Artificial intelligence has recreated what the Tisa River looked like more than two centuries ago

Researchers at Eötvös Loránd University (ELTE) in Budapest used artificial intelligence to reconstruct the historical appearance of the Great Hungarian Plain prior to large-scale river regulation and the draining of wetlands. To create the reconstructions, the researchers used historical maps, modern photographs, drone imagery, and other research materials.

Today, images can be generated using artificial intelligence in a matter of minutes. However, the speed of generation does not guarantee the scientific accuracy of the results. 

That is why ELTE researchers decided not to limit themselves to standard text prompts but instead provided the AI with a substantial amount of scientific data. Thanks to this, they were able to produce reconstructions with accurate topography of the area.

Historical maps served as the basis for creating the images. At the same time, modern photographs of the area and drone images helped recreate its actual natural features.

Additional sources of information included old vegetation inventories, pollen analysis results, and studies of charcoal and macrofossils.

According to experts, this approach makes it possible to create reliable, understandable, and informative visualizations that help reveal interconnections that are difficult to grasp based solely on text and tables.

In addition to reconstructing historical landscapes, the method can be used to predict the consequences of wetland restoration and changes in the natural environment caused by drought.

Just a few centuries ago, the Tisa River looked completely different. It was rich in meanders and side channels, and the surrounding areas were regularly affected by flooding. As a result, the Great Hungarian Plain was covered with floodplains, marshes, and other wetlands that were periodically flooded.

However, in the 19th century, during the Habsburg Monarchy, there arose a need to put the 35,000–40,000 square kilometers of land periodically flooded by the Tisa River to agricultural use.

To this end, efforts began to drain the floodplains and straighten the river’s course in order to free up land for agriculture. By the 20th century, most of the former floodplains had already been converted into farmland.

Large-scale river regulation significantly altered the region’s natural landscape. In particular, the length of the Tisa River was reduced from 1,211 to 758 kilometers, and that of the Keresh River, including its tributaries, from 1,041 to 462 kilometers.

As a result of these changes, regular flooding ceased, and a significant portion of the territory that had previously been covered by forests, swamps, and meadows disappeared.

While forests originally covered 37% of the Great Hungarian Plain, they now account for less than 1%. The area of floodplain meadows and marshes has also decreased sharply—from 44% to 3–6%.

This is stated in a report on the university’s official website.

Previously, an analysis of fossilized remains of ancient humans helped scientists reconstruct the appearance of Denisovans and determine the territories where they lived.

Scientists have found a way to transmit quantum information between distant qubits.

Chinese scientists have created a 100-channel photon detector—South China Morning Post.

The calorie content of coffee has increased by 40% due to plant-based cream, according to scientists.