Remote Sensing with Machine Learning for Mosquito Vector Surveillance

Guest scientist blog by Ishaan V., 2022 NASA SEES Intern

Have you ever wondered why NASA has chosen to follow water? NASA’s James Webb Space Telescope has captured the distinct signature of water outside our solar system. NASA's Artemis program is examining the presence of water on the Moon in advance of sending and establishing a sustainable human presence there.  The answer is simple - water is key to life as we know it, including on planet Earth. Only 2.5% of the water on Earth is fresh, and only a tiny fraction of that is available for use. NASA and other space agencies’ remote sensing satellites provide vital information to monitor the water cycle and inland water quality.  NASA SEES Internship program gave me the opportunity to learn about this and much more!

Learning and immersion started with the mandatory and optional courses we had to complete before starting the research project. These courses helped us understand climate science, the role of Earth-observing satellite missions, and multispectral scanners across the electromagnetic (EM) spectrum. One of my favorite sections was to understand why water is so incredibly weird, yet one of the most desired things in the universe. I got to experience our home planet earth from the eyes of a remote sensing spacecraft orbiting at altitudes between 700km - 36,000km, which provided a perspective I never had before.

As part of an optional series of python coding assignments, I got an opportunity to work with the Andromeda galaxy data to develop a color-magnitude diagram and perform simulations to study the effect of various factors such as dispersion, and inclination, etc. Thanks to Dr. Raja Guaha and his astronomy students for giving us the experience of working with spatial data analysis and visualization programmatically. This not only helped me improve my programming skills for scientific research but also made me curious to explore satellite data for the research project.

In the SEES Earth Explorers and Mosquito Mappers cohort, I learned about the importance of field research remote sensing to collect land cover data and environmental data in different regions. My field research involved visiting multiple sites with water bodies (creeks, ponds, lakes, canals, ditches, etc.) in the neighborhood. During my field study, I observed that I found more mosquito larvae in dirty ditches under an open sky. Based on my observations, for my mosquito traps, the specific variables I chose to control were centered around the quality of water and trap wall materials. I shared my successes and failures with my peers and mentors and updated weekly progress reports through the Earth System Explorers research forum.  Through my experiment, I was able to identify the mosquito species as Aedes aegypti which does transmit diseases.

Participation in land cover observations field research gave me the opportunity to have first-hand experience in using ArcGIS online tools and GLOBE observer apps. NASA and the GLOBE programs provide free access to GLOBE Observer data collected by citizen scientists and scientists. I am proud to contribute to this program for data to be used by peer SEES interns as well as by scientists globally.

For the final project, my research team and I decided to build a machine learning prediction model for the West Nile viruses by observing algae bloom using remote sensing satellite images. Eutrophication (algae bloom) is a major source of food for mosquito larvae, leading to vector-borne diseases.  Using the multispectral visible and near Infra-Red (V-NIR) images to train the ML prediction model, WNV outbreaks can be predicted in advance to help improve vector-borne disease surveillance and preparedness at the county and state levels. We used Sentinel-2 MSI images from 2019 to 2021 for the Sacramento and Fresno, CA regions to generate Normalized Difference Water Index (NDWI) and Normalized Difference Chlorophyll Index (NDWI) and combined them with GLOBE land cover data (water, vegetation, etc.) to build a feature matrix for the prediction model. The accuracy of the ML predictive model ranges from 0.7 to 0.95, depending on the algorithm and the length of time series used for the training of the model.

            The guest speaker’s series provided us the opportunity to listen to and interact with NASA astronauts and scientists working on project such as Mars Exploration, Artemis Program, and Space Force. Understanding how Project AEDES, winner of the NASA International Space Apps Challenge in 2019, used a data-driven approach to improve dengue surveillance and public health helped me understand the real-life impact of scientific research.

            Finally, having access to scientists, mentors, and fellow researchers from all over the nation allowed me to learn and collaborate with like-minded researchers with lots of hands-on fieldwork and data analysis projects.  This 8-week internship program not only helped me achieve the 3 goals I had set for myself this summer, but more importantly, the program helped me understand the importance of NASA remote sensing satellites in improving the quality of life on our home planet Earth.

​​​​​​​Ishaan V. is a  high school student at Bridgewater Raritan Regional High School, NJ. His virtual internship is part of a collaboration between the Institute for Global Environmental Strategies (IGES) and the NASA  Texas Space Grant Consortium (TSGC) to extend the TSGC Summer Enhancement in Earth Science (SEES) internship for US high school ( He shares his experience this summer in this guest blog post.

More Blog Entries