12-05, 10:00–10:30 (America/Belem), Room II
The State of California has a recurring issue of devastating wildfires that spread out of control, causing significant damage to the well-being of its citizens, environment, and economy.
Due to this recurring issue, the State of California and the USDA—United States Forest Service committed to treating one million acres of land (roughly 400,000 hectares) to reduce the occurrence and severity of these wildfires. Planscape was created to prioritize and guide this one million-acre effort.
Planscape is a freely available tool built to help landscape planners, public agents and the general public prioritize and learn about landscape interventions that minimize fire risk and cost while maximizing ecological benefits.
Planscape utilizes zonal statistics data gathered from over 200 raster layers, applying Linear Programming techniques to determine the best locations for planners to intervene in the landscape by performing treatments such as mastication, prescribed burns, etc.
This hands-on session will discuss several options for obtaining Zonal Statistics for Planscape. In this benchmark, we can compare a diverse range of technologies and techniques utilized to reduce the necessary time to obtain zonal statistics for roughly 24 million regular polygons covering the state of California and, over 200 million zonal statistics records.
This talk benchmarks the following methods: PostGIS + PostgreSQL, Python RasterIO, rasterstats, custom code.
George Rodrigues da Cunha Silva is a seasoned software engineer with a unique background that bridges the worlds of geography and software development.
Originally trained as a geographer, George has leveraged his deep understanding of geographical information systems (GIS) into a successful software developer and technical leader career.
With over a decade of experience in software development, George has a proven track record of delivering high-quality, maintainable code and architecting complex, distributed systems.
His expertise encompasses various programming languages and technologies, including Python, Django, Clojure, Elixir, AWS, Terraform, Kubernetes, and PostgreSQL. George has worked extensively in back-end engineering, particularly with distributed services, event-driven architecture, design patterns, and domain-driven design. His creative and resourceful approach to software development is complemented by his strong leadership skills. He has managed and grown teams of developers to achieve significant project milestones and operational efficiencies.
George’s journey from geographer to developer has been marked by significant contributions to various GIS tools, including QGIS and PostGIS, where he has engaged in bug reporting, feature development, and documentation translation efforts. His work has advanced the capabilities of these tools and supported the global GIS community by enhancing accessibility and usability.
In his recent roles, George has been responsible for determining initial architectures, building and maintaining features, and managing diverse teams of developers across back-end, front-end, and data science domains. He has successfully led projects that automated critical business processes, resulting in substantial cost savings and operational improvements.
His passion for delivering business value through clean and maintainable code and extensive experience in GIS and software development positions him uniquely at the intersection of these fields.