That mankind is changing the climate is commonly accepted.

The claim is supported by increasing greenhouse gas (GHG) emissions from fossil-fuel burning, a global temperature increase from 1975-1998 and future temperature projections derived from climate models. The Kyoto protocol based on model projections aims to keep future temperature elevations below 2ÚC by severe reductions (decarbonisation) in GHG emissions, and fossil-fuelled electricity generation is considered a primary emitter. But the alternatives, so-called renewables, are very expensive and area-hungry.

Biofuels from cereals have actually increased world food prices; wood has higher emissions than coal. Nuclear is the best alternative but costly to initiate, although cheaper overall.

Paleoclimatology in contrast has shown climate and atmospheric GHG concentrations have changed radically in the past without human intervention. The contribution of natural variation to present climate change remains uncertain.

Some 2.5 billion people go to bed hungry each night and nearly nine million die each year from malnutrition. About 80% of mankind has replaced animal labour and fire with electricity but one third are still without hygienic cooking facilities, and smoke damages health.

It is estimated 60% of the world's primary energy in 2020 will be from fossil fuels. Cheap electricity is essential for the economic activity and development needed to lift people out of poverty, increase food production and boost life expectancy.

Given the enormous and crippling economic cost of decarbonising, worst for the poorest, the source of prediction about future climate temperature had better be accurate.

Current assessments from some 70 global climate models (GCMs) have been summarised by the Intergovernmental Panel on Climate Change. For a CO2 doubling, they give a temperature rise of 1.5ÚC to 4.5ÚC. Those so inclined choose the highest to headline and call for radical changes in our way of life. But other much lower and more optimistic scenarios are also indicated.

However the problem with all GCMs is that their short-term predictions have proved inaccurate, disagreeing with observation for the last 16 years and are therefore, in some or many respects, wrong. Current GCMs do disagree strongly with each other over the contributions of clouds and precipitation to climate temperature.

But these are fundamental to the greenhouse effect of water vapour, which provides the great majority of heat retained by the planet. A newer and simpler approach relies totally on satellite observations of incoming and outgoing heat and conservation of energy to measure the sensitivity of climate temperature to CO2 changes. These observations indicate that a doubling will raises the temperature by only 1.5ÚC.

At equilibrium, when emissions and sequestration eventually equal each other, it will only be 2ÚC. Observation is scientifically far superior to inaccurate models and much more believable.

It is against the reality of grinding poverty that Kyoto attitudes must be judged. Supporters greatly exaggerate fears and often claim a scientific consensus about climate change. If there is a consensus it's certainly not science and if science, it's not consensus. Not enough is known to make any prediction certain. Forecasting as a science is littered with gross failures.

Humanist manifestoes emphasise the reduction of poverty as a priority. Many religious organisations agree but do not realise belief in climate models contributes to the problem.

The problems of poverty and hunger are real whereas those underpinned only by climate models are possibilities and dependent on interpretation, not fact.

The Kyoto protocol exaggerates the precautionary principle and demands expensive energy sources including biofuels. The poorest parts of the world need cheap electricity to help lift them out of poverty. Given the uncertainties in climate predictions, Kyoto should be put on hold until errors are corrected and climate better understood. To do otherwise is surely immoral.