Almost 20% of India's land surface has toxic levels of arsenic in its groundwater, reveals a recent IIT Kharagpur study supported by artificial intelligence (AI)-based prediction modelling. Researchers found a high presence of arsenic in areas along the Indus-Ganga-Brahmaputra river basin and in pockets in Peninsular India. This exposes more than 250 million people across the country to the poisonous element. 

The study, published in the journal Science of the Total Environment, has shed light on the ground situation and the researchers convey that there is a higher population and a larger spread of the arsenic zone than previously estimated by various government and non-governmental organisations. Researchers of the study call for a more rigorous and thorough sampling of arsenic levels across the country. 

Among the states, Punjab, Bihar and West Bengal have the largest area of elevated groundwater arsenic zones at 92%, 70% and 69% respectively. While states such as Assam, Haryana, Uttar Pradesh, and Gujarat have 48%, 43%, 28% and 24% of arsenic content in their groundwater level, Madhya Pradesh, Karnataka, Odisha, Maharashtra and parts of Jammu and Kashmir have sporadic occurrences of arsenic presence in single digits at 9%, 8%, 4%, 1% each, respectively.

The study is the first-of-its-kind, based on the field-sampled arsenic distribution patterns along with an arsenic prediction model using advanced AI across India. The team used advanced AI to model the presence of arsenic in groundwater above the permissible limit as per India, 10 microgrammes per litre (μg/L), across India. The study focused on various geologic, hydrogeologic and anthropogenic factors that have been known to control the groundwater arsenic distribution in the aquifers.

"A total of more than 250 million people are estimated to be exposed to high arsenic in India...Our study indicates a strong influence of irrigational abstraction (of groundwater) and regional geology on arsenic distribution patterns within India," Abhijit Mukherjee, Associate Professor at the Indian Institute of Technology Kharagpur, West Bengal and the lead author of the study told PTI. 

The researchers have used Random Forest, an advanced AI model, which has been known to be extremely efficient in predicting groundwater arsenic in one of our previous studies. The researchers studied 27 lakh field measurements of the government's Jal Jeevan Mission and undertook the research to help the mission be able to provide clean drinking water to households.

Arsenic is an extremely toxic element in an inorganic form. According to the World Health Organisation (WHO), long-term exposure to arsenic, present in drinking water and food can cause cancer and skin conditions along with other disorders. 

 "Our AI model predicts pervasive arsenic contamination in major parts of the Himalayan mega-river Indus-Ganges-Brahmaputra basins, however, it also occurs in several more-localised pockets, mostly related to ancient tectonic zones, igneous provinces, aquifers in the modern delta and known sulphide mineralised mining regions," said Mukherjee.

The scientists hope that the study is helpful to policymakers and administrators in identifying safe drinking water sources in arsenic affected areas of India.