Your analysis of the long-term investment strategy of the US National Science Foundation (NSF) fails to recognize the extent to which the agency's academic-research fleet is one of the most vital elements of its oceanographic infrastructure (Nature 501, 461; 2013).

The increasing complexity of large-scale global programmes on climate and ocean science, such as CLIVAR, GEOTRACES and GeoPRISMS, calls for modern vessels that can host large research groups and that use sophisticated on-board sampling and processing equipment. These vessels are needed for operating state-of-the-art observation systems, interpreting satellite data, and launching and recovering the gliders and autonomous underwater and aerial vehicles used by the ocean research community.

The replacement strategy that you criticize is in fact insufficient to replace a research fleet, the age and escalating costs of which mean that it is barely capable of meeting projected scientific requirements — even if all three planned regional research vessels are constructed over the coming decade.

It is difficult to provide infrastructure, which takes many years to design and build, with budgets that are unpredictable and constantly changing. Core science budgets should be kept separate from funding for the operation and maintenance of crucial research infrastructure, and not just from its construction. These are complementary rather than competing projects, and both need greater investment.

Properly informed, scientifically based policies and strategies will save the public much more money than would be saved by cutting costs on essential scientific infrastructure.