18. Satellites, Data and Earth Observation: Signal from Noise – with Dan Hammer
SYNOPSIS:
How can satellite data and computation fundamentally shift how we understand our place on a changing Earth, and amongst other species? Can we use all that newfound knowledge, transparency, and intelligent data architecture to become better stewards? Allowing the earth to behold itself and its own lifeworld in a whole new way… And what are the ethical implications of having the power of such oversight? In whose hands?
Today our guest is Dan Hammer, Managing Partner at Ode, a data and design agency for the environment, and prior chief data scientist at the World Resources Institute, where he co-founded Global Forest Watch, a tool that tracked and monitored global deforestation patterns. He is founder of Spaceknow, a satellite image analytics start-up, and was a senior advisor in the Obama White House, a Presidential Innovation Fellow at NASA, creator of Global Plastic Watch and Amazon Mining Watch.
His work has used direct earth observation to locate every wastewater pond in rural Alabama; to watch illegal mining unfold in the Amazon; and to find every plastic waste site along rivers in Vietnam. He created the application Climate TRACE for former Vice President Al Gore, the first facility-level global inventory of greenhouse gas emissions, and much more.
In this episode speak about his new endeavour which is attempting to create an open source foundation model for nature – where you can “start to query the landscape like you would Google Maps”. I ask Dan how he manages to strike a balance between high level global information layers, and local relevance, and whether is it really possible that a global model can actually help people on the ground develop a deeper intimacy and action with the lifeworlds of where they reside.
Your browser doesn't support HTML5 audio
Satellites, Data and Earth Observation
QUOTES:
You can't see gender inequality from space. You can't see income vulnerability. But you can see these sort of breadcrumbs, and it takes a lot of ground truthing as well to be able to interpret what’s really happening there.
Over the course of the first couple years, there was a steady beat of around 1,500 to 2,000 people coming every day to the Global Forest Watch site, but that steady beat was punctuated by these massive spikes in usage when a journalist used the data in the service of reporting on an environmental or humanitarian event.
Alexa: It's this act of translation that I'm very interested in. How do we speak on behalf of the forests using this data? What do people want to see? How do people use the data?
We hear of this amazing work from IBM and NASA being able to do these foundation models for on earth observation, but there's a gap between that and having a local journalist reporting on artisanal mining in the Amazon being able to actually leverage that power.
Alexa: To give another metaphor, it’s you're training these algorithms or these neural networks to become like animal trackers or human trackers that are learning how to read signs across a landscape. It's fascinating that we're creating these artificial intelligences and we're training them to explore the land from space in the same way that we do physically inside the forest. It's a similar process of very local pattern recognition across scales
For a long time there was asymmetry in the information that has been created by space born sensors. And so to give open access to that data I think helps to rebalance the power. I think it's a progressive rather than a regressive technology.
SHOW LINKS
Dan Hammer - danham.me/r