A high-tech crystal ball maps future development on the Delaware River

Attention planners!!!

Wanna buy a crystal ball?

It sees into the future (say, 2070) and can predict where growth will happen. Interested?

Well check out this website: drbproject.org.

By now, the talented people who are reading this at the Center for Land Use and Sustainability at Shippensburg University are cringing -- and who can blame them? There is no crystal ball, just years of hard work, science and data to support this goal:
"We aim to develop a land-cover mapping, modeling, and monitoring system for the Delaware River Basin in support of maintaining and restoring water resources, and to address the impact of future development and environmental change in the DRB."

That's taken from the introduction to the website, and what a site it is.

Using satellite images from the past and comparing those to what the images are now, the land-use specialists can make "educated guesses" on how the future might look.

It's not that complicated in theory, but enormously complicated to carry out.

So, you start with those early images, then introduce hundreds - thousands - of possible scenarios to those images, watch how development is indicated by those "pretend" scenarios, and compare the results of those scenarios with what the actual state of play is right now.

The smart part is that you then take the scenario that's given you the closest result to what actually happened and that becomes the route map for what might happen in the short and long term.

Pretty genius, huh?

Here's the process:

First, mapping: High-resolution LiDAR-based (LiDAR is an extremely accurate way to measure distance with pulsed laser light) land cover data for all 43 counties in the watershed. That's a real-time look either at the past or the present. The product produced in the mapping stage is a high-powered, densely detailed tool for conservation practitioners and planners to use. Because of the size and detail of the data, it wasn’t used directly in the modeling work.

Then, modeling: Developing models of land-cover change. This is the guts of the project. Thousands of possible changes are run to establish the most likely pattern of further growth. Eventually the models will include climate change, hydrology, and tree species.

Monitoring: Feasibility analysis used to establish a long-term land-cover monitoring program.

So the first two steps are done, the last is a bit trickier. It's not up to the Center for Land Use and Sustainability to dictate how towns and counties plan their communities.

After all, each and every municipality in the Delaware River watershed is dynamic and unpredictable. Pipelines and changes in transportation needs can alter what the future might look like. How counties and municipalities -- and their residents -- change the zoning is also a big factor. Dr. Claire Jantz, director of the CLUS and professor in the Department of Geography-Earth Science explained that the changes like these mean the mapping project has to be an on-going effort -- incorporating these unexpected changes will result in a better result for maps of the future.

Jantz also explained that the starting point for a model like this is good set of satellite photos that show what the earth looked like a decade or more ago.

The starting date for this model is 1996, using the National Land Cover Data Set. Then the thousands of different possible variations are run, then it was validated in 2011 -- which means comparing the reality with the projections. From there a further projection to 2070 was prepared, in three different scenarios: one based on current trends; one on sprawling development along transportation corridors; and one on compact development around existing or historic centers.

Anyone can go online and see the maps, but they are not in fine detail. You won't for example, see your house.

Jantz explained the target is "Anywhere there's a built environment. The satellite cell size is a quarter acre. So, we aren't looking at a cabin in the woods, but four or five houses on a crossroads.

"The model grows it forward thousands of times, maybe 3,000 different times, each of those times using different rules for development.

"At the end of all those we look for a fit.

"It will map clusters and new developments," she said, "but what you're looking for is the closest match to what we actually have. We prove the validity of the model to use it in forecasting.

"Then you take that proven version and use it to predict that likely growth at 2070."

The maps, she said, can help government manage processes.

"When we're modeling into the future, we say, 'We think population growth will be this much.' The question for planners is 'What kind of development do you want, and what should be the policy to encourage it where you want it?' "

There are really two faces to this project, both are open to public use. The first are the maps which anyone can look at without any sort of professional expertise.

The second is the data itself, which can be downloaded and used by GIS (geographic information system) professionals who work in planning departments or for conservation groups. The tool can help conservation groups see where its activity would have the best bang for the buck.

You can explore the effect that amenities would have, or commuting patterns. It doesn't give impervious land cover changes but registers a highly developed area, which is termed urban.

When Lauren McGrath, the field study site coordinator for the Willistown Conservation Trust, first explored the model, she said, "It drove home the importance of land preservation and how the protection of small parcels really adds up to maintain the landscape, and therefore the water.

"We were looking at development in our area, and found that one of the larger parcels that did not have an easement during the time of the map's creation and was predicted to be developed, did in fact, become developed during the predicted timeline," she said. "It was really fascinating to see how accurately the modeling worked."

All of this work was supported by a $1 million grant from the William Penn Foundation to the Center for Land Use and Sustainability at Shippensburg University. Sub grantees were the University of Vermont Spatial Analysis Lab and the United States Geological Society.

Antonia Price, the project manager at CLUS, said that there were several models of possible land and/or water use that have been developed but there's bit of a stumbling block in combining them since they all use different frameworks.

"We continue to stay in touch with both of these organizations as they finalize their products and hope to have the opportunity to collaborate in the future," she wrote in an e-mail.

This project joins two other tools that are helping to give planners, conservationists, scientists -- and anyone interested -- a better sense of what's happening now with our water. The first is the Stream Reach Assessment Tool, (https://www.streamreachtools.org) which is a product of the Academy of Natural Sciences of Drexel University, also funded by the William Penn Foundation. It supplies a way to "see" a type of pollution coming from a specific source -- called point-source pollution. Price pointed out the effectiveness of both tools would be increased if they could be combined -- at this point they don't talk each other's language, so to speak. That's another huge step, and is so far not funded.

The last in this trio is the wikiwatershed.org , a web toolkit developed by the Stroud Water Research Center, funded in part by the William Penn Foundation as well as the National Science Foundation among others.

The CLUS team which includes its own GIS professional, Alfonso Yáñez Morillo, also created an estimate of the combined effect of storm surge and sea-level rise in the Delaware Bay -- the bay shores of New Jersey and Delaware.

This was folded into its model of the whole Delaware River Basin. Jantz was reluctant to recommend this model to predict the variety of changes possible as the effects of climate change come to pass.

"We don't know the absolute effect,"Jantz said. "A lot more rain is predicted for fall and a lot less in the summer -- how will that affect flooding, or dams, or our forests? We're going to need new models.

"But the goal for this project," she said, "at the very least is to spark conversations about our future."

Check out the website http://drbproject.org and you can be part of that conversation.

About Meg McGuire

Meg McGuire has been a journalist for 30 years in New York and Connecticut. She started in weekly newspapers and moved to full-time work in dailies 25 years ago. She knows about the tectonic changes in journalism firsthand, having been part of what was euphemistically called a "reduction in force" six years ago. Now she's working to find new ways to "do" the news as an independent online publisher of news about the Delaware River, its watershed and its people.

Leave a Comment