Climate change news always seems grim. A week before a major climate change conference, it seems grimmer. The last month was full of news about melting glaciers, shrinking ice in the poles, increasing rate of carbon dioxide emissions, possibility of large-scale methane emissions and so on. Buried among these were a few bits that should have made people in India sit up and take notice. It said something that has been said before but in a different way: the Indian monsoon is in trouble.

A few weeks ago, scientists at the Potsdam Institute of Climate Impact Research in the Netherlands published the results of a study that said that the Indian monsoon could switch gears in no time, from a high-precipitation mode to long dry periods. The study comes a few months after another at Purdue University, the US, which said the South Asian monsoon will onset later and later, and cause long dry periods and suppression in certain parts of the country. Other studies in Indian institutions have also indicated that the monsoon will change, increasing in intensity in places where they cause floods now and weakening in places that are drought-prone. “The study has major implications for people in north India and Pakistan,” says Moetasim Ashfaq, lead author of the Purdue study and a post doctoral researcher at Stanford University.

The impact on India is huge, although different studies tell different stories, and it is not all disaster. The Potsdam study is more disturbing because it talks about an abrupt transition. The Purdue study is also bad news for agriculture, but it is consistent with the observation that overall rainfall will increase over the subcontinent. But the increase may not be during the monsoon season, which means agriculture could be seriously affected. Variation in rainfall over the subcontinent is also bad news, because both floods and droughts will increase.

The Potsdam group observed that the Indian monsoon had gone through several abrupt transitions in the last 11,000 years, and decided to investigate these changes. As is known, the monsoon is fed by temperature differential between the land and the sea. Hot air rises above the land, and moist air from the sea moves in towards the land. But as the air drops this moisture over the land as rain, it also releases latent heat over the land. This causes a further temperature differential, which draws in more moist air, and the cycle intensifies. This self-amplification, vital to a good monsoon, is very vulnerable to external influences, the Potsdam researchers found.

A critical amount of sunlight is necessary for the monsoon to start. If this level falls below the threshold due to pollution, the monsoon may not start. But they also found that there are two states even when it starts, and the changing climate can push the monsoon from one state to the other.

For example, if the monsoon current weakens for any reason, the self-amplification can work the other way: a weak current has less moisture, and thus less latent heat is released, which means there is lesser temperature differential. Aerosol emissions could also reduce monsoon stability. “Oscillations between months with heavy rainfall and those with weak rainfall could strain people’s ability to adapt,” says Jacob Schewe, co-author of the study, in a release.

However, the effect of aerosols may not be significant. Says Krishna Kumar, a scientist at the Indian Institute of Tropical Meteorology Pune, “Though it may seem theoretically possible, the likelihood of such tipping through aerosols is remote. That’s because the increasing levels of greenhouse gases compensate the cooling effects of aerosols over the region.”

The Purdue study used different models and methods, and found something equally interesting. The major input for it was a model with a high resolution. Most models use general circulation models that take into account global patterns in circulation but not local details like topography. The Purdue model supposedly used models with the highest amount of detail so far. It showed an overall weakening of the monsoon and a delay of five to 15 days by the end of the century. This model predicts the past accurately, and so could be accurate for the future as well. “We need to investigate this further to understand some of the new observations,” says Noah Diffenbaugh, assistant professor at the department of environmental earth system science, Stanford University.

India has had a weak monsoon this year, but many parts have had good rains after its withdrawal. This is consistent with the observation that overall rainfall will increase even as the monsoon weakens.

What does it mean for the Indian farmer? A change in the nature of agriculture, perhaps. The changing rainfall may do some good as well, by bringing rain during what is usually a dry period for India. Whatever it is, it is time to prepare to adapt.