Geography and the coded world: GEEting to know Google Earth Engine

By Jacob Jacela

Last October 28, 2019, the Geography 190 class (Map and Air Photo Representation) hosted the 5th Spatial Technologies for Advancing Research Techniques (START) technical colloquium-workshop that featured introductory exercises on the use of Google Earth Engine (GEE) for image processing and analysis. The event was the 2nd of three workshops for this semester. Ms Fatima Pamittan, formerly a Science Research Specialist of the UP Training Center for Applied Geodesy and Photogrammetry (TCAGP) for the DIME Project, was the invited resource speaker for the session.

GEE combines a multi-petabyte catalog of satellite imagery and geospatial datasets with planetary-scale analysis capabilities and makes it available for scientists, researchers, and developers to detect changes, map trends, and quantify differences on the Earth's surface. It makes use of the Java language as a medium to input codes that enables the user to command the presentation of remotely-sensed imagery, maps and other GIS applications. 

Participants learned how to perform basic processing methods including cloud masking, filtering and band ratioing (e.g., Normalized Difference Vegetation Index or NDVI). In cloud masking, we were able to adjust the different characteristics and qualities of cloud data obtained from several images, including the cloud likelihood threshold, cloud probability band and cloud cover. For filtering, we learned how to specify the parameters for the data we wanted to analyze and to filter them by date or bounds / points, among other filtering methods. NDVI was also derived during one of the exercises to measure the greenery and health of vegetation observed in the set of images. 

In our day and age of artificial intelligence and machine learning, coding is actually a very valuable skill, and it was nice to have these kinds of crash courses even with just the basic commands. It was a very interesting session because it gave us insights on how coding is used in GIS projects, which could give us more freedom in working with our datasets. You can also see how proficient the speaker was in her field, which made the session more engaging. This session may have been challenging to take and learn, especially for those who are not that much familiar with basic programming, but the skills obtained could come in handy when encountering remote sensing systems that require knowledge of the Google Earth Engine. Overall, the session offered another way for students, intellectuals and researchers to understand and appreciate the field of remote sensing through a higher-level analysis using programming and spatial technologies. The use of GEE as a remote sensing tool could further be explored by inviting more researchers to consider it in their studies and integrating it more fully into remote sensing systems and spatial technologies.