Wrapping up the Summer of Earth Engine
By Devaja Shah, Project Coordinator, Geo for Good India
Everyone has the power to influence change in the world — even students at the very beginning of their careers — when connected with ideas and people that inspire change and have the ability to create real-world impact on the ground. The Google Summer of Earth Engine program, wrapping up this week in India, shows us what can happen when students gain access to the right technology and mentors, and set their minds to solving real-world problems.
At this week’s student mini-summit, we’ll be sharing five projects chosen for the Summer of Earth Engine program and each student will be training 100 students on topics like EE Apps, Tensorflow and Classification learnt during the program. While the projects are all very diverse — from using satellite images for a water quality study, to mapping locations of human-wildlife conflicts — they are all solving some of India’s most pressing environmental and conservation issues, using remote sensing and the power of cloud computing with Earth Engine.
Each of these projects were proposed by dedicated mentors from our research partner organizations in India such as WRI India, ICRISAT, and ATREE, who ensured that each project mapped back to the organizations’ missions. The mentors were responsible for providing the students with guidance on each of these scientific disciplines, along with the ground reality of the issues the students were trying to solve.
Google Earth Engine, the geospatial data-analysis platform, was used to perform the modelling and mapping for all five projects. Earth Engine’s cloud computing power and ease of access to public datasets can inspire many more students studying remote sensing and GIS to try their hands at impactful projects.
Summer of Earth Engine projects
Water quality monitoring using Sentinel-2 data
Jyoti Dhwanikar, a student at the National Institute of Technology Karnataka, worked with her mentor, Raj Bhagat, GIS and remote sensing analyst for WRI India, to study solids in water bodies, using satellite data to complement traditional ground monitoring stations. The ground stations are in short supply and create gaps in data — whereas satellite data is more consistent and doesn’t require constant check-ins.
“I’m keenly interested in researching the possibilities that can be created using satellite data,” Dhwanikar says. “This is when we need a cloud platform to analyze such large amounts of data, in less cost and time. This is where Earth Engine fits right in.”
For Bhagat,the project will help WRI India achieve the first step of its “Count it, change it, and scale it” strategy. “As an organization researching ways to improve water quality across the country, we need to begin monitoring the ground situation,” Bhagat says. “I believe this project could trigger further research on satellite-based water quality estimations.
Building an app to visualize change in conservation areas
Rajat Shinde, a student in geoinformatics at the Indian Institute of Technology Bombay, developed EcoViz, an Earth Engine app to study how world heritage sites like the Western Ghats mountains — earmarked for conservation — fare over time, using satellite imagery. Rajat was mentored by Mysore Doreswamy Madhusudan, co-founder and director of the Nature Conservation Foundation. Shinde hopes EcoViz will help people outside the GIS community, including administrators, journalists and citizen groups, better understand ecological status and change in the key conservation areas. Check out the app here.
“EcoViz can be used to visualize changes in forest cover, changes in precipitation content, changes in aerosol content, and atmospheric parameters,” Shinde explains. “It can be very difficult to monitor these parameters for remote places. We sincerely feel that EcoViz would be of great value in such conditions.”
Tracing wildlife conflict through cropping patterns
Abhishek Potnis, a student in the GeoComputational Systems Lab at the Indian Institute of Technology Bombay, helped the Centre for Wildlife Studies and its Wild Seve program identify croplands where farmers are likely to be in conflict with wildlife. The Wild Seve program compensates farmers for livestock injured or killed by wild animals.
With guidance from his mentor Anubhav Vanamamalai, the Centre’s program manager, Abhishek defined algorithms used to map crops such as horse gram and finger millet — with the goal of analyzing crops across multiple seasons and gaining a better picture of the landscape. “From writing the proposal to wrapping up on this Summer of Earth Engine project, it’s been a great learning experience for me,” Potnis says. “I’ve been inspired and touched by the work of the Wild Seve program, and it was an honor for me to contribute.”
With guidance from his mentor Anubhav Vanamamalai, the Centre’s program manager, Abhishek defined algorithms used to map crops such as horse gram and finger millet — with the goal of analyzing crops across multiple seasons and gaining a better picture of the landscape. “From writing the proposal to wrapping up on this Summer of Earth Engine project, it’s been a great learning experience for me,” Potnis says. “I’ve been inspired and touched by the work of the Wild Seve program, and it was an honor for me to contribute.”
“We’re happy to have scientists such as Abhishek help move our research forward and add to our understanding of the landscape,” says Vanamamalai, Potnis’ mentor. “This project laid the groundwork for a lot of future research, which will ultimately help in improving the coexistence between our people and wildlife.”
Mapping the spread of invasive plant species in India
Shristi Singh, a student at the Indian Institute of Technology Patna, worked with the Ashoka Trust for Research in Ecology and the Environment (ATREE) to map the spread of Prosopis Juliflora in Tamil Nadu. It’s an invasive plant that’s replacing areas once filled with natural forests as well as plant sources for animals. Shristi and her mentor, ATREE fellow Veena Srinivasan, worked to correlate the presence of the plant with variables like temperature and soil moisture.
“Invasive species disrupt the natural vegetation of an area, and have caused lots of damage to the farmers and agricultural economy,” Singh says. “I hope that with my model, we’ll be able to better detect these kinds of species so we can get to their root causes and eradicate them.”
Identifying crop irrigation methods through satellite images
Parmita Ghosh, a student at the Indian Institute of Technology Bombay, studied crop-specific domain mapping with guidance from Murali Krishna Gumma, head of the GIS and Remote Sensing Lab at ICRISAT (International Crops Research Institute for the Semi-Arid Tropics). Ghosh’s research helps identify crop types from satellite images using machine learning algorithms, and also identify crops that are irrigated and which are rain fed.
“In India, canal water management is becoming challenging — and at times, due to the high water usage by farmers, leads to water scarcity,” Ghosh explains. “If the precise water requirement for each field is available, then water management can be handled in a more efficient manner.”
“SoEE was a good platform for students and mentors. The way the program enabled the execution of a 3-month project solely through virtual meetings was amazing. It was an enriching experience for me as a mentor and I’m sure for the students as well” says Dr Gumma, Parmita’s mentor for the program.
Bring your projects to life using Earth Engine
We’re proud of the projects created by the Summer of Earth Engine students. We want to encourage everyone — students, nonprofit researchers, or anyone passionate about solving India’s current challenges and conserving natural resources — to bring their own vision and ideas to Earth Engine. The Earth Engine tools are free to use: Just bring your time and creativity. To begin your learning with Earth Engine, check out our tutorials and code labs.
