AI Systems Predict Dust Storms with Higher Accuracy in Asia

Spring is considered a beautiful season world round, but many Asian countries face the problem of dust storms in spring every year. In March, some areas of China faced the same problem when the people living there were told to not to go out as fast winds will be blowing, reaching 100 kilometers per hour.

First, Inner Mongolia in China got affected by the dust storm, and then it was Beijing where a warning was issued to shut the windows and stay inside as dust reached the city after storming through the plains of China’s Inner Mongolia and some areas of Mongolia. In many cases, visibility got reduced to less than 90 meters during the storms as the skies turned yellow and gloomy.

Chinese meteorologists and scientists have developed a few forecasting systems over the years, starting back in the 1990s. But they are advancing their efforts to have a better idea and more accurate information on where the dust gets picked at first, in how much quantity, and how it changes in route as the storm progresses from one area to another. The systems they rely on are not without their own errors, as all systems are.

A dust storm above Arabian sea. Source: NASA.

Asian researchers and scientists have been utilizing climate modeling and now artificial intelligence to get better estimates and predictions for this phenomenon that they have to face every year. There are economic consequences to these climatic conditions, and with better predictions, millions of dollars could be saved every year. As it is estimated that in only one quarter, the losses can reach up to more than 4 million dollars, including damages to the infrastructure, farms, and housing units.

Causes, sources, and consequences

The Sahara in Africa, the world’s largest desert, is also the largest source of sand for these storms. While in Asia, the Gobi Desert is one of the largest sources, which lies partly in Mongolia and partly in China. Now scientists are working on solutions to predict the sand storms that are based on artificial intelligence. Chinese researchers already have significant data about sand storms which they are utilizing for forecasting, which is through satellites and ground observation centers, also data from simulation models of different types.

When fast winds blow across dry areas, it lifts dust particles from the ground into the air, at times as high as 1500 meters, and takes them from one area to another. According to experts, the fast moving combination of dust and air can travel quite long distances. And they also take bacteria and metal particles, which are toxic, along with the dust, this combination is quite damaging for the health of people and also for the environment.

As during these storms, mortality rates increase, not only from wind caused accidents but also from cardiovascular and respiratory diseases. Deaths from cardiovascular problems increase by 25%, and those from respiratory problems increase by 18 percent.

There is also a climatic cost to these dust storms, estimates are that they reduce water and soil nutrition and it could result in lower crop yields. Mongolia estimates a 24% reduction in the yields of its crops from these very storms. On a global scale, 334 million people get affected by sand storms every year.

AI systems for predicting dust storms

Chinese AI systems can forecast dust storms in 13 countries across Asia, including China, Tajikistan, and Pakistan, 12 hours earlier and on an hourly basis. When they started trials last year, AI based systems showed better results as compared to non AI based systems, with 13% less errors. Researchers are also working on bringing these systems to mobile apps for the public to get timely dust storm forecast and alerts.

Chinese scientists have also developed a system called DAPS (Dust Assimilation and Prediction System). This system can provide forecasts for 48 hours. The system utilizes data assimilation, in this process, observational data is integrated with AI model calculations for the accuracy of forecasts. This system makes the process almost automatic. Deep learning algorithms are used for predictions by removing bias in the observational data. The system is helpful in giving forecasts for China, Mongolia, Japan, South Korea, and also North Korea, with detail as to how dense the storm could be and how the dust will spread across different areas.

A dust storm sweeping across part of Pakistan, Iran, and Afghanistan. Source: NASA.

Climate change also plays a role in worsening weather conditions, but Arctic amplification, which changes wind circulation patterns and is a meteorological occurrence, has reduced dust levels in south and west Asia. But the same phenomenon has worsened dust conditions in Pakistan, where Karachi and Lahore, two of its major cities, are facing air quality degradation, often resulting as most toxic and in top ten around the world.

Tree plantations play an important role in controlling the desertification of land, which is the major cause of dust storms. As agree that dust storms mostly affect remote areas with less development. Anti desertification efforts must be improved with government support to reduce the impact of storms and reduce dust levels, along with technological advancement.


Earn more PRC tokens by sharing this post. Copy and paste the URL below and share to friends, when they click and visit Parrot Coin website you earn: https://parrotcoin.net0


PRC Comment Policy

Your comments MUST BE constructive with vivid and clear suggestion relating to the post.

Your comments MUST NOT be less than 5 words.

Do NOT in any way copy/duplicate or transmit another members comment and paste to earn. Members who indulge themselves copying and duplicating comments, their earnings would be wiped out totally as a warning and Account deactivated if the user continue the act.

Parrot Coin does not pay for exclamatory comments Such as hahaha, nice one, wow, congrats, lmao, lol, etc are strictly forbidden and disallowed. Kindly adhere to this rule.

Constructive REPLY to comments is allowed

Leave a Reply