Heavy Weather: JT Reager and Gravity-Based Flood Prediction


Award-winning NASA scientist J.T. Reager studies water resources using satellites that monitor the Earth’s gravitational field.

A rendering of the GRACE mission in orbit. Image Credit: NASA JPL

A rendering of the GRACE mission in orbit. Image Credit: NASA JPL

The continental Earth is like a bucket: It can only hold so much water. As the so-called bucket of soil fills with water, it’s eventually going to overflow.

According to NASA hydrologist John T. Reager, overflow that happens when the soil profile of a given area is fully saturated with water can cause severe large-scale floods that threaten lives and property. As a hydrologist who works with satellite gravimetry and water circulation patterns, Reager uses his work to help others predict those large-scale floods. In March of 2016, Reager won the NOAA David Johnson Award for his work in seasonal flood prediction.

Accurate flood prediction, he said, ‰ÛÏDepends on knowing how much water is in that proverbial bucket and knowing how saturated the soils are, where the actual water table is in the soils, and how much snowpack you have in the mountains.‰Û

Knowing how much water is in a given location tells researchers important information about terrestrial water storage (TWS), and when an area is more susceptible to floods.

Reager is a principal investigator on the Gravity Recovery and Climate Experiment (GRACE) Science Team at NASA’s Jet Propulsion Lab (JPL), where his research focuses on using slight changes in the Earth’s gravitational field to map changes in water circulation and concentration.

The Earth’s surface concentrates mass in different locations ‰ÛÒ mountains are heavier than plains, for example, and wet soils are heavier than dry ones ‰ÛÒ so its gravitational pull is variable. When satellites pass over areas of higher mass, the increased gravitational force in those regions pull the satellites closer to the Earth than areas with lower mass do.

GRACE global gravitational anomaly animation. Image Credit: NASA/JPL/University of Texas Center for Space Research

GRACE global gravitational anomaly animation. Image Credit: NASA/JPL/University of Texas Center for Space Research

Launched on St. Patrick’s Day 2002, GRACE is a system of two satellites orbiting the Earth in tandem at about 450 kilometers orbit height. The satellites orbit about 250 km apart from each other, observing each other’s movements as they experience the differing gravitational pulls of the Earth’s surface below. The pair has a microwave ranging device that can measure and record changes in distance smaller than the width of a human hair, said Reager. That data is then combined with information from the GPS systems atop both satellites in order to create a spatial gravity map.

Reager and his team use the data from GRACE to assess ‰ÛÏflood potential‰Û ‰ÛÒ essentially, how full the ‰ÛÏbucket‰Û is of a given area.

Once a month, GRACE completes a global gravity field map, which is compared to previous maps and analyzed for trends. If a given geographic location displays stronger gravitational pull over time, it is likely that the area is steadily filling with water. That means that it has a high flood potential and, if the area experiences a certain volume of rain above its soil’s carrying capacity, it will likely overflow and cause major flooding.

GRACE data allows the researchers to establish what U.S. Geological Survey economist Dr. Richard Bernknopf calls a ‰ÛÏTWS antecedent condition‰Û for a given area. When a baseline gravitational field is mapped onto a grid of that area, Bernknopf said, it can be converted into a prior probability of TWS so that specific precipitation and snowmelt events are considered in terms of how much water is already in the ground. åÊ

To retroactively test their methodologies, the GRACE team observed catastrophic floods in the Missouri River basin causing billions of dollars in damage and displacing thousands of people in 2011. Looking at the GRACE data from just before the floods, Reager said, his team noticed that ‰ÛÏthe bucket was fuller than it had ever been before in the GRACE record two months before the flood occurred. This river basin was already experiencing record high storage.‰Û A better understanding of gravitational density might have predicted the devastating floods ahead of time.

Flooding on the Mississippi River in May of 2011. Image Credit: Bill Borghoff

Flooding on the Mississippi River in May of 2011. Image Credit: Bill Borghoff

‰ÛÏI actually didn’t know that the Earth’s gravitational field is different in different places before I started working on this mission,‰Û Reager said. He started his career with a master’s degree in physical oceanography, where he learned another important fact: He has a predisposition toward terrible seasickness.

Even off the boat, the localized data analysis his oceanography program featured didn’t seem like the right path to follow.

‰ÛÏI wanted to answer the big questions,‰Û Reager said. ‰ÛÏI wanted to understand how the planet worked.‰Û He decided against pursuing physical oceanography any further, but didn’t stray far. ‰ÛÏWater is a key component of (how the planet works). And I knew that water is a scarce resource in some places.‰Û

Thinking about how climate change might affect water resources and certain populations, Reager knew he could help contribute something to the world if he devoted his time to studying water patterns and circulation. Because water is one of the heaviest things on Earth that can freely circulate and affect the global mass distribution, observing water resources using GRACE fit right into this ideal.

According to Bernknopf, GRACE data can be used to help analyze investment risks in flood mitigation. Looking at the flood potential of an area can inform the way that experts determine how much needs to be invested in avoiding a threshold of damage that would render a community unsustainable following a disaster, and therefore helps to avoid over- or under-investment in the resources necessary to mitigate flood damage and protect human life.

Some GRACE data is already used for drought monitoring, and Reager hopes that it can be integrated into national flood prediction systems in the future.

Currently, Reager said, much of the flood prediction in the United States and the rest of the world is based on weather forecasts alone. ‰ÛÏWeather forecasters are looking at whether or not they’re going to have a big rainstorm coming in the next three to 10 days,‰Û he explained. ‰ÛÏAnd if we have a lot of rain coming then we look at historically when we have that much rain and say things like ‰Û÷Well, that much rain could cause floods.’‰Û

Assimilating GRACE data into higher-resolution hydrology models could help forecasters diminish uncertainty and boost understanding of flood potential, allowing them to predict catastrophic floods months ahead of time.

Getting that data into the hands of decision-makers is a complicated negotiation, however. Incorporating GRACE into existing hydrology models means a lot of time, effort and new training. In addition, Bernknopf said that because GRACE data operates as a statistical model, all ‰ÛÏGRACE-based forecasts will contain uncertainty.‰Û The limit of acceptable uncertainty has yet to be determined, however, and Bernknopf explained that users need to determine it for themselves.

Another question that complicates implementation is the nature of forecasts, according to Bernknopf. Forecasters have to take into account the length of time that would make a warning based on GRACE data socioeconomically valuable. ‰ÛÏHow much of a benefit is a 1-week or 90-day warning?‰Û Bernknopf asked as a hypothetical consideration. ‰ÛÏAnd how would benefits be measured as the result of economic behavior in response to the warning?‰Û

There are concerns about whether having advance notice about warnings is beneficial on an individual level as well as an economic one: Does increasing lead time really mean that people will be more prepared? Reading a weather alert too early could lead to overhyping or unnecessary stress. However, having environmental data in the face of a catastrophic flood or similar harsh weather event is crucial to protecting people in that event’s path. It is still important to get those reports out to the general public so that they understand the risks of natural disaster, and the specifics of dissemination can be left to operational analysts.

Reager and the GRACE team are currently in talks with the National Oceanic and Atmospheric Administration and the National Weather Service to see if there’s a way to keep their data in use operationally. In the meantime, GRACE will continue to survey the globe and collect data for future use, keeping Reager and his team busy as they try to track the water patterns indicative of a changing planet.