on compaq at gsfc i ran with -C to check array bounds for r1 global_ocean_90x40 it worked, no errors!!! what a team. however, when i increase number of time steps from 20 -> 200 it crashes in time averaging accumulation routine - anybody know of problems here? chris