We spoke with Scott Klein, assistant managing editor, and Al Shaw, news application developer at ProPublica about the editorial decisions and satellite imagery used in “Losing Ground.” To read more about the science behind the imagery, read a case study by Fergus Pitt here.
Smitha: Where did the idea for “Losing Ground” come from?
Scott: Mapping is an important part of what we do. We hadn’t done all that much with satellite—maybe an image or two.
I attended the Tow Center’s Remote Sensing Seminar, a two-day conference, and I started thinking, people have used this in news, but no one’s used this to do an investigative story. I started talking to Fergus Pitt about what makes sense for an investigative story and among the things we thought would be interesting was using satellites–using satellites not just to illustrate a story but also as a key way of analyzing information. So Fergus and the Tow Center very generously offered to help us.
Another thing that happened is that we had been working in the Mississippi river delta for a long time, and we knew that there was a big story happening in Southeast Louisiana that was not known outside the delta which was this soil erosion, subsidence, land loss issue.
Al: Bob Marshall at the Lens had been writing about this extensively. I had also read a book called Bayou Farewell, which is a fantastic book about the people who live outside the levees and watching the land disappear, so there were a lot of different inspirations.
Scott: One of the things we wanted to look at with satellites was the Mississippi delta. In our reporting, we found that NOAA, the National Oceanic and Atmospheric Administration, maintained a list of places to remove from maps because of climate change and soil erosion. These are settlements, marshes, bays and rivers and things like that that are now just open water, so that NOAA knows they don’t have to provide weather alerts there, because you don’t have to provide weather alerts for places that don’t exist.
So this is fascinating for us, and we had seen that nobody had covered this before. This really inspired us to start asking the question, what’s being lost?
If you drew a circle around New Orleans, everyone inside the circle knows that this is happening, everyone outside the circle doesn’t.
We said this is a story we want them all to hear, we really think there’s a compelling reason to use satellites to analyze the information and then to use the satellite images to tell the story and let them see what’s being lost, both from a 30,000 foot view, almost literally, all the way down to personal stories.
And that’s when we started thinking, who has done really compelling work in southeast Louisiana? Who has the contacts and the means and the understanding to be able to go and tell personal stories of the people whose settlements, whose culture, whose livelihoods are being destroyed?
Al: Bob Marshall has grown up and spent his entire life down there.
He actually has a boat and knows these bayous by heart. When we first started talking about this, I went down there and asked him, “Say we had satellite imagery— what are the areas that you would like to see most illuminated?” and we kind of went through a map and he said, “This is one of the most important places.” We actually started drawing boxes around places, and that’s how the reporting built up from there.
Smitha: So you had a lot of local knowledge from Bob to rely on.
Scott: That was really a key thing that we didn’t have here in New York. Again, thanks to the Tow Center, Al and some other folks got some pretty intensive training from a studio that does satellite work as well as from Lela Prasad, a woman who used to work for NASA.
It was sort of like, when the people in the Matrix are asked, if each of you had to buy a helicopter: just over three days, there was an incredibly intensive knowledge exchange and Lela Prasad taught us exactly how to understand and work with satellite images.
And that was when things really got started.
Smitha: How easy was it to pick up this technology? Was it challenging, or was there a steep learning curve?
Al: It is somewhat of a steep learning curve. When you download these images right from NASA, they don’t look like anything really. And the tools are still somewhat rudimentary. You kind of have to cobble stuff together. We had to actually write a bunch of new software.
Scott: Now there are also a number of satellites with different capabilities and different caveats that you have to understand.
Al: The sediment kind of looked indistinguishable from the land, so that didn’t tell our story that well. We had to color correct the images because satellites don’t make water look like what we think of as water, like bright blue, or bright green for land. The images are also shot on different days, shot from different angles, there’s cloud cover. To turn that into the big image you see on the site took a fair amount of processing.
The 1922 map, that’s a United States Geological Service (USGS) map we got from Louisiana State University, and we geo-referenced it, which basically means adding geographic data to their standard image in order to line up with the 2014 land set satellite image.
Smitha: So there was a lot of stitching together geospatial information from very different sources as well.
Al: In the intro slideshow, there are overlays from the levees, the canals, oil and gas infrastructure, and pipelines, which came from different sources in USGS, from government sources in Louisiana, from a dissertation that an LSU student had done. So a lot of different sources went into it.
Smitha: One thing that really struck me about the way the piece is put together is that it’s very simple. Even though there is a lot of complex information, it feels very easy to navigate. How did you approach the issue of usability, the user’s experience of the site?
Scott: From the beginning we knew we wanted to make something where the maps were the main kind of metaphor. The maps were going to be the biggest thing on the page, the central spine of the interaction.
A few weeks ago we did some semi-formal user testing. We put a tweet out and sent an e-mail asking for volunteers to come and take a look at this. We watched them navigate through a draft of the app on a big screen and we asked them questions. It taught us a lot, and we cut out a lot of stuff. There was a whole different navigation metaphor that we left out.
Smitha: What are your hopes for the policy implications of this piece?
Scott: Our job as journalists is to inform the debate and give people as much information as they need to make really good decisions. Our hope for the policy piece is that we inform the debate in Louisiana. More importantly, I think focusing national attention on this will bring it needed scrutiny.
Smitha: What challenges are unique to working with satellite imagery?
Al: The raw size of the data is a big one. We went through tens of gigabytes of satellite imagery and other sources and being able to chew through that is a big barrier itself.
Smitha: The collaboration with The Lens, based out of New Orleans—-is this something you do a lot, working closely with local papers?
Scott: We do it very extensively. It’s a long tradition for us to work with local newsrooms.
Smitha: Has the cost of working on an interactive piece like this been prohibitive at all, or has it been a worthy investment?
Scott: It has absolutely been a worthy investment. The only costs have been staff time, travel—the imagery is all from the government.
Smitha: Do you have any projects similar to “Losing Ground” currently in the works or in gestation?
Scott: We do! We can’t talk about it. And this isn’t even the end of the Louisiana project, so we will have more to come.