It is commonly assumed that biomass fuel cycles based on renewable harvesting of wood or agricultural wastes are greenhouse-gas (GHG) neutral because the combusted carbon in the form of CO 2 is soon taken up by regrowing vegetation. Thus, the two fifths or more of the world's households relying on such fuels are generally not thought to play a significant role in GHG emissions, except where the wood or other biomass they use is not harvested renewably. This review examines this assumption using an emissions database of CO 2 , CO, CH 4 , NMHC, N 2 O, and total suspended particulate emissions from a range of household stoves in common use in India using six biomass fuels, kerosene, liquefied petroleum gas, and biogas. Because typical biomass stoves are thermally inefficient and divert substantial fuel carbon to products of incomplete combustion, their global warming commitment (GWC) per meal is high. Depending on time horizons and which GHGs are measured, the GWC of a meal cooked on a biomass stove can actually exceed that of the fossil fuels, even if based on renewably harvested fuel. Biogas, being based on a renewable fuel and, because it is a gas, being combusted with high efficiency in simple devices, has by far the lowest GWC emitted at the stove per meal and is indicative of the advantage that upgraded fuels made from biomass have in moving toward sustainable development goals. There are a number of policy implications of this work, including revelation of a range of win-win opportunities for international investment in rural energy development that would achieve cost-effective GHG reduction as well as substantial local benefits.

Abstract. Diurnal variation of number size distribution (particle size 3-800 nm) and modal parameters (geometric standard deviation, geometric mean diameter and modal aerosol particle concentration) in a highly polluted urban environment was investigated during October and November 2002 in New Delhi, India. Continuous monitoring for more than two weeks with the time resolution of 10 min was conducted using a Differential Mobility Particle Sizer (twin DMPS). The results indicated clear increase in Aitken mode (25-100 nm) particles during traffic peak hours, but towards the evenings there were more Aitken mode particles compared to the mornings. Also high concentrations of accumulation mode particles (>100 nm) were detected in the evenings only. In the evenings, biomass/refuse burning and cooking are possible sources beside the traffic. We have also shown that nucleation events are possible in this kind of atmosphere even though as clear nucleation events as observed in rural sites could not be detected. The formation rate of 3 nm particles (J3) of the observed events varied from 3.3 to 13.9 cm −3 s −1 and the growth rate varied from 11.6 to 18.1 nmh −1 showing rapid growth and high formation rate, which seems to be typical in urban areas.