Via Planet’s Education and Research Programresearchers around the world are finding new ways to harness Planet satellite data to better understand how our societies and Planet’s natural systems are intimately intertwined at all scales, from revealing the impacts of climate change on thawing permafrost to model how war affects rural production and markets With our data sets, these researchers can make novel discoveries and accelerate humanity toward a greater understanding of our complex world.
Permafrost thawing
As the climate continues to warm, researchers urgently need means to monitor permafrost degradation and its consequences in cold regions. One such region is the Qinghai-Tibet Engineering Corridor that contains a major highway and railway across the Tibetan Plateau, where the permafrost is shallow and therefore vulnerable to degradation.
This thawing permafrost can be seen as thaw collapses, ice slope fault areas that subsequently retreat and expand, which can cause the release of greenhouse gases previously trapped in frozen ground and damage infrastructure. and the surrounding ecosystems. However, conducting research on these depressions has been challenging in these high mountainous areas, and scientists are now turning to Planet for remote sensing data from these regions.
“Our recent study used PlanetScope imagery and deep learning to generate the first comprehensive map of melt depressions along the Tibetan Plateau engineering corridor,” said Dr. Lin Lui of the Chinese University of Hong Kong.
With the help of PlanetScope data, the study looked at 875 melt depressions with 38 of them located near roads or railways. The research team used these comprehensive datasets in combination with artificial intelligence to create a baseline dataset for analysis of these rapidly changing troughs and to assess the impact this may have on the critical road and rail infrastructure system. . This framework seeks to help stakeholders mitigate potential hazards in the region.
“We chose PlanetScope imagery not only for its excellent resolution, but also for its comprehensive coverage and high quality. Those advantages provide critical support for the identification of small melt falls by visual inspections and deep learning networks in our framework,” said the research team from the Chinese University of Hong Kong.
Mapping of urban forests
Urban sprawl in Chattanooga, Tennessee, has been occurring rapidly since 1987. With this continued expansion also comes the fragmentation of local temperate forests, degrading habitat connectivity and reducing forested areas that are culturally, ecologically, and recreationally significant in the southeastern United States. However, in collaboration with the City of Chattanooga and the City Forester, University of Tennessee Chattanooga master’s student Will Stuart is conducting research to map the urban extent of the city’s forest canopy.
“Determining the extent of the urban forest canopy in a rapidly growing city could help direct sustainable development efforts towards areas that are urbanizing at a higher than average rate, especially if the city administration is open to the principles of sustainable development. Stuart explained.
Using SkySat satellites, the research team obtained high-resolution images of the city in a quick time frame compared to traditional field sampling. Their research allowed them to perform a spatio-temporal analysis of the extent of the modern urban forest and to compare it with data from a 37-year period. Their findings indicated that the impervious surfaces in the city have increased over the 37 years by 134% and the area of temperate forest decreased by 43%.
Thanks to the fine spatial resolution of [SkySat] images, a technique known as object-based classification can be successfully employed, in which a selected machine learning algorithm (we use a support vector machine) classifies real-world objects within an image instead of classifying pixels. individual in themselves,” Stuart said. “The fine spatial resolution of the multispectral images obtained by the SkySat constellation facilitates the successful classification of trees planted in urban landscapes with very small crowns that could otherwise be grouped in an adjacent developed class,” he further explained.
Preliminary research presented at ESRI User Conference 2022:
Simulation of lava flow paths
Satellite technology is increasingly used to monitor volcanic activity, especially those in remote areas. Since lava flows can have devastating impacts on regional lives and economies, understanding the potential dangers of a volcano eruption is a top priority for researchers. Using satellite imagery, scientists seek both to observe the changing intensity of an active eruption and to work on eventual forecasting of areas that may be vulnerable to future eruptions.
Researchers from the Istituto Nazionale di Geofisica e Volcanologia, Sezione di Catania, Osservatorio Etneo in Italy are using a variety of remote sensing sources, including data from Planet, to assess the 2011 Nabro volcano eruption in Eritrea. The long lava flows from this eruption had a significant impact on the inhabitants along the Eritrea-Ethiopia border.
“[Our research institute] worked with PlanetScope data to update topographic changes at active volcanoes and quantify and measure the extent and thickness of volcanic deposits,” said Dr. Gaetana Ganci. “Planetscope data is also used to map vents and ephemeral vents during eruptions and to discover the presence of lava tubes,” she continued.
From these observations, the researchers were able to generate lava flow simulations to help understand the forward speed and extent of the lava flow, and observe the volcano’s eruption activity despite its remote nature. This research further demonstrated a step forward for volcano monitoring and hazard mitigation, even in regions without ground-based volcano sensor networks.
“[Our paper] uses PlanetScope imagery to compute a post-eruption digital surface model that allowed us to estimate the volume of lava that erupted during the 2011 Nabro eruption,” said Dr. Ganci. “Quantifying erupted volumes provides insight into emplacement processes and can aid computer simulations of lava flow paths for hazard assessment studies, providing input data for predictive lava flow modeling and tools for iterative validation or refinement of parameters”.
The full study is at Remote Environment Sensing.
Rural market modeling
In areas affected by conflict and poverty, changes in markets can be particularly challenging for rural communities. Farming and market trading are vital to the livelihoods of small communities, but to develop policies that support these economies, more information is needed on the impacts of disruptions such as lockdowns and violent conflict.
Dr. Tillman Von Carnap-Bornheim of Stockholm University used satellite data from Planet to monitor this market activity in Kenya and Ethiopia, adding temporal and geographic detail to existing research in this field. The study noted that Planet’s remote sensing data offers a solution for monitoring such markets when field analysis is not feasible, such as in times of war and violence.
“PlanetScope’s high frequency and resolution allow up-to-date information on what’s happening in rural markets, which are key places for farmers to sell their crops and buy the products they need. I am focusing on East Africa, particularly how the war in Ethiopia is affecting activity in the markets there,” said Dr. Carnap-Bornheim.
By monitoring crowd activity and market village infrastructure via satellites, the study was able to identify both the locations and attendance of communities at markets, as well as the timing of market days. This work mapped against rainfall shocks to look at the impacts of rainfall on agricultural yields and thus on market activity, drawing parallels between the three economic factors. The investigation further revealed that market activity decreased during the COVID-19 lockdowns and then again in the regions close to the ongoing war in Tigray when the violence occurred.
“Researchers and policymakers rarely have access to high-frequency data on economic activity in low-income countries. However, such information would be useful in understanding what policies work to reduce poverty or to target humanitarian assistance in times of crisis,” explained Dr. Carnap-Bornheim.
The full study is at SSRN.
Throughout the year, researchers and scientists are harnessing Planet data to better understand the complexities of the world we live in. We are proud to support such vital work and look forward to showcasing more researchers soon. If you are using data from Planet in your research and are interested in being featured in a future publication, please email [email protected].
Learn more about how Planet’s Research and Education Program could help you with your research here. Be sure to check out our Publications page to view scholarly articles published using Planet data, and don’t forget to follow us on Twitter to see the latest featured articles on #ScienceWednesday.