and that's 3 for the inner nests...) Then, the locations of the nested grids are specified relative to their own parents, as

  i_parent-start=1,18,18,
  j_parent-start=1,18,18,

which, as can be seen, steps in one-third of the way (17 out of 51) and will be a third the actual width (but still the same computational grid size)...

So it looks like someone's notes for that interview were confused or in error :-)

By the way, if there is no overlap between grid squares to allow for "edge conditions", that would give an area of around 2.9 million square kilometres, which would represent about 12% of the broadest categorization of Sub-Saharan Africa (as per Wikipedia...)

I would expect that a lot of that 24 million square kilometres is of less interest (true desert, entirely mountainous regions, massive urban complexes?), but it would be interesting to know what areas are covered...

Cheers - Al.