IBM DeepCurrent Predicts Environmental Changes in 3-D

Share this post:

From ocean bays in Ireland to fresh water lakes in New York, IBM “smarter water” scientists have developed a way to forecast the health of bodies of water — in 3-D. IBM DeepCurrent laps up data from coastlines to underwater depths via sensors attached to everything from buoys to tagged fish. It creates 3-D models that can zoom in on a location or event to provide accurate predictions of environmental dynamics, such as salt runoff or invasive species. So, business owners, environmental agencies and public administrators can use it to improve their accuracy in marine monitoring to better manage and forecast for potentially destructive environmental events.

Emanuele Ragnoli

“IBM DeepCurrent is a geophysical fluid dynamics tool that focuses on fluid motion in the earth’s water systems. It simulates the flow of water and corresponding changes in its properties, such as variations in temperature, salinity and nutrient concentrations.”

“And it can also tell us about water flows, water quality, and other environmental parameters by using physical models, machine learning and control theory algorithms, combined with sensor data, in a particular coastal area,” said Emanuele Ragnoli, an IBM Environmental Analytics researcher in Dublin, Ireland.

Galway Bay is a popular coastal area on the west coast of Ireland that’s been eulogized in movies to music, but it’s under threat of pollution, flooding, and other environmental factors. Emanuele’s team is applying IBM DeepCurrent prediction and modelling to study the live dynamics of oceanic flows into the bay, detecting the oceanic flow’s interaction with freshwater discharges, integrating radar imagery of surface currents, and even monitoring the wind-driven currents to offer insights into the ecological planning and economic management of the bay. Now, scientists can see, in a 3-D model, how live weather data will impact everything from tributaries into the bay, to how currents distribute marine nutrients — or contaminants — across the bay. So far, IBM DeepCurrent deployment has shown a 10 to 20 percent error reduction of predicted flows within the bay, relative to past sensor measurements.

On the other side of the Atlantic, IBM DeepCurrent researchers examined the effects of Hurricane Sandy on Chesapeake Bay, which straddles the states of Virginia and Maryland. The team combined a disparate array of sensor data with model simulations to reproduce the changes of water elevation during the movement of this super storm, and mapped the resulting environmental effects it had on cities and regions around the bay. The deployment validated the performance of the system in extreme events. Now, the team hopes to apply IBM DeepCurrent to similar situations in real-time to assist city managers and urban planners with emergency planning and management.“The coastal ocean represents perhaps the most complex natural system, subject to a nonlinear and chaotic set of conditions that make predictions extremely challenging. The accurate quantification of flow processes in the upper layers of the ocean requires a holistic consideration of system-forcing variables and accurate field sampling data.”

“But this is exactly what IBM DeepCurrent does. It provides industry, local government and relevant stakeholders with real-time information to better manage the coastal environment and mitigate anthropogenic impacts,” said Fearghal O’Donncha, IBM Environmental Analytics researcher.

Fearghal O’Donncha

IBM DeepCurrent’s latest effort is its first in fresh water: the IBM Jefferson Project at New York’s Lake George. The “Queen of American Lakes” is threatened by salt, algae, and invasive marine species affecting the fish and vegetation. And IBM DeepCurrent’s role in the joint project with Rensselaer Polytechnic Institute and the FUND for Lake George is to help study the circulation of salt runoff from de-icing nearby roads in the winter.

In October, as part of the European Commission’s Frontier and Emerging Technologies initiative, the team from Ireland will use IBM DeepCurrent to model the Deepwater Horizon oil spill in the Gulf of Mexico. The work will not only help further validate IBM DeepCurrent’s predictive capabilities, but also help continue to account for oil still in the gulf five years after the deadly explosion.

More stories

A new supercomputing-powered weather model may ready us for Exascale

In the U.S. alone, extreme weather caused some 297 deaths and $53.5 billion in economic damage in 2016. Globally, natural disasters caused $175 billion in damage. It’s essential for governments, business and people to receive advance warning of wild weather in order to minimize its impact, yet today the information we get is limited. Current […]

Continue reading

DREAM Challenge results: Can machine learning help improve accuracy in breast cancer screening?

        Breast Cancer is the most common cancer in women. It is estimated that one out of eight women will be diagnosed with breast cancer in their lifetime. The good news is that 99 percent of women whose breast cancer was detected early (stage 1 or 0) survive beyond five years after […]

Continue reading

Computational Neuroscience

New Issue of the IBM Journal of Research and Development   Understanding the brain’s dynamics is of central importance to neuroscience. Our ability to observe, model, and infer from neuroscientific data the principles and mechanisms of brain dynamics determines our ability to understand the brain’s unusual cognitive and behavioral capabilities. Our guest editors, James Kozloski, […]

Continue reading