-
3D in Geospatial: NeRFs, Gaussian Splatting, and Spatial Computing
Neural radiance fields and Gaussian splatting provide new ways to create 3D replicas. In this post I explore how they relate to traditional methods such as photogrammetry and 3D engines, their relationship to geospatial, and what applications we might see them in.
-
Large Language Models and Geospatial
Large language models (LLMs) like ChatGPT are at the center of a lot of conversations currently. I’m particularly interested in how they will impact the world of geospatial. The aim of this post is to think through some of the ideas I’ve had regarding those impacts. Due to a combination of slow writing and a fast moving field, it’s also a tour of some neat proof-of-concepts.
-
Can Diffusion Models imagine Natural Disasters?
Natural disasters impact the built environment. In this post I look at whether we can use the latest generative models to imagine the consequences of an earthquake, flood, and fire on a specific location.
-
ArtGIS: A satellite view of British Columbia
Satellite imagery provides an unprecedented view of the Earth. And while there are extensive practical applications, it's also neat to look at for its own sake. So I set out to create images of British Columbia using free and open access remote sensing products. I produced four images which highlight BC's terrain, nighttime light footprint, and the difference between winter and summer seasons.
-
Data Science Shouldn't Wait
Data science should not be delayed until after data governance and other supporting initiatives are implemented. While this will require additional effort from stakeholders, the costs are outweighed by the increase in data-driven thought, organizational familiarity with data science workflows, and impact on the bottom line.
-
Automating Traffic Analysis with Machine Learning
Data science and machine learning present new opportunities for improving public spaces. Leveraging these technologies for smart cities can make our communities more livable, more sustainable, and benefit local economies. They can assist with understanding key questions in city planning and urban design such as how public spaces are used, how many users there are, and who the users are. In this post, we'll look at a proof-of-concept system we implemented to answer these questions using machine learning for video analysis.
-
The Future Hydrology of the Pacific Northwest
Water is a critical resource in the Pacific Northwest. The effects of climate change on hydrological resources will be felt by a wide range of communities. Cities, First Nations, farmers, and the environment will all be affected. Effective management of water resources will become increasingly important to meet the diverse needs of the region.
-
Mapping Industrial Facility Greenhouse Gas Emissions in BC
Visualizations are a great tool for gaining intuition about a dataset. When the dataset has a geo-spatial component, overlaying the data on a map can be a starting point for exploratory analysis. In this blog post I want to share a map of industrial facility greenhouse gas emissions in British Columbia. It's based on this 2016 dataset, wich covers industrial facilities emitting 10,000 tonnes or more of carbon dioxide (C02) equivalent per year, as well electricity import operations in British Columbia.
-
Understanding Neural Networks
Neural networks generate a lot of interest. However, it's not always clear to people outside of the machine learning community the problems they're suited for, what they are, or how they're built. We'll address these topics in this blog post, aiming to make neural networks accessible to all readers.