The value of observations - Part I: data denial experiments for the Atlantic and the Pacific