Indian Monsoon: Contribution to the Stern Review Andrew Challinor, Julia Slingo, Andrew Turner and Tim Wheeler University of Reading Introduction: India’s economy and societal infrastructures are finely tuned to the remarkable stability of the monsoon 1 , with the consequence that vulnerability to small changes in monsoon rainfall is very high. In 2002 the monsoon rains failed during July, resulting in a seasonal rainfall deficit of 19% and causing profound loss of agricultural production with a drop of over 3% in India’s GDP. Neither the prolonged break in the monsoon, nor the seasonal deficit was predicted. In the future, the pressures of an increasing population will bring additional stresses on society and the environment, with implications for water resources, health and food security. Consequently, climate change and the potential for the monsoon to become more volatile have major implications for India, itself, and for economies, worldwide. Current levels of skill: Current capabilities, world-wide, in predicting the behaviour of the monsoon within a season, for a season ahead and for the coming century are very limited. This is due to an incomplete understanding of monsoon processes and the poor performance of current climate models in this challenging and critical region where the ocean, atmosphere, land surface and mountains all interact. Yet we know that biases in the mean simulation of the monsoon compromise our ability to represent year-to-year variations and their links with El Nino (Turner et al. 2005), as well as introducing additional uncertainties in projections of monsoon behaviour in the 21 st century under global warming (Turner et al. 2006a). Future changes in mean monsoon: The general consensus amongst climate models is that the mean summer rainfall for All India will increase slightly, by about 10% by the end of the century, largely because of the warmer Indian Ocean and the fact that warmer air can hold more water. This increase in rainfall will not necessarily be accompanied by stronger monsoon winds. There are, however, likely to be much larger regional variations across India, with indications that the northern states will see much of that increase, although these changes are considerably more uncertain. As part of the more intense hydrological cycle, some models indicate that the intensity of heavy rainfall events may increase (the Mumbai floods of 2005 may be an example) whilst the number of rainy days may decrease. This suggests changes in the temporal characteristics of the water cycle which could have profound effects on agriculture and management of water resources. Temperatures will increase for all months. Consequently, during the dry pre-monsoon months of April and May, the incidence of extreme heat is likely to increase, leading to greater mortality (DEFRA, 2005). Higher mean temperatures during the wet season will also have implications for the viability of some crops. Meltwater from Himalayan glaciers and snowfields currently supplies up to 85% of the dry season 1 Over the past 100 years the standard deviation of the seasonal mean monsoon rains has been close to +/- 10% for All India. More regionally however the interannual variability tends to be larger particularly in the more arid parts of NW India where the standard deviation approaches +/- 30% (see Pant and Rupa Kumar 1997).