Chinese scientists turn to artificial intelligence as potential 1,000km seam of rare earth minerals found in Himalayas
- Machine which has accuracy of more than 90 per cent could help locate deposits along the Tibetan border, scientists say
- The discovery also has strategic and environmental implications in an area where there is an ongoing border dispute with India
in Beijing Updated: 1:27pm, 21 Jun, 2023
Locating and extracting minerals in the Himalayas will prove a major challenge. Photo: AFP
Chinese geologists recently discovered a huge potential reserve of rare earth minerals in the Himalayas that could significantly strengthen
.
But the belt of minerals is believed to be more than 1,000km (600 miles) long and locating deposits in such a vast, remote area could take years, if not decades.
A further headache is caused by the location – along the southern border of Tibet, where China has a long-standing territorial dispute with India.
As a result, Chinese government geologists have said that the quicker a country can pinpoint the deposits, the more “strategic advantages” it will have.
One possible solution is to use artificial intelligence. Since 2020, with financial support from the central government, a research team has been building AI that can automatically process nearly all raw data collected by satellites and other means to locate the rare earth deposit on the Tibetan plateau.
The scientists from the State Key Laboratory of Geological Processes and Mineral Resources in the China University of Geosciences in Wuhan said the machine has achieved an accuracy rate of 96 per cent.
“China’s demand for bulk mineral resources such as iron, copper, aluminium, coal, and cement that support industrialisation and urbanisation is expected to sharply decline in the next 15 to 20 years. The focus of mining will mainly shift to rare earths,” professor Zuo Renguang, the project’s lead scientist, wrote in a peer-reviewed paper published in Chinese-language journal Earth Science Frontiers last week.
“Rare earth metals are irreplaceable in emerging industries such as new materials, new energy, defence and military technology, and information technology, making them a key strategic mineral resource in global competition.”
What the AI looks for is a unique form of granite that has a lighter-than-usual tone. This may contain rare earths such as niobium and tantalum, which are highly sought after for hi-tech products, but also significant amounts of lithium, which is vital for making electric cars.
The Chinese geologists have found such granite all over the place in the Himalayas, including around Mount Everest, but until recently it was not believed to contain any mineable materials.
But about 10 years ago, Chinese geologists accidentally discovered an usually rich presence of rare earth and lithium in some rock samples collected from Tibet, and started to question the received wisdom.
China currently has a major rare earth production base in Inner Mongolia, and one scattered further south in provinces such as Guangdong, Jiangxi and Sichuan.
But scientists now believe the rare earth reserve in the Himalayas could be equal to, if not bigger than these and may even help China re-establish its position in the global market.
While China previously held a dominant position, with around 43 per cent of global reserves in the 1980s and 1990s, its share had declined to around 36.7 per cent by 2021, according to industrial estimates.
Meanwhile, rare earth resources outside China have seen significant growth, more than doubling from 40 million tons to 98 million tons.
When Zuo’s team built the AI more than two years ago, it was trained using a limited data set, such as satellite images, to identify the light-coloured granite.
Initially it only had an accuracy rate of about 60 per cent, but the researchers gradually broadened its knowledge base by increasing the accuracy of their algorithms.
The additional data sets that the team fed into the machine included those related to the chemical composition of rocks and minerals, their magnetic or electrical properties, spectral data collected by aircraft and geological maps of the Tibetan plateau.
Finding patterns in different data sets can pose a challenge to AI but the researchers taught the machine several techniques that can be used to overcome these issues, such as data normalisation, feature selection and data fusion.
The researchers noticed a rapid self-refinement of the AI model, which achieved an accuracy rate of over 90 per cent within months.
Although this suggests that the AI could be used immediately in the field, Zuo and his colleagues still have to overcome a further problem. They wrote in the paper that the way the machine chose locations “cannot be explained” and they did not feel comfortable trusting its decisions until they can understand the rationale behind them.
Mineral resources within the Himalayan rare earth belt have not only economic value but also strategic implications due to their potential impact on regional dynamics and resource competition, according to a study by researchers with the China Geological Survey last year.
“The Himalayan mineralisation belt is located in the southern part of China’s Tibet region and extends into countries such as India, Nepal, and Bhutan. Therefore, the mineral resources within this belt not only have economic significance but also, to some extent, possess strategic importance,” said the government study.
The study did not further discuss the strategic importance of the discoveries, but the extraction and processing of rare earth and lithium minerals requires the establishment of infrastructure such as roads and power supply.
The development of rare earth and lithium resources could also contribute to the economic growth, which in turn could increase the population of the area.
But a “rare earth rush” could increase the risk of geopolitical conflict – particularly with India – due to territorial disputes and environmental concerns, said a Beijing-based researcher who requested not to be named due to the sensitivity of the issue.
China has also promised to protect the environment in Tibet, but mining would have a significant ecological impact, such as putting pressure on already limited water resources. The environmental scientists also warned that it will be difficult to enforce proper waste management in such a remote area.