Using satellite monitoring data, the researchers developed a deep learning algorithm that can provide real-time monthly land use and land cover maps for parts of India.
India is one of the 10 most forest-rich countries in the world, with some 80.9 million hectares of trees covering the country – about 25% of the country – but this is a significant drop compared to the past few years. Between the 1890s and 1990s, a combination of rapid economic development and overexploitation of local resources caused India to lose nearly 80 percent of its native forest area. Now, as India’s forests continue to disappear, researchers are working to help protect what’s left.
“Our work aims to help the Indian government and industry improve the country’s attempts at forest sustainability,” said lead author Ying Zuo, a graduate student in geosciences at Ohio State University.
The land-use monitoring system was trained using data provided by the Norwegian International Climate and Forest Initiative (NICFI), a Norwegian government enterprise that aims to reduce the destruction of tropical forests, in part by providing high-resolution imagery of the world’s tropics . This product was generated using imagery from PlanetScope, a satellite constellation that takes images of the entire Earth on a daily basis.
By combining data from NICFI products with global land cover maps produced by Tsinghua University, their deep learning model was able to obtain more detailed types of regional basemaps.
“To combine the two datasets into the same system, we resample them to the same spatial resolution and align each pixel to create an image-labeled paired training dataset,” Zuo said. “This process helps us ingest both datasets so they can be used to train our deep learning models.” This essentially combines thousands of small images into one larger basemap.
After training their deep learning model on these new satellite images, the team was able to process 10 basemaps of the region from January to October 2022.
The research poster was presented last week at the annual meeting of the American Geophysical Union. In her presentation, Zuo said that using the maps, the team was able to detect seasonal changes across India, such as changes in barren land, how farmland is affected by monsoons during the rainy season, and the distribution of forests in mountainous regions.
One conclusion of the study is that ecologists must look more closely at the seasonal effects of monsoons on forest cover in India. Understanding these seasonal changes can help scientists understand the effects of climate change on forests.
“As Earth’s average temperature increases, natural disasters will become more frequent, so we can use these maps to help everyone understand how this problem affects life on Earth,” she said.
In addition, if the team can expand the time span of these basemaps from months to years instead of months, Zuo says the better results could help scientists study other annual changes around the globe, such as floods.
“The characteristics of native forests and their surrounding habitats may vary in other regions,” Zuo said. “But with the help of more detailed datasets, our work could easily be used in areas of the world where detection and alerting the public to forest degradation and its side effects is a priority.”
Co-authors of the poster include CK Shum, Rongjun Qin, Yuanyuan Jia, Guixiang Zhang and Shengxi Gui, all Ohio State researchers. This work was supported by the USAID Forest Sustainability Program.
Source of story:
Material Provided by the Ohio State University. Original book by Tatyana Woodall. NOTE: Content may be edited for style and length.
