Select Page

Indian scientists found that only nine of 42 global models could capture the declining rainfall trend in South Asia since 1951

Indian scientists caution that inconsistencies between global and regional models on the impact of climate change on the Indian monsoon need to be resolved. (Image by Bobinson KB)

Indian scientists caution that inconsistencies between global and regional models on the impact of climate change on the Indian monsoon need to be resolved. (Image by Bobinson KB)

Meteorological models to simulate the impact of climate change on rainfall at a more focused, regional scale in India are throwing up results contrary to what global models project, which throws planning for adaptation measures and for agriculture into a quandary.

At a recent conference organised by the Tata Institute of Social Sciences in Mumbai, Indian scientists cautioned that inconsistencies between global and regional models on the impact of climate change on the Indian monsoon need to be resolved. “We have to be careful in planning adaptation measures,” Subimal Ghosh, assistant professor at the department of civil engineering at Indian Institute of Technology (IIT), Mumbai, said.

When Ghosh’s team analysed the global circulation models (GCMs), they found that out of 42 GCMs, only nine could capture the declining trend in Indian rainfall observed on ground since 1951. “This is a limitation” of global models, he said.

Global circulation models cannot model accurately for regional-level impacts, Ghosh said. “For impact assessments, for example, water domains, water resource management, hydraulic structure design, economic risks and vulnerability assessments of extreme events, we need regional-scale hydro-meteorological (water and weather) variables.”

Ghosh’s team has developed a regional model that showed a decreasing trend in rainfall after 1950, which matched ground observations and records.

“For regional modelling, we use statistical methods. It (the regional model) not only goes for fine resolution but captures the impacts of orography (physical geography) on possible changes,” Ghosh told The parameters used for the two sets of models differ too, he added.

Statistical downscaling techniques involve developing a numerical or quantitative relationship between large-scale atmospheric variables and local variables. The IIT team captured statistical properties of observed data in India and co-related it with rain gauge stations and physical topography.

“As climate models are still not good for Indian monsoon, first development of correct models and their critical evaluation is required before taking any adaptation measure,” Ghosh said.

Bhupendra Nath Goswami, former director at the Indian Institute of Tropical Meteorology (IITM), Pune, concurred that studies from India showed a decline in rainfall over India since 1941.

Also, extreme events (more than 10-15 cm continuous rainfall) have been increasing in India in recent years, while moderate events with 2-7 cm continuous rainfall have been decreasing. The intensity of extreme events too has been increasing. “This has implications for Indian agriculture,” Goswami told the conference.

The intense, extreme events occur over a small area, and are of shorter duration. “Increasing trend of these events means increased run-offs,” he pointed out.

In contrast, “weak and moderate rain events are responsible for groundwater recharge. Decreasing trend of these events means decreased ground water recharge capacity. When combined with the fact that groundwater is already decreasing in most places, the two trends mean a significant increase in susceptibility to drought,” Goswami cautioned.

There are two kinds of discrepancies. One is between global and regional models. The other is between models and actual ground observations. Scientists said both kinds of discrepancies arise due to intrinsic problems with modelling precipitation (rainfall and snowfall).

There is a problem of scale – how effectively is a small-scale event, for example, rainfall in a specific Indian state or district, represented in larger-scale models. Govindswamy Bala, associate professor at the Centre for Atmospheric and Oceanic Sciences (CAOS), Indian Institute of Science (IISc), Bangalore, said, “This problem will remain with us for some time.”

Then there is a lack of understanding how cloud droplets and ice crystals form, how they grow and precipitate as rain – all of which play a key role in clouds, weather and climate models. “The (cloud) microphysics is poorly understood,” said Goswami.

Indian scientists admit these problems limit their ability to forecast monsoon performance over South Asia, over the next season, or more long term.

At the conference, scientists stressed the need for more accurate projections of the impacts of climate change on the monsoon over India, given the enormous implications for poor people’s livelihoods and the country’s economy.

Well over half the population of South Asia is dependent on agriculture for a livelihood, and around two-thirds of the agriculture is rain-fed.

Climate models indicate that extreme temperature episodes are shifting more towards early in the summer. The wheat crop, for example, may experience more heat in future, says a 2014 study led by T. Jayaraman of TISS.

“There is consensus that Indian monsoon may get 5-10% more variability in the future,” says the study, which was presented at the Mumbai conference. “This will have serious implications on agricultural practices, especially in regions where the monsoon is already low and have high variability component.”

Accurate projections on the impact on Indian agriculture will require detailed analysis of various agriculture-related issues such as cropping patterns; asset and land use based information; information on agriculture extension facilities’ inputs such as seed, fertiliser and labour; various agricultural operations during cultivation; agriculture credit; market access and market rates.

An example of how extreme events cause havoc in village economies is a TISS study in village Melanjippattu in the southern state of Tamil Nadu, hit by cyclone Thane on December 29-30, 2011.

The cyclone hit the state during the harvest, leading to widespread damage to standing crops and killing 53 people.  It damaged about 55% of the state’s cultivated area; around 73% of the paddy growing area; and 92.5% of the groundnut crop.

The study found that all households made net losses in cultivation in the disaster year; and the loss per acre was higher for smaller farm sizes, though the total volume of losses varied directly with the size of the farm.

The relative losses (loss per acre or loss as a percentage of normal year incomes) are higher for smaller farmers, and the smaller farmers are more vulnerable to the impact of income shocks caused by disasters.

Agricultural workers could not find employment during harvesting time and in the subsequent seasons when crop production declined due to disruption of power supply that affected irrigation.

Given such situations, it is crucial to know whether the decreasing rainfall trend observed in recent decades is temporary and part of multi-decadal oscillations – rise and declines over decades – Goswami said. Knowledge of the trends, in turn, is crucial, for planning adaptation measures.

Share This