Water management could be transformed by new web application

Representatives of the Environmental Defense Fund, National Aeronautics and Space Administration, the Desert Research Institute and Google recently announced their web application in development, OpenET, to help transform water management in the western United States.

Robyn Grimm, senior manager, Water Information Systems, Environmental Defense Fund, said there are thousands of people across the West making decisions about water every day. This includes farmers, local and regional water managers, state and federal planners.

“With water supplies becoming increasingly scarce, many of them are constantly being asked to do more with less,” she said. “And this makes good data and information that much more important.”

Grimm said it’s surprising that “one of the most important pieces of information for water management in the West,” is difficult to obtain. That data is the amount of water that’s consumed by crops and other vegetation as it grows.

“OpenET aims to fill this huge data gap,” she said. “For the past three years scientists from NASA, EDF, the Desert Research Institute, USGS and the USDA have been working with Google Earth Engine and many university partners and dozens of stakeholders on OpenET.”

OpenET is a web application that aims to enable improved water management by providing field scale evapotranspiration data across 17 western states. Evapotranspiration is the process by which water evaporates from the land surface and transpires from plants. Evapotranspiration is the second largest component of the water cycle and can be thought of as the opposite of precipitation.

“It’s water that leaves the land surface and goes back to the atmosphere,” Grimm said. “And as such, it’s a measure of the water that’s lost to a local system, after being applied to the landscape.”

The water that’s applied to the landscape might run off back into a local stream, canal or recharge the groundwater basin, but evapotranspiration is essentially the water that’s effectively gone.

“Trying to manage water without ET data is kind of like trying to balance or manage your checking account without knowing how much you’re spending every day or where that money is going out,” Grimm said.

There are multiple ways to generate estimates of ET and they’ve been refined over decades of research, according to Grimm, but the estimates from different approaches are often varied and, at times, can be large. Because the data is so fragmented and generated with such different approaches, only those with the technological or financial capability can afford it.

“That can lead to some confusion and lack of trust in the data and generally an access barrier that keeps it out of the hands of many who could really stand to benefit from its availability,” she said.

Availability

OpenET will address this challenge, Grimm said, and it is publically available data from multiple satellites, weather stations and the Google Earth platform to bring together several models for calculating ET into a single platform for the first time.

“We brought the scientific community behind those different ET approaches together, and are using best available science to make consistent, near real time and field scale ET data easily accessible for the entire western U.S.,” she said.

Grimm expects OpenET to go live in 2021, and anyone should be able to go to a website, zoom in on any geographic region in 17 western states and get near real time ET data for that location.

There’s a lot of potentially exciting applications for OpenET, she said, and the first is to allow for the development of more realistic water budgets so decisions about water allocations in a certain area are based on a “really clear understanding of current and historical uses.”

The second is it enables local water managers to develop incentive driven conservation programs, driven by real data and further that data to help credit those who have made investments in conservations and done so successfully.

“Similarly, the ability to track water use in near real time can significantly reduce the transaction costs for water trading programs and make sure those are done in a well–informed manner,” Grimm said.

And finally, the applications of ET data at the field scale, where it can be incorporated into ranch management or other irrigation management software to maximize crop per drop and reduce costs for the farmer.

From the outset OpenET was driven by three goals—reliable data that’s available; broad trust in the data from water managers and decision makers at all scales; and a combination of those two things so a variety of sustainable resource management practices are enabled at a scale that wouldn’t be possible without OpenET.

“We recognized from the very beginning how critical that middle trust component is,” Grimm said. “And as such, we’ve prioritized engagement with end user communities from the very beginning.”

The groups have worked with more than 100 farmers, practitioners, water managers, state and federal planners to guide and test development.

“In fact, one of the first things we did was talk with all of those folks and develop a use requirements report that really forms the basis of the design for the platform,” she said. “We’ve also worked to ensure trust by bringing together leaders in the scientific community behind the methods for generating those ET estimates.”

From NASA

Forrest Melton, program scientist for the NASA Western Water Applications Office, said one of the strengths of OpenET is it relies entirely on publicly available data as inputs. It begins with satellite and mineralogical data that’s stored and updated daily on the Google Earth Engine platform.

“OpenET uses this data to drive an ensemble of ET models,” he said. “Model outputs are stored at the original satellite resolution, which is 30 meters by 30 meters, 0.22 acres per pixel, and data also summarized for millions of individual field boundaries for fields across the west.

Results are stored in a geo database and can be accessed through web-based graphical interfaces and through an application–programming interface, which gives access to the OpenET data services. Melton said one question he gets asked is why is this being developed.

“There have been two really important advances over the last roughly five years,” Melton said. “First is the availability of the Earth Engine computing platform, which is focused specifically on enabling analysis of Earth observation data using Google’s cloud platform. And second, really hard work by teams at NASA, USGS and European Space Agency to make satellite data freely available and available with very low latency, as close to real time as possible.”

There are other models that calculate ET from satellite images developed by others, but they require multiple inputs and other data. They can have information on land cover, topography and can be computationally extensive to run, according to Melton.

“Earth Engine is making it possible for the first time to operationally run, not just one model but an ensemble of ET models driven with satellite data over and over an area that’s a millions of square miles, covering 17 western states,” he said. “Earth Engine also allows us to bring all these inputs together on a cloud–based computing platform that allows the teams to collaborate and work together on the development of the highest quality input possible.”

Then they’re able to rapidly conduct comparisons among the different approaches and models, making it possible to identify the strengths of each approach. The application–programming interface is important because of integration with other software tools and systems. This includes irrigation scheduling software tools and farm and ranch management applications already used by local, state and federal water data information systems, water trading and water accounting platforms.

“We recognize that for many applications ET data is most powerful when it’s combined with other sources of information,” Melton said. “And this is especially true for irrigation management. Irrigators have to account for factors including fertilizer and pesticide management, harvest dates and coordination of activities across dozens of irrigation blocks.”

Open ET will allow farmers and irrigators to access data directly or via other farm and ranch management software they might already be using.

Google’s part

Tyler Erickson, developer advocate, Google Earth Engine, said the Earth Engine is one of Google’s cloud-based platforms that combines a very large catalog of data—earth data—with geospacial functions that it can use to analyze data. All of this runs at scale in the cloud in Google data centers and it can be used to make reductions of large amounts of satellite data or weather model data into information that decision makers can use.

“At Google we created this platform to help scientists address many of the global environmental and social issues like deforestation and water scarcity,” Erickson said. “And because Earth Engine runs the cloud, researchers are able to collaborate and just move science forward faster because they can share their data sets, their analysis codes and the outputs of their analysis with their collaborators or with the general public.”

The OpenET team is taking advantage of the Earth Engine running their algorithms over the western U.S., comparing and continually refining results and improving models.

“But it all starts with data,” Erickson said.

From a farmer

Ten years ago, Mark Owens of Harney County, Oregon, didn’t even know what evapotranspiration was.

“I thought ET was a movie, and I couldn’t pronounce it,” Owens said.

Now an Oregon state representative, the water situation in his basin was one of the reasons he got into politics, starting as a county commissioner. In 2015, the Oregon Water Resource Department came to producers in the basin where Owens farmed and said the water was over–allocated for what’s available.

“They showed us a simple water budget, the amount of water that came in to the amount of water is being used,” he said. “We’re a closed basin and all sorts of water was not escaping. We had 160,000 acre feet of water available for all uses, and they allocated approximately 170,000 acre feet.”

Owens and others in the area knew there was an issue, and the USGS conducted a five year comprehensive groundwater study, that’s just now finishing up. The study told them they were “over–allocated by approximately 120,000 to 130,000 acre feet.” The amount of pivot irrigation in the basin is about 120,000 to 130,000 acre feet just by itself.

“There’s a big issue,” he said. “We’re going to have to take some time and work through this issue and figure out how we can reach a reasonably stable groundwater level.”

At an alfalfa conference Owens attended, in 2015, he listened to a speaker talking about converting a normal pivot system to a low elevation spray application system. At the time of the conference, water allocation talks were just starting in Harney County. The speaker at the conference told Owens he could possibly reduce his water use by 18 to 20% just by converting his irrigation system.

“So that made a lot of sense to me,” Owens said.

He began looking into system designs and how to manage his water more effectively through irrigation. He tried to look up evapotranspiration and weather information in his area, and hit a brick wall.

“We didn’t have any good ET data,” Owens said.

But since the USGS study was going on in his area, he took advantage of some of the teams available and began to start to design irrigation systems to try to meet the evapotranspiration of the crop.

“So instead of designing irrigation systems at 7.5 gallons per minute per acre. We went down to 6. We went down to 5.7. We went to 5.8,” Owens said. “We were trying to hit that sweet spot where we could have conservation with efficiencies, without reduction of yield.”

Owens and those in his basin are going to have to have a lot of other tools in order to solve their water problem and the availability of the ET data is a very important part of the puzzle to help their water situation.

“I’m thankful this information is going to come online, because it is the biggest data gap that we have in the western United States,” Owens said.

For more information visit www.OpenETdata.org.

Kylene Scott can be reached at 620-227-1804 or [email protected].