In the past few decades, humanity has sent thousands of satellites into orbit. Some take pictures or map the Earth, others are used to improve the GPS in our phones, and some are equipped with very specific scientific instruments. This is the case with Synthetic Aperture Radar (SAR) satellites.
SAR satellites emit a radar pulse and then measure the reflection or backscatter. This reflection depends on physical properties such as height, roughness, or structure. SAR is especially useful because, unlike other types of satellites, it works just as well in any type of weather. But, it also has its own problems.
SAR data has limitations, including signal noise and image distortion, which complicate data interpretation. Urban studies, in particular, have struggled with the scarcity of long-term and high-resolution backscatter data.
Most urban monitoring efforts rely on optical images, which offer clear, two-dimensional views of surface changes but fall short of capturing the three-dimensional aspects crucial for comprehensive urban analysis. The solution would be to combine SAR and optical data, and this is exactly what researchers have done here.
“We set out to fill the knowledge gaps in the reconstruction of long-term backscatter coefficient data. This study developed a method that would reconstruct the data over the past three decades (1990 through 2022) based on Sentinel-1 Ground Range Detected data and long-term Landsat data in the Jing-Jin-Ji region. This could significantly support three-dimensional dynamics in urban domains,” said Xuecao Li, a professor at the College of Land Science and Technology at the China Agricultural University in Beijing, China.
JJJ, China
They looked at the Jing-Jin-Ji (JJJ) region of China, which includes Beijing and Tianjin and other major cities, between 1990 and 2022. Because this region has changed so drastically, it’s a good candidate for this type of research.
In terms of optical data, the team didn’t just use visual images. They used several optical properties, including spectral bands and vegetation indices. In terms of satellite SAR data, they used Sentinel-1 GRD data, long-term Landsat images, and ALOS World 3D data.
The approach was able to show how cities have changed and grown and mapped building heights with accuracy. For instance, Beijing’s urban core has seen a noticeable upward expansion, particularly in districts like Haidian and Chaoyang. The total building volume in Beijing increased at a rate of 0.420 km³ per decade, highlighting the city’s rapid vertical growth. In contrast, cities like Cangzhou exhibited much slower growth rates, underscoring diverse urbanization patterns within the region.
<!– Tag ID: zmescience_300x250_InContent_3
–>
Plenty of applications
The researchers did more than estimate heights. They used long-term data to classify urban surfaces and for flood monitoring. The ability to capture this type of data over decades is important for urban planners and policymakers.
“In addition to estimating building height, the long-term and high-resolution backscatter coefficient also has promising potential for urban studies regarding impervious surface classification, change detection, and flood extent mapping. The proposed approach in this study allows researchers to estimate the backscatter with finer resolutions for decades regarding seasonal or monthly changes.”
For the future, researchers have big plans. They want to expand the approach to a global scale and characterize even more urban parameters.
“The estimated building height in the Jing-Jin-Ji region can be further used for investigating the urban environmental issues regarding the 3D perspective, such as the urban heat island, dynamic building height, and carbon emissions, which are expected to provide useful information for achieving sustainable development goals,” said Li.
Journal Reference: Bo Yuan et al, Reconstructing Long-Term Synthetic Aperture Radar Backscatter in Urban Domains Using Landsat Time Series Data: A Case Study of Jing–Jin–Ji Region, Journal of Remote Sensing (2024). DOI: 10.34133/remotesensing.0172
Thanks for your feedback!
