Just as the National Broadband Network has been subjected to rigorous review and reframing, we should expect nothing less of the direct action scheme.

The debate thus far is not encouraging. The Climate Institute, an independent body previously supportive of Rudd-Gillard government initiatives, has weighed in to the debate with a submission to the Senate environment and communications references committee.

The institute argues that Australia's targets are not enough and, even if matched by the rest of the world, could bring about global warming disasters including "droughts in southern Australia occurring up to five times more often than present, and the destruction of the Great Barrier Reef". It does not attempt to review the physical science on which such projections are based.

Maurice Newman and Mohamed Nasheed have added to the debate on this page this week, with neither considering the underlying science and both adding more heat than light: Newman speaks of anthropogenic global warming as "scientific delusion" and Nasheed counters with sceptical views as being "antediluvian denialism".

I identify five segments of science - all detailed in peer-reviewed journals in the past three years - which demand scrutiny before we believe current global warming projections.

First, climate sensitivity is generally defined as the change in global temperature produced by a doubling of CO2 in the atmosphere. A range of studies across the past five years indicates this may be below, or significantly below, present values quoted by the UN's Intergovernmental Panel on Climate Change, in which case published modelling projections of future global warming and sea-level rise become overstated.

Second, the disconnect between CO2 increase and global temperature change since 1900 is especially evident in the global warming hiatus of the past 17 years. The mechanisms for this hiatus are not adequately described by consensus science, but there is increasing evidence to suggest natural cyclic change plays a major role in this dichotomy between projections from climate modelling based on anthropogenic global warming theory, and systematic measurement using terrestrial and satellite observation platforms.

Third, cyclic variations in global sea level suggest natural cycles of about 60 and 30 years in length. Such cycles, which are deserving of considerable further study, suggest a significant fraction of the observed rate of sea-level rise of past decades may be attributable to the upswing of natural cycles. The consequence, if proven, on projections of future sea-level rise and associated planning and land-use policy is large.

Fourth, natural cycles in climate change are increasingly evident from precise studies of temperature records imprinted in cave deposits, ice cores, corals and deep-sea sediments. These provide mounting evidence that current global warming is not abnormal in a historical context, and variations are subject to a range of natural cyclic phenomena with periods ranging from about 60 years to millennia.

Finally, causative mechanisms for natural cycles in climate change are an essential complement to observational data showing natural cycles in climate change. Mechanisms involving highly complex interactions of solar physics, magnetic fields and cosmic rays are on the cusp of delivering insights into possible mechanisms.

The issue of global sea-level rise has made news in the past week because of some local councils in NSW placing restrictions on frontage properties because of perceived flood and erosion risk. Such restrictions may have major impacts on land values, affecting individual owners, and it is worth considering some of the unknowns affecting such decisions.

A useful baseline number for global sea-level rise is a mean of 1.7mm a year for the period 1900-2010. Faster rates of rise in the order of 3.2 mm a year are recorded for 1993-2010, and 1930-1950. Projections of sea-level rise of 90cm by the year 2100 - used to justify planning restrictions on coastal property - require the rate of rise to increase to 15mm a year.

While the evidence for a small underlying trend of an increasing sea level is compelling, faster rates of rise and projections based on these need to be reconsidered in the light of evidence of multi-decadal cyclic variations in global sea level.

By my reading of the relevant literature (such as by Don Chambers of the college of marine science, University of South Florida, and colleagues, and by Ben Hamlington and colleagues at the University of Colorado, Boulder), the cycles arguably account for 30 per cent to 50 per cent of the observed sea-level change of the past 20 years, which leaves a residual rate of global sea-level rise near to the rate observed in the past 200 years.

The past 20 years appear to lie dominantly on the upswing of the natural cycles, thus raising the question of whether an apparent increasing rate of sea-level change in past decades is a new trend or whether it could be another facet of natural cyclic change of earth systems.

More alarmingly, these cyclic variations are not yet built into any published sea-level projections.

While the work of Chambers is mentioned in the IPCC's Fifth Assessment Report, it is not yet incorporated into any of the published modelling scenarios for future sea-level rise. Hamlington's work is more recent than the report.

Local councils that rely on the older IPCC Fourth Assessment Report not only risk injustice to their ratepayers, they are out of date by five years on scientific grounds.

The Abbott government is committed to spending $5 billion annually on its direct action emissions reduction program.

The Senate inquiry would do well to recommend some thousandths of this sum be spent re-examining which projections are credible, which natural changes require mitigation of effect rather than cause, and what cost-benefit parameters apply to programs targeting residual anthropogenically related climate change.