Hydrosat is a geospatial intelligence company that provides intelligence for food, security, public safety, and environmental purposes. Using that intelligence, the Washington-based organization aims to help Earth manage its most valuable resource: water. Surface temperature is largely driven by water availability, so Hydrosat’s thermal imagery can help to map and monitor crop water stress, drought, wildfire, urban heat islands, and many other application areas.
To achieve their operational objectives, Hydrosat will eventually launch a 16-satellite constellation. However, they first needed to demonstrate the power of their future data collection to early adopters and clients by providing synthetic daily high resolution land surface temperature products.
Lacking the in-house means to architect and deploy the Hydrosat Fusion Hub — the processing pipeline and data cataloging solution that produces these early data products — the team knew they would need the help of a third party. Consequently, Hydrosat reached out to Element 84 based on their technical domain expertise, reliability, and agile, iterative development approach.
Making waves with a custom processing pipeline
“Hydrosat came in with a conceptual idea,” explained Arthur Elmes, Senior Geospatial Engineer at Element 84. “They were aware of what Element 84 had done on other projects, so they knew what parts they wanted to use and that our team had the expertise to implement quickly.” Therefore, the project team began by deploying FilmDrop, Element 84’s suite of open source tools that ingest, archive, process, analyze and distribute geospatial data in the cloud. Building from the existing FilmDrop infrastructure as code (IaC), the engineers were then able to tailor its components specifically for the Hydrosat Fusion Hub.
Deployed in AWS (Amazon Web Services), the bespoke processing pipeline included data discovery from multiple sources: upstream inputs, data refinement and preparation, and science algorithms that were modified for large-scale cloud deployment. To facilitate data discovery and cataloguing, the Element 84 engineers also implemented a STAC (Spatio Temporal Asset Catalog) server, which captures systematic metadata and processing provenance.
“When we talk about the Hydrosat Fusion Hub, the ‘fusion’ refers to the fact that the data is a mashup of several data sources, each of which provides one desired attribute to the end product,” said Arthur. Subsequently, the processing pipeline was designed to be invoked for target Areas of Interest (AOIs) that provide Hydrosat and their early adopters sample data to analyze on an on-demand basis.
As an AWS-based solution, the team was able to use various AWS services, such as Step Functions, Lambda, and OpenSearch, across the pipeline. These services were assembled in order to coordinate a complex series of data queries and processing procedures, and to efficiently store and index outputs for secure data accessibility.
Understanding the importance of usability and security, Element 84 also built out a custom user management interface for the Hub. The interface grants different permissions for administrators, users, and guests, which protects Hydrosat’s data holdings by requiring authentication to view and download the daily data products.
The user management UI provides a simple means of adding, removing, and adding privilege levels to specific users.
Arthur Elmes, Senior Geospatial Engineer at Element 84
Finally, Element 84 deployed a JupyterHub environment to host Jupyter Notebooks. Jupyter Notebook is an open source web application that can be used to create and share documents containing live code, equations, visualizations, and text. As such, Hydrosat’s data scientists and early adopters had access to all the necessary data within the Hub.
Enabling clients to test the waters before launch
With Element 84’s geospatial processing pipeline, Hydrosat was able to provide early adopters with a live data product, comprising many thousands (n>10,000) of high resolution land-surface temperature (LST) granules that were produced in the Fusion Hub. The product gave potential customers an opportunity to see what the new data would look like and how it could be leveraged — from wildfire prediction and monitoring to disease vector mapping.
By combining the processing pipeline with their internal data stream, Hydrosat will also be able to augment their solutions and provide a richer data offering to customers. What’s more, since the solution was built on AWS, Hydrosat can easily scale it in the future to meet increased demand.
Today, Hydrosat maintains ongoing validation efforts, comparing the Fusion Hub outputs to ground-based LST measurements at four atmospheric science research stations.
Hydrosat’s first two satellite missions will be making a splash soon as they are scheduled to launch in early 2024!
Want to learn more about how FilmDrop can help you?
At Element 84, we love hearing about your projects — whether they’re napkin sketches or almost-ready-to-launch.
Whatever stage you’re at, our FilmDrop suite can help you manage your geospatial data, from small image collections through petabytes of complex SAR and remote sensing data.