Student Research - Mission Earth
Open social media sharing options menu
Dimensionality Reduction and Inference for Field-based Surface Temperature (DRIFT-ST)
Organization(s):United States of America Citizen Science
Country:United States of America
Student(s):Daksh Mamnani, Adhvaith Vinothkumar, Marlin Wong
Grade Level:Secondary School (grades 9-12, ages 14-18)
Educator(s):
Contributors:
Report Type(s):International Virtual Science Symposium Report, Mission Earth Report
Protocols:Surface Temperature
Presentation Video:
View Video
Presentation Poster:
View Document
Language(s):English
Date Submitted:2025-07-17
The GLOBE Program collects environmental data from
students and scientists around the world. Our project focused
on surface temperature data, which includes measurements
and metadata like date, time, location, and instrument type.
While the dataset was rich, it was also messy, with missing
values and inconsistencies. We used dimensionality reduction
to simplify the data and uncover patterns. This technique
reduces the number of variables while keeping the important
structure. Our goal was to determine which dimensionality
reduction technique best revealed the underlying structure of
the surface temperature data. Ultimately, we wanted to use
this structure to identify outliers and inform flag generation in
environmental datasets