[House Hearing, 116 Congress]
[From the U.S. Government Publishing Office]
A TASK OF EPIC PROPORTIONS:
RECLAIMING U.S. LEADERSHIP
IN WEATHER MODELING AND PREDICTION
=======================================================================
HEARING
BEFORE THE
SUBCOMMITTEE ON ENVIRONMENT
OF THE
COMMITTEE ON SCIENCE, SPACE,
AND TECHNOLOGY
HOUSE OF REPRESENTATIVES
ONE HUNDRED SIXTEENTH CONGRESS
FIRST SESSION
__________
NOVEMBER 20, 2019
__________
Serial No. 116-57
__________
Printed for the use of the Committee on Science, Space, and Technology
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Available via the World Wide Web: http://science.house.gov
______
U.S. GOVERNMENT PUBLISHING OFFICE
38-332 PDF WASHINGTON : 2020
COMMITTEE ON SCIENCE, SPACE, AND TECHNOLOGY
HON. EDDIE BERNICE JOHNSON, Texas, Chairwoman
ZOE LOFGREN, California FRANK D. LUCAS, Oklahoma,
DANIEL LIPINSKI, Illinois Ranking Member
SUZANNE BONAMICI, Oregon MO BROOKS, Alabama
AMI BERA, California, BILL POSEY, Florida
Vice Chair RANDY WEBER, Texas
LIZZIE FLETCHER, Texas BRIAN BABIN, Texas
HALEY STEVENS, Michigan ANDY BIGGS, Arizona
KENDRA HORN, Oklahoma ROGER MARSHALL, Kansas
MIKIE SHERRILL, New Jersey RALPH NORMAN, South Carolina
BRAD SHERMAN, California MICHAEL CLOUD, Texas
STEVE COHEN, Tennessee TROY BALDERSON, Ohio
JERRY McNERNEY, California PETE OLSON, Texas
ED PERLMUTTER, Colorado ANTHONY GONZALEZ, Ohio
PAUL TONKO, New York MICHAEL WALTZ, Florida
BILL FOSTER, Illinois JIM BAIRD, Indiana
DON BEYER, Virginia JAIME HERRERA BEUTLER, Washington
CHARLIE CRIST, Florida FRANCIS ROONEY, Florida
SEAN CASTEN, Illinois GREGORY F. MURPHY, North Carolina
BEN McADAMS, Utah
JENNIFER WEXTON, Virginia
CONOR LAMB, Pennsylvania
VACANCY
------
Subcommittee on Environment
HON. LIZZIE FLETCHER, Texas, Chairwoman
SUZANNE BONAMICI, Oregon ROGER MARSHALL, Kansas, Ranking
CONOR LAMB, Pennsylvania Member
PAUL TONKO, New York BRIAN BABIN, Texas
CHARLIE CRIST, Florida ANTHONY GONZALEZ, Ohio
SEAN CASTEN, Illinois JIM BAIRD, Indiana
BEN McADAMS, Utah FRANCIS ROONEY, Florida
DON BEYER, Virginia GREGORY F. MURPHY, North Carolina
C O N T E N T S
November 20, 2019
Page
Hearing Charter.................................................. 2
Opening Statements
Statement by Representative Sean Casten, Presiding Chairman,
Subcommittee on Environment, Committee on Science, Space, and
Technology, U.S. House of Representatives...................... 7
Statement by Representative Roger Marshall, Ranking Member,
Subcommittee on Environment, Committee on Science, Space, and
Technology, U.S. House of Representatives...................... 7
Written Statement............................................ 8
Written statement by Representative Lizzie Fletcher, Chairwoman,
Subcommittee on Environment, Committee on Science, Space, and
Technology, U.S. House of Representatives...................... 9
Written statement by Representative Eddie Bernice Johnson,
Chairwoman, Committee on Science, Space, and Technology, U.S.
House of Representatives....................................... 10
Written statement by Representative Frank Lucas, Ranking Member,
Committee on Science, Space, and Technology, U.S. House of
Representatives................................................ 10
Witnesses:
Dr. Neil Jacobs, Assistant Secretary of Commerce for
Environmental Observation and Prediction, performing the duties
of Under Secretary of Commerce for Oceans and Atmosphere,
National Oceanic and Atmospheric Administration (NOAA)
Oral Statement............................................... 12
Written Statement............................................ 15
Dr. Cliff Mass, Professor of Atmospheric Sciences, University of
Washington
Oral Statement............................................... 19
Written Statement............................................ 21
Dr. Peter P. Neilley, IBM Distinguished Engineer and Director of
Weather Forecasting Sciences and Technologies, The Weather
Company, An IBM Business
Oral Statement............................................... 55
Written Statement............................................ 57
Dr. Thomas Auligne, Director of the Joint Center for Satellite
Data Assimilation, University Corporation for Atmospheric
Research (UCAR)
Oral Statement............................................... 65
Written Statement............................................ 67
Discussion....................................................... 79
Appendix I: Answers to Post-Hearing Questions
Dr. Neil Jacobs, Assistant Secretary of Commerce for
Environmental Observation and Prediction, performing the duties
of Under Secretary of Commerce for Oceans and Atmosphere,
National Oceanic and Atmospheric Administration (NOAA)......... 90
Dr. Cliff Mass, Professor of Atmospheric Sciences, University of
Washington..................................................... 96
Dr. Peter P. Neilley, IBM Distinguished Engineer and Director of
Weather Forecasting Sciences and Technologies, The Weather
Company, An IBM Business....................................... 101
Dr. Thomas Auligne, Director of the Joint Center for Satellite
Data Assimilation, University Corporation for Atmospheric
Research (UCAR)................................................ 107
Appendix II: Additional Material for the Record
Letter submitted by Representative Sean Casten, Presiding
Chairman, Subcommittee on Environment, Committee on Science,
Space, and Technology, U.S. House of Representatives........... 112
A TASK OF EPIC PROPORTIONS:
RECLAIMING U.S. LEADERSHIP
IN WEATHER MODELING AND PREDICTION
----------
WEDNESDAY, NOVEMBER 20, 2019
House of Representatives,
Subcommittee on Environment,
Committee on Science, Space, and Technology,
Washington, D.C.
The Subcommittee met, pursuant to notice, at 1:58 p.m., in
room 2318 of the Rayburn House Office Building, Hon. Lizzie
Fletcher [Chairwoman of the Subcommittee] presiding.
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Mr. Casten [presiding]. This hearing will come to order.
Without objection, the Chair is authorized to declare recess at
any time. As some of you know, due to caucus votes at 3 p.m.,
I'm going to try to keep my introductory remarks brief, and ask
to submit the rest for the record, in order to get witness
testimonies in as quickly as possible. The Chairwoman and
Ranking Member of the full Committee have also agreed to submit
their statements for the record.
As we have previously discussed in this Subcommittee.
Americans depend on the data and services provided by the
National Oceanic and Atmospheric Administration, or NOAA, and
the National Weather Service every day. Earlier this Congress,
in this Subcommittee's hearing on the NOAA Fiscal Year 2020
proposed budget, we heard from Acting Administrator Dr. Jacobs
that the U.S. is not currently the global leader in weather
forecasting. Considering how important accurate weather
forecasting is to all Americans, this is extremely concerning.
The National Integrated Drought Information System
Reauthorization Act, NIDIS,which was signed into law in January
2019, directed NOAA to establish the Earth Prediction
Innovation Center, or EPIC. EPIC is tasked with creating a
collaborative, community-driven, global weather research
modeling system. The system will be publicly accessible,
allowing those outside of NOAA to access and contribute to a
community developed model. At today's hearing I look forward to
a discussion with our distinguished panel of experts about how
EPIC will leverage the skills and expertise across the public,
private, and academic sectors of the United States weather
community to bolster modeling and forecasting. Since EPIC is
still in its infancy, this hearing will provide a timely
opportunity to discuss the future of its organization,
management, and governance, and examine each sector's vision
and short- and long-term goals for EPIC.
I cannot overstate the importance of improving U.S. weather
modeling and prediction capabilities. EPIC represents what some
experts in the weather community have claimed as America's last
chance to get this right, and restore our leadership in global
weather prediction. I look forward to today's discussion about
how EPIC is going to accomplish this. Thank you.
The Chair now recognizes Ranking Member Marshall for an
opening statement.
Mr. Marshall. Thank you for holding this hearing. I want to
thank our witnesses for appearing before the Subcommittee,
especially Dr. Jacobs, who is in front of all of us now for the
third time this year in Congress. And thanks for all of you on
the panel for sharing your perspectives.
Weather prediction is something that affects the
constituents of every Member up here, from the fields of Kansas
to the Outer Banks of North Carolina. Anticipating the strength
and conditions of the next weather event can save lives and
property, as well as be the difference between a profitable
year for a farmer or a catastrophic loss. I'm proud to say the
Science Committee acted decisively last Congress by passing the
Weather Research and Forecasting Innovation Act, the Weather
Act, and the National Integrated Drought Information System
Reauthorization Act. The Weather Act was the first authorizing
legislation to address weather forecasting in 25 years, and
prioritized improving weather data, modeling, computing, and
forecasting. I'd like to extend to my gratitude to Ranking
Member Lucas for introducing what is now a law, and for his
continued leadership on this issue.
The NIDIS Reauthorization Act established the Earth
Predication Innovation Center, EPIC, the topic of our hearing
today. EPIC, when completed, will crowdsource the expertise of
the private sector and the research communities to improve our
forecasting models. This aligns with Congress' vision for the
program by leveraging the weather enterprise to provide
knowledge and skill on numerical weather prediction. The
Federal Government should be doing more to utilize resources of
private companies and university researchers, who are often the
leading sources of innovations. In addition to having world
class facilities and minds, private companies and academics are
extremely flexible in research and development and cost
effective in their methods. It is in the best interest of
Kansas farmers, ranchers, emergency personnel, and everyday
residents to have more accurate forecasts, and EPIC is an
important step in the improvement of our forecasting ability.
In 2012 Hurricane Sandy caused nearly $70 billion in damage
as it made landfall in Cuba and the Northeast Coast of the
United States. This was the catalyzing weather event which
caused Congress to examine how we could improve weather
forecasting. We don't know when the next superstorm will be,
but it's my hope that, through EPIC, NOAA and the National
Weather Service will be fully prepared to predict, respond, and
recover from the next severe weather event. While NOAA has
taken the initial steps to implement EPIC, we must see a
stronger sense of urgency moving forward. Because it's designed
as a community approach to weather prediction and modeling. I
look forward to hearing how Dr. Mass and others have been
involved in implementing the center, and getting their
feedbacks on how to ensure a successful and timely completion.
Thank you, Mr. Chairman, and I yield back.
[The prepared statement of Mr. Marshall follows:]
Thank you for holding this hearing, Chairwoman Fletcher. I
want to thank our witnesses for appearing before the
subcommittee, especially Dr. Jacobs who is in front of the
Committee for the third time this Congress, and all of you on
the panel for sharing your perspectives.
Weather prediction is something that affects the
constituents of every Member up here. From the fields of Kansas
to the Outer Banks of North Carolina, anticipating the strength
and conditions of the next weather event can save lives and
property.
I'm proud to say the Science Committee acted decisively
last Congress by passing the Weather Research and Forecasting
Innovation Act (the Weather Act) and the National Integrated
Drought Information System (NIDIS) Reauthorization Act.
The Weather Act was the first authorizing legislation to
address weather forecasting in 25 years and prioritized
improving weather data, modeling, computing, and forecasting.
I'd like to extend my gratitude to Ranking Member Lucas for
introducing what is now a law and for his continued leadership
on this issue.
The NIDIS Reauthorization Act established the Earth
Prediction Innovation Center (EPIC), the topic of our hearing
today. EPIC, when completed, will crowdsource the expertise of
the private sector and the research communities to improve our
forecasting models. This aligns with Congress' vision for the
program by leveraging the weather enterprise to provide
knowledge and skill on numerical weather prediction.
The Federal Government should be doing more to utilize the
resources of private companies and university researchers, who
are often the leading sources of innovations. In addition to
having world-class facilities and minds, private companies and
academics are extremely flexible in research development and
cost-effective in their methods.
It is in the best interest of Kansan farmers, ranchers,
emergency personnel, and every day residents to have more
accurate forecasts. And EPIC is an important step in the
improvement of our forecasting ability.
In 2012, Hurricane Sandy caused nearly $70 billion in
damage as it made landfall in Cuba and the Northeast coast of
the United States. This was the catalyzing weather event which
caused Congress to examine how we could improve weather
forecasting. We don't know when the next "superstorm" will be,
but it is my hope that through EPIC, NOAA and the National
Weather Service will be fully prepared to predict, respond, and
recovery from the next severe weather event.
While NOAA has taken the initial steps to implement EPIC,
we must see a stronger sense of urgency moving forward. Because
it is designed as a community approach to weather prediction
and modeling, I look forward to hearing how Dr. Mass and others
have been involved in implementing this center and getting
their feedback on how to ensure a successful and timely
completion.
Thank you, Madam Chair. I yield back.
Mr. Casten. If there are Members who wish to submit
additional opening statements, your statements will be added to
the record at this point.
[The prepared statement of Chairwoman Fletcher follows:]
Good afternoon, and welcome to the Subcommittee on
Environment's hearing entitled "A Task of EPIC Proportions:
Reclaiming U.S. Leadership in Weather Modeling and Prediction."
I would like to thank all of our witnesses for being here today
to discuss the current state and future of the Earth Prediction
Innovation Center, or EPIC, and its role in improving U.S.
weather forecasting capabilities.
As we've previously discussed in this Subcommittee,
Americans depend on the data and services provided by the
National Oceanic and Atmospheric Administration (NOAA) and the
National Weather Service every day. Much of these data are
utilized in the weather products offered by private companies,
such as weather apps on our cell phones or local news
forecasts. Earlier this Congress, in this Subcommittee's
hearing on the NOAA Fiscal Year 2020 Proposed Budget, we heard
from Dr. Jacobs that the U.S. is not currently the global
leader in weather forecasting. Considering how important
weather forecasting is to all Americans, this is extremely
concerning.
A devastating display of this was in 2012, when the U.S.
model failed to predict Hurricane Sandy's sharp left turn and
landfall over the East Coast. The European model got it right,
demonstrating to the nation that U.S. weather forecasting
abilities were far behind those of Europe. As we've discussed
in this Committee, severe storms like Sandy are increasing in
frequency and intensity due to climate change, making accurate
forecasts even more critical.
A major difference between the U.S. and the European
systems is that in Europe, the entire weather community
contributes to a single model. In the U.S., the public,
private, and academic sectors operate in isolation from each
other, each working on their own weather prediction research
and contributing to their own models. Even within the federal
government, multiple agencies work on their own models in an
uncoordinated way, and resources and expertise are fragmented.
As a result, the U.S. Air Force abandoned the U.S. global
weather model in 2015, preferring the United Kingdom's Unified
Model. It is of the utmost importance that the U.S. weather
community immediately act to catch up with its European
counterpart.
Congress recognized the need to better leverage the skills
and expertise across the public, private, and academic sectors
of the U.S. weather community to create a single global model
that is stronger than any of the individual models. The
National Integrated Drought Information System Reauthorization
Act, which was signed into law in January 2019, directed NOAA
to establish the Earth Prediction Innovation Center, or EPIC.
EPIC is tasked with creating a collaborative, community-driven
global weather research modeling system. The system will be
publicly accessible, allowing those outside of NOAA to access
and contribute to a community-developed model.
On top of improvements to global weather prediction, EPIC
could also serve as a vehicle to improve other, specialized
modeling systems, such as rainfall and flooding prediction.
This has implications for places like my district, Texas's 7th
Congressional District in Houston, that has been experiencing
increasingly frequent and intense precipitation events in
recent years. Leveraging the capabilities of the community to
improve precipitation modeling could provide my constituents,
and others who live in flood-prone areas, more precise
information about the timing and intensity of forecasted
rainfall, thus protecting lives and property. I know all of our
constituents look to the Weather Service as the national
authority in issuing life-saving forecasts, watches, and
warnings. While EPIC is intended to leverage the expertise of
the non-federal weather community, the provision of official
watches, warnings, and forecasts should remain with the
National Weather Service.
At today's hearing, I look forward to a discussion with our
distinguished panel of experts from across the U.S. weather
community about how EPIC will combine each sector's expertise
to bolster U.S. modeling. Since EPIC is still in its infancy,
this hearing will provide a timely opportunity to discuss the
future of its organization, management, and governance and
examine each sector's vision and short and long-term goals for
EPIC.
I cannot overstate the importance of improving U.S.
modeling and prediction capabilities. EPIC represents what some
experts in the weather community have called America's last
chance to get this right and reclaim our leadership in global
weather prediction. I look forward to today's discussion about
how EPIC is going to accomplish this.
Thank you.
[The prepared statement of Chairwoman Johnson follows:]
Thank you, Chair Fletcher.
We have had many discussions this Congress, and Congresses
in the past, about the importance of accurate and timely
weather forecasts.
Weather forecasting is complex and relies on first
collecting as many observations and data as possible that are
then assimilated into cutting edge weather models that are
tested and verified. NOAA, the lead civilian agency for
operational weather forecasting, participates in all aspects of
this process, including the development of our weather models.
Despite being at the forefront of the development of numerical
weather prediction, the accuracy of U.S. forecasts and
numerical weather prediction has fallen behind that of other
countries. But this isn't just a matter of pride; accurate
weather forecasts save lives and protect property.
We recently had a devastating tornado touch down in Dallas
that ripped through densely populated areas of the Metroplex in
and near my district. Fortunately, there were no deaths or
severe injuries related to this outbreak, but the tornadoes did
cause an estimated $2 billion in property damage.
Timely forecasts, watches, and warnings from the National
Weather Service were instrumental to keeping Texans safe during
this tornado outbreak. Thank you to Dr. Jacobs and the
dedicated employees at NOAA and the National Weather Service
for their great work in protecting Americans every day.
As Texans, Chair Fletcher and I are very familiar with
extreme weather events, as are Ranking Members Lucas and
Marshall. This Committee held a hearing earlier this year on
how to improve the understanding and forecasting of extreme
weather events in a changing climate. Many of the witnesses at
that hearing shared that leveraging the capabilities and
resources of our robust weather enterprise through a community
approach would be critical to addressing extreme weather
forecasting challenges. NOAA's Earth Prediction Innovation
Center, or EPIC, has thepotential to support the goal of
regaining U.S. leadership in global weather forecasting through
a community driven approach.
The Weather Research and Forecasting and Innovation Act of
2017 was a significant step toward improving weather
forecasting. This was followed by the National Integrated
Drought Information System, or NIDIS Reauthorization Act of
2018 that amended the Weather Act and authorized EPIC at NOAA.
It is vital for Congress to conduct oversight of federal
programs that we have authorized to ensure they are being
implemented as Congress intended. Hearings like this are
important if we are to be good stewards of taxpayer dollars. I
am looking forward to hearing from a broad group of
stakeholders from the weather community this afternoon on how
we can leverage a program like EPIC to achieve a common goal of
improving our weather forecasts to better protect our
constituents.Thank you and I yield back.
[The prepared statement of Mr. Lucas follows:]
Thank you, Chairwoman Fletcher, for holding today's
hearing. I've said before that the continued improvement of
weather forecasting is one of the most important topics in this
committee's jurisdiction. Accurate forecasting not only helps
our businesses make strategic plans, but it helps us to protect
lives and properties during severe weather events. We need an
accurate and trustworthy system.
The United States was once the world's leader in numerical
weather prediction, but we can't credibly make that claim
today. This was apparent in 2012, when American forecasts
predicted Hurricane Sandy would weaken over the Atlantic, while
the European forecast model correctly saw Sandy making
landfall.
Congress saw the need for rapid improvement in U.S. weather
forecasts. In the supplemental appropriations package passed in
response to Sandy in early 2013, Congress provided more than
$20 million to NOAA to help improve forecast modeling and
computing resource needs. While this assistance resulted in
some improvements to our forecasting abilities, we needed to do
more.
This committee passed the Weather Act during the 115th
Congress, which was signed into law in April 2017. The Weather
Act, the most significant weather legislation passed by
Congress in more than 25 years, provided authorities and
direction for NOAA in its weather research and forecasting
efforts. One of the most consequential provisions in the bill
was direction for NOAA to begin purchasing more commercial data
in creating forecasts. This came in response to a recognized
need for NOAA to better utilize the knowledge and expertise of
the private sector and the research community.
An extension of the Weather Act was signed into law earlier
this year. Included in this legislation was an authorization of
the Earth Prediction Innovation Center - known as EPIC. This
center represents a new way of weather modeling for NOAA by
utilizing the computing resources and expertise of the academic
community, private enterprise, and others who want to help the
U.S. regain leadership. It will also utilize new computing
resources, a significant reason why the U.S. has lagged in its
forecasting abilities.
The authorizing legislation for EPIC became law in January.
While NOAA has taken initial steps to implement EPIC, progress
has been slow. We must move forward quickly to implement this
legislation and begin closing the gap with the Europeans,
Canadians, and others who have surpassed us. Our panel of
witnesses will help us identify potential bottlenecks in
implementing EPIC and what we can do to help the process move
forward quickly.
Dr. Neil Jacobs is no stranger to our committee, and I want
to thank him for again appearing before us today. He has made
the quick and effective implementation of EPIC a personal
priority. His education and professional background will be
invaluable as we continue to improve the accuracy of our
weather forecasts and I look forward to working with him on
this effort.
I again want to thank Chairwoman Fletcher for conducting
today's hearing and I also want to thank Chairwoman Johnson for
her shared commitment to helping the U.S. again be the world
leader in weather forecasting.
Thank you and I yield back.
Mr. Casten. At this time I would like to introduce our
witnesses.
Our first witness is Dr. Neil Jacobs. He is the Assistant
Secretary of Commerce for Environmental Observation and
Prediction, performing the duties of Under Secretary of
Commerce for Oceans and Atmosphere. Prior to joining NOAA, Dr.
Jacobs was the Chief Atmospheric Scientist at Panasonic
Avionics Corporation. He was also the Chair of the American
Meteorological Society's Forecast Improvement Group, and served
on the World Meteorological Organization's aircraft-based
observing team. Dr. Jacobs has a master's and doctoral Degree
in atmospheric science from North Carolina State University.
Our second witness, Dr. Cliff Mass, is a Professor of
Atmospheric Sciences at the University of Washington. His
specialty is numerical weather and climate prediction, and the
meteorology of the western United States. Previously Dr. Mass
was a faculty member at the University of Maryland's
Meteorology Department. Dr. Mass is a Fellow of the American
Meteorological Society, a member of the Washington State
Academy of Sciences, and has published over 120 papers. Dr.
Mass received his Ph.D. in atmospheric sciences from the
University of Washington. Welcome.
Our third witness, Dr. Peter Neilley, is an IBM
Distinguished Engineer, and Director of Weather Forecasting
Sciences and Technologies for The Weather Company. He
specializes in developing state-of-the-science technologies in
weather forecasting for public use and weather-dependent
markets. Dr. Neilley worked as a scientist at the National
Center for Atmospheric Research, and the Chief Scientist at
Weather Services International Corporation. Dr. Neilley
recently served on NOAA's Science Advisory Board's
Environmental Information Services Working Group. He was a
longtime member and Chair of the American Meteorological
Society's Committee on Weather and Forecasting. Dr. Neilley
holds a master's degree and a Ph.D. in meteorology from MIT.
Welcome.
Our final witness, Dr. Thomas Auligne, thank you, is the
Director of the Joint Center for Satellite Data Assimilation, a
research center based on a multi-agency partnership between
NOAA, NASA (National Aeronautics and Space Administration), the
U.S. Navy, and Air Force. He is responsible for the mission to
accelerate and improve the quantitative use of satellite data
in weather, ocean, climate, and environmental analysis and
prediction systems. Dr. Auligne has held research positions at
the National Center for Atmospheric Research, the European
Center for Medium-Range Weather Forecasting, and Meteo-France.
Dr. Auligne earned a master's in meteorology and a Ph.D. in
atmospheric physics in France.
As our witnesses should know, you will each have 5 minutes
for your spoken testimony. Your written testimony will be
included in the record for the hearing. When you all have
completed your spoken testimony, we will begin with questions.
Each Member will have 5 minutes to question the panel. We will
start with Dr. Jacobs.
TESTIMONY OF DR. NEIL JACOBS,
ASSISTANT SECRETARY OF COMMERCE FOR
ENVIRONMENTAL OBSERVATION AND PREDICTION,
PERFORMING THE DUTIES OF UNDER SECRETARY OF
COMMERCE FOR OCEANS AND ATMOSPHERE, NOAA
Dr. Jacobs. Good afternoon, Chairman Casten, Ranking Member
Marshall, and Ranking Member Lucas. Thank you for the
opportunity to testify at this hearing. NOAA is entrusted with
the responsibility to provide environmental information and
prediction to the public to enable informed decisions on a
range of phenomenon spanning a broad spectrum of temporal and
spatial scales. Part of NOAA's core mission is to protect lives
and property, and to safeguard the national economy. With such
an important task, it is imperative that NOAA provide accurate
and timely weather information. We strive to produce the best
weather forecast in the world, underpinned by cutting edge
research, collaborative external partnerships, and thousands of
dedicated scientists.
Following Hurricane Sandy, Congress provided supplemental
funding for NOAA to take the first large step toward increasing
computing capacity and improving its global forecast models.
The desire to improve NOAA's weather mission culminated in
congressional interest, and the passage of the Weather Research
and Forecasting Innovation Act of 2017. This groundbreaking
legislation contains a number of important directives for NOAA,
including focusing transitioning research to operations, sub-
seasonal and seasonal weather forecast improvement, and
satellite data innovation. Since coming to NOAA, implementing
the Weather Act has been my top priority.
One section in the Weather Act I would like to draw
attention to is the mandate to make NOAA's numerical weather
prediction code publicly available. While NOAA complied with
this directive in spirit, it has been unable to fully implement
it. The existing version of the code is unique to NOAA
computers. This means that, while the public would have access
to the code, without access to NOAA's internal computers, they
would not be able to actually run the model.
To solve this problem, NOAA needs a strategy to allow for
greater accessibility by the public. To achieve this NOAA, will
need to port its weather model code to commercial cloud, where
it can be hosted by one or more providers. Making NOAA's model
code available to the public will allow external world class
scientists and researchers the opportunity to collaborate on
new improvements, and this is a new way of thinking. Instead of
keeping research and development inside of NOAA, the entire
weather enterprise will be able to work with us to improve our
modeling system, thereby accelerating advancements to our
mission of protecting life and property. This strategy is the
core principle of NOAA's new Earth Prediction Innovation
Center.
Building on the tenets of the Weather Act, and recently
authorized in the National Integrated Drought Information
System Reauthorization Act of 2018, EPIC will serve as the core
research to operations to research hub for building and
maintaining a community modeling framework. EPIC's innovative
structure will link scientists and software engineers in
academia, the private sector, and partner agencies with
research, development, and operational activities inside the
agency. Doing so will help accelerate model improvements,
enhancing NOAA's ability to provide accurate warnings of
weather-based threats, and helping to re-establish the U.S.
preeminence in numerical weather prediction.
Once integrated into the infrastructure of NOAA, EPIC will
be used with the Unified Forecast System to improve the
forecast skill of NOAA and other modeling initiatives, such as
climate and ocean models. EPIC's public accessibility through
highly scalable commercial cloud-based HPC (high-performance
computing) architecture will enable external research partners
to develop, test, and provide feedback on the American modeling
system. Structured as a virtual center, EPIC will also manage
model evaluation, source code, and user training. Where
appropriate, NOAA will look to partner with other Federal
agencies and academia to further this initiative.
The President's Fiscal Year 2020 budget proposed $15
million for EPIC. NOAA recognizes that importance of the EPIC
program and has already started implementing several steps to
plan for its future. Last month NOAA held an industry day to
engage outside collaborators, ranging from universities to
cloud vendors. NOAA has also issued a request for information
on governance structure of the program itself, and has
conducted extensive market research with external stakeholders.
With adequate funding, NOAA looks forward to issuing a request
for proposals, and moving forward with this critical program.
Chairman Casten, Ranking Member Marshall, Ranking Member
Lucas, and Members of the Subcommittee, thank you again for
inviting me to participate today, and I would be pleased to
answer any questions you may have.
[The prepared statement of Dr. Jacobs follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Mr. Casten. Thank you. Dr. Mass, you're recognized.
TESTIMONY OF DR. CLIFFORD MASS,
PROFESSOR OF ATMOSPHERIC SCIENCES,
UNIVERSITY OF WASHINGTON
Dr. Mass. Thank you. Good afternoon, Chairman Casten,
Ranking Member Marshall, Members of the Subcommittee. My name
is Cliff Mass, and I am a Professor of Atmospheric Sciences at
the University of Washington. The U.S. is behind in numerical
weather prediction, and we are not catching up. NOAA's global
model is either third or fourth in skill, behind the European
Center, the U.K. Met Office, and often the Canadian model. The
U.S. has the leading weather research community in the world,
and our Nation invests heavily in weather prediction. We should
be far ahead, consistent with the state of the science, but we
are not, and our global model is not the only problem. U.S.
weather prediction trails in other crucial aspects, including
high-resolution ensembles, and model post-processing.
In 2012 the Nation became aware of the problem during
Hurricane Sandy, and Congress responded with additional funds.
Seven years later objective numbers show that we are not
catching up, and the cost to the American people of the
stagnation is huge. State of the science forecasting will save
lives, greatly aid the U.S. economy, and serve as the first
line of defense for severe weather. So why is the U.S. failing
in this crucial arena? The causes are duplication of effort,
poor organization, and lack of leadership, plus insufficient
computer resources.
The enormous weather research resources of the United
States are spread over too many modeling systems. NOAA has
three groups working on such models, the Environmental Modeling
Center, and NOAA's ESRL (Earth System Research Laboratory) and
GFDL (Geophysical Fluid Dynamics Laboratory) labs. NASA and the
Navy have both developed both global and regional models. The
Air Force has acquired a foreign modeling system, and the
National Center for Atmospheric Research, which encompasses the
academic community, has developed another global modeling
system, in addition to the well-known WRF (Weather Research and
Forecasting) model.
The U.S. research community has mainly worked with NCAR's
(National Center for Atmospheric Research's) weather models,
and NOAA has used its own. They are not generally working
together, and thus NOAA has been cut off from the innovations
and energy of the U.S. academic community. Such a division of
effort has undermined U.S. weather prediction, resulting in a
large number of subcritical, inferior efforts. But there's
more. NOAA has been starved for computer resources, crippling
research and testing, and blocking the operational application
of promising approaches. My analysis, supported by colleagues
at NOAA, is that the National Weather Service could effectively
use 100 times its current computer allocation.
All of these problems can be turned around quickly if our
Nation reorganizes how we develop, test, and run numerical
weather prediction models. And the key to it all is bringing
resources and personnel together in one national effort. EPIC
can be a big part of the solution. EPIC must become the center
of U.S. model development and testing, and resources should be
concentrated there. It must be a physical center located
outside of NOAA, and serve all agencies and groups in the
Nation.
EPIC needs resources, independence, autonomy, stability,
and, most importantly, responsibility to deliver the best
modeling system in the world. It must be an exciting center of
discovery, science, and technology that will attract the best
scientists, and our best students. EPIC needs sufficient
computer resources for development and testing. It must entrain
the efforts and capabilities of the U.S. research community,
most importantly that of the National Center for Atmospheric
Research. Finally, EPIC must develop and support a national
community model that is freely available to the Nation. EPIC
can easily fail if it is not given primary responsibility and
resources to create the best weather prediction system in the
world. It will fail if its goals are too narrow, or designated
to serve a single agency.
Our nation was the first in numerical weather prediction,
but we threw away leadership by dividing our efforts. It is
time, through EPIC, to combine the national resources, and
rationalize how we develop forecast models with extraordinary
benefits to the American people. Thank you very much.
[The prepared statement of Dr. Mass follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Mr. Casten. Dr. Neilley, you are recognized for 5 minutes.
TESTIMONY OF DR. PETER P. NEILLEY,
IBM DISTINGUISHED ENGINEER AND DIRECTOR OF
WEATHER FORECASTING SCIENCES AND TECHNOLOGIES,
THE WEATHER COMPANY, AN IBM BUSINESS
Dr. Neilley. Thank you, Chairman Casten, and Ranking Member
Marshall, and Ranking Member Lucas, and all Members of the
Subcommittee, thank you for the opportunity to address the
Subcommittee today. My name's Peter Neilley, and I am the
Director of Weather Forecasting Sciences and Technologies at
the Weather Company, part of IBM. I oversee a team of
scientists and engineers that develop a broad suite of
technologies that are used to create and distribute weather
forecast products and services for both the U.S. and the rest
of the world. The U.S. is fortunate to have the most vibrant
weather enterprise anywhere in the world, with deep
partnerships between Federal, academic, and private sectors
creating the delivery services for the Nation. The Weather
Company and IBM are proud to be active contributing members to
that weather enterprise.
Numerical weather prediction, or NWP, is the foundational
technology used to create nearly all weather forecasts today.
At The Weather Company we employ many of our own numerical
weather prediction models, but are also heavily dependent on
the forecasts from numerical weather prediction models by NOAA
and others. Because of that dependency, The Weather Company has
deep interest in the quality of numerical weather prediction
forecasts produced by NOAA. Accordingly, I actively participate
in numerous advisory boards, committees that discuss and make
recommendations on means of improving numerical weather
prediction capabilities for the Nation, including the Earth
Prediction Innovation Center. The community workshop for the
Earth Prediction Innovation Center held this past summer
gathered nearly 300 stakeholders from inside and outside of the
Federal Government to inform NOAA and the community on early
ideas for implementing EPIC, and a summary report of those
findings, I believe, was released just yesterday by NOAA.
The U.S. has a rich history in developing and operating
numerical weather prediction systems that date back to the
1950s. Today numerical weather prediction capabilities are
developed and deployed in numerous places, including NOAA, the
Department of Defense, the Department of Energy, NASA, at
National Science Foundation-funded institutions, academic
organizations, and the private sector. Each corner of this
numerical weather prediction community has created modeling
capabilities tailored to the mission of each one of those
individual institutions. Unfortunately, there is no overarching
national strategy guiding the organization interaction of these
activities, which has led to less than optimum efficacy in any
one of them.
Further, there is no obvious existing place within the
enterprise where such a strategy might be formulated and
executed. As a result, there is a very broad set of NWP
capabilities across the Nation. Some of them good, but few of
them as good as they could be. In fact, when compared to models
developed by international counterparts in Europe and the U.K.,
our global numerical weather prediction systems from NOAA and
other members of our national enterprise are materially less
accurate, and have been for decades. As a result, our Nation is
significantly less prepared, and less resilient to the adverse
impacts of weather and climate than we could be.
Today I wish to convey four key points to the Subcommittee.
First, under the leadership of Acting Administrator Jacobs, and
his vision for EPIC, we have before us a generational
opportunity to address the shortcomings of our Nation's NWP
capabilities, and elevate them to the world's best.
Second, EPIC, as proposed by Dr. Jacobs, envisions the
creation of a state-of-the-science institution for the
community, and by the community, where the numerical weather
enterprise collaboratively works together using a common
framework of tools and technologies. This would enable the most
effective, and efficient, development of advanced numerical
weather prediction capabilities in support of both NOAA, and
all the other numerical weather prediction stakeholders across
the Nation.
Third, to execute this vision, NOAA must construct EPIC as
a semi-autonomous and externally managed national institution
that will establish, catalyze, organize, and manage a large and
diverse scientific and technical community collaboratively
working toward the betterment of NWP. NOAA would be a major
constituent in the EPIC community, participating in both its
government and scientific endeavors, but would only be one of
the many stakeholders drawing value from its accomplishments.
Finally, in order to achieve this vision, numerical weather
prediction stakeholders across the Federal Government must
embrace and actively participate in EPIC. As outlined in my
written testimony, I think Congress plays an important role in
fostering and incentivizing such participation, as well as
helping to ensure the long-term success of EPIC. Thank you for
the opportunity to address the Subcommittee today, and I also
look forward to your questions.
[The prepared statement of Dr. Neilley follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Mr. Casten. Dr. Auligne, you are recognized for 5 minutes.
TESTIMONY OF DR. THOMAS AULIGNE,
DIRECTOR OF THE JOINT CENTER FOR SATELLITE DATA
ASSIMILATION, UNIVERSITY CORPORATION FOR ATMOSPHERIC RESEARCH
(UCAR)
Dr. Auligne. Good afternoon, Chairman Casten, Ranking
Member Marshall, Ranking Member Lucas, and Members of the
Subcommittee. Thank you for the opportunity to testify today. I
am Dr. Thomas Auligne, Director of the Joint Center for
Satellite Data Assimilation at the University Corporation for
Atmospheric Research. As a trained meteorologist, I care deeply
about improving the quality of our weather models, which help
build a weather-ready nation, and save lives and property. My
experience in academia and operational centers gives me a
unique perspective on the so-called valley of death separating
research and operations.
For more than 30 years, weather prediction in the United
States has been trailing behind other international centers,
most notably the European Center. Previous actions and
additional funding have failed in regaining U.S. leadership.
This leads me to propose a disruptive vision for EPIC,
reconsidering organizational roles, governance, and funding
models. My view is that only with radical change is it
realistic to expect radical improvement.
Drawing from my previous experience at the European Center,
I have concluded that the secret sauce fueling their success
story has the following ingredients: Focus, innovation,
excellence, and accountability. While the U.S. weather
enterprise is often described as the uncoordinated giant,
plagued by fragmentation of resources, the Europeans rally
behind the strength of a common goal. The success of EPIC lies
in a clear, non-overlapping mission, with clear responsibility
and accountability.
EPIC should launch a focused effort with one goal, develop
the best weather prediction system for the Nation. Success
should be directly measured, and EPIC's director should be held
accountable. We need a center of excellence, attracting the
best talents that can drive the Center's goals, guide the
community, and work toward operational requirements. This dream
team will be supported by lean decisionmaking, efficiency-
driven operations, and strategic allocations of resources. On
this aspect, we need massive investment in high-performance
computing, leveraging the elasticity of the cloud.
EPIC should provide a collaborative environment, where
scientists from the government, academia, and the private
sector can gather to conduct innovative code development, and
explore high-risk, high-reward research. This requires building
a research model accessible by the entire community, and paired
with an effective process to transfer research to operations.
As EPIC focuses on encouraging and incorporating innovative
science, it should also utilize an innovative business model. I
am convinced that EPIC can draw from the success of the Joint
Center for Satellite Data Assimilation. Its distributed
structure, following a hub and spokes approach, increases the
government's ability to engage world class scientists and
engineers. Its agile team, at the intersection of multiple
Federal agencies, is reinventing collaboration, and exploring
innovative pathways. In fact, the Joint Center is already
applying the European secret sauce to better assimilate
observations to initialize model forecasts. This major science
problem is the highest priority for EPIC.
I dream of EPIC as an agile center, where scientists can
focus on science, red tape is reduced to a minimum,
decisionmaking is streamlined, and community collaboration is
entirely result-driven. The implementation of EPIC should be
delegated to a single trusted partner that has strong
connections to the community and the government, building a
bridge across the valley of death.
In conclusion, EPIC represents a unique opportunity. We
have one shot to get it right, and business as usual is not an
option. We need to reinvent the way we transition weather
research to operations. The breadth and depth of the U.S.
research community is second to none. EPIC can use is ingenuity
to reach, and even surpass, forecast improvement goals, and
collectively reclaim American leadership in weather modeling
and prediction.
Thank you for your attention. I look forward to answering
any question you have.
[The prepared statement of Dr. Auligne follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Mr. Casten. Thank you. At this point we will begin our
first round of questions. The Chair will recognize himself for
5 minutes.
I want to start by thanking you, Dr. Jacobs, for a really
productive meeting we had in my office last week on NOAA, and
the future of weather forecasting, the role of science, the
Enlightenment. We had a wide-ranging conversation, I appreciate
it. And I really just want to emphasize again how much I
appreciate you taking the time, as we discuss in greater depth
today the EPIC program, and the future of U.S. weather
modeling. And I also want to make sure that we keep in mind the
great work of the National Weather Service forecasters, and
their efforts to ensure that communities receive timely and
accurate forecasts of major weather events 24/7/365.
To that end, Dr. Jacobs, I will again echo my concerns
about NOAA's FY20 budget request to eliminate 110 full-time
equivalents, and I just want to reiterate the ask I made in our
meeting, and my subsequent letter, which I will use my power as
Chair, with unanimous consent, to enter into the record. I just
would ask that you follow up with the Committee on the issues
raised in that letter. Thank you.
Dr. Jacobs, in your testimony you discussed how, once
integrated into the infrastructure of NOAA, that EPIC will be
used within the Unified Forecast System to improve the forecast
skill of NOAA's other modeling initiatives, including climate
and ocean models, which, as a scientist myself, I geek out on
how that would all work, and the idea of actually having a
model that can both research and model many types of events,
hydrological changes, sea level rise, fisheries, and harmful
red tide. Can you help us understand, what is your timeline
under which EPIC can improve these other economically and
ecologically important forecasts, and especially with tying in
near-term weather to crucial longer-term climate models?
Dr. Jacobs. Sure. Thank you very much for the question.
Very much appreciate the time we spent together, and appreciate
your interest in NOAA and numerical weather prediction. The
Unified Forecasting System is a way to sort of streamline our
production suite. Inside of the National Center for
Environmental Prediction, we run a lot of different models,
from high-resolution short-range convection, to dynamic climate
models, medium-range models, and then we have wave models,
ocean models, hydrological, biological, ecological models.
We're trying to get all of these in a unified system, and this
sort of hinges on the NCAR/NOAA MOA (memorandum of agreement),
where we were looking at a common code base, and a common
infrastructure. The Finite Volume Cubed Sphere, FV3, dynamic
core was actually written as a dynamic climate model, but we
realized we could actually use it at high resolutions as a
weather model. What's appealing to me here is when we can unify
a lot of the code architecture, then we eliminate a lot of
these redundant kind of parallel development programs, and have
the same amount of people all focused on a unified forecasting
system.
The timeline is essentially--we really wanted to focus on
the global model first because the global model provides the
boundary conditions for all of the other models. In other
words, if you're interested in looking at a forecast for
harmful algal bloom, we have biological models for that, but
those models depend on a hydrological model, which models how
there's runoff, and, of course, that depends on the weather
model, which forecasts precipitation. So the first sort of
foundation of this is implementing the global model, and then
all of the other models that use that for initialization will
then be implemented.
Mr. Casten. And, I'm sorry, just because I'm tight on time,
and I want to get one more question in, any ballpark on
schedule, or is it too soon to say?
Dr. Jacobs. Well, we're already pretty far down the road in
developing the Unified Forecast System. We're expecting--we've
been--we've actually just had a public release of the
seasonal--the sub-seasonal version of the code. It's on GitHub
now, but it's supported yet for the community, so we're really
trying to work on the model support for that. These upgrades
will be coming out on GitHub as I speak. We're hoping to get
the RFP (request for proposals) for EPIC out early next year.
Mr. Casten. OK. Dr. Mass, with the little time I have left,
I want to raise with you a conversation I had with Dr. Jacobs,
given your numerical modeling background. I sit on the AI
(artificial intelligence) Task Force on Financial Services, and
we have this continuing conversation around--in a world of
machine learning and AI, there's this tradeoff between
precision and accuracy and transparency of algorithms. As we
build out more and more sophisticated weather models, given
your background in modeling, do you see tradeoffs coming in the
weather forecast model as we get more precision with machine
learning, but potentially start to separate from the
fundamentals in the models that we've relief on that have some
level of physical understanding? And is there anything that we
should be concerned about as a Committee if that break happens?
Dr. Mass. I don't know if concern is the right word, but I
think the marriage of AI with modeling is very powerful. AI
is--and machine learning's very powerful for quality control,
but just as importantly for post-processing and model output,
so you need both. You need the dynamical models, but you need
machine learning on both sides, so--together they're much more
powerful.
Mr. Casten. OK. And I see I am out of time, so I now
recognize Ranking Member Marshall for 5 minutes, who we have
missed--Mr. Lucas, you are recognized for 5 minutes.
Mr. Lucas. Thank you, Mr. Chairman. I appreciate the
opportunity, and, Dr. Jacobs, you've been before the panels
enough times to know that occasionally we ask about the same
thing, we just come at it from slightly different angles. And,
with that in mind and in spirit, I'd like to note that I am
concerned that NOAA may be approaching EPIC implementation as
something of a software issue, or the agency views the biggest
challenge to improving weather forecasting is simply improving
existing software. What assurance can you give the Committee
that NOAA's engaging the academic community and the private
sector to ensure that this is truly a community-based weather
prediction model?
Dr. Jacobs. Well, the primary assurance I can give you is
that it'll have to exist outside of NOAA, and having the
involvement of industry and academia is essential. This will
all be built into the request for proposals. We had a big
meeting in August. We just released the report of that meeting
yesterday, and we had a lot of stakeholders from all the
sectors involved and contributing, so the whole point and
design of this is to--is a stakeholder-run program, with an
operational outcome that NOAA should benefit from.
Mr. Lucas. And along that line, Doc, I'm a Member of a body
where there seems to be an ever-increasing turnover, so when I
ask this, I ask this in the politest of terms. What assurance
can you give the Committee that EPIC will continue past your
tenure at NOAA?
Dr. Jacobs. Well, once the RFP--once----
Mr. Lucas. And I'm not predicting the length of anyone's
tenure. I'm just asking about continuality.
Dr. Jacobs. No, this was a top priority for me, and making
sure this lives past my tenure is extremely important, and
that's another reason why having it, you know, externally
managed would ensure that if it's inside of NOAA, then NOAA,
you know, has complete and total direction. If it's external,
even if--I mean, I expect that NOAA will have a seat at the
table in guiding what it does, but largely driven by external
stakeholders will ensure that, even if I'm not at NOAA anymore,
it will still be successful.
Mr. Lucas. And along that line, if you could expand a
little bit on your current acquisition strategy, and the
timeline that we're very concerned about here about
implementing EPIC?
Dr. Jacobs. So there's the RFP, which we expect to go out
early next year, first month or two, and then an additional--so
there's two sort of acquisition strategies we're concerned
about. One of them is where does EPIC live? We'll learn that
when the award goes out. Then there's the acquisition of cloud-
based compute resources. This is a little bit trickier. So we
have a need to procure cloud-based resources, but right now,
with the Federal acquisition regulations, it's very complicated
for us to try to figure out how to buy cloud compute on a
demand that fluctuates. It's not just like buying a rack of
servers.
Mr. Lucas. And as an elected official, I would be remiss if
I didn't ask how universities, such as The University of
Oklahoma, would have a chance to participate in EPIC?
Dr. Jacobs. Well, my hope----
Mr. Lucas. They're are homegrown questions, you know,
wherever we come from.
Dr. Jacobs. I, you know, so obviously their expertise is
largely in convective weather forecasting, and there's a
component of this that will deal with that weather forecasting.
I would hope that they would be both contributing model
improvements, as well as benefiting from some model
improvements, as well as the Mesonet Program, and all of the
different observing systems. We'll be able to test those in
this open cloud-based architecture, which I think will benefit
not just the forecasting, but the observation systems going
into the models.
Mr. Lucas. And one more time, touch on what you would
define as the milestones that will reflect our progress toward
closing the gap with the Europeans?
Dr. Jacobs. So the first initial milestones of success
would just be how many external stakeholders are using the
code, so we would want to monitor how many downloads, what's
the user interest? And after that we would, you know, we would
hope that we would start to see contributions and improvements
coming back, but the initial steps are just getting the
external stakeholders engaged in the program.
Mr. Lucas. And, Dr. Mass, you're not exempt from these kind
of questions either. I ask this, and I think I know the answer,
but I'm going to ask, does the broader weather research
community support EPIC?
Dr. Mass. I think the answer's clearly yes, but we want
EPIC to be something that does serve the entire community.
That's crucial.
Mr. Lucas. Thank you, Doctor, and with that I yield back,
Mr. Chairman.
Mr. Crist [presiding]. Great. Thank you very much. Thank
you to the witnesses for being here today. We all appreciate
your attendance.
The ability to forecast hurricane tracks has greatly
improved since the 1960s. Forecasting hurricane intensity has
also improved, but less so than the track forecasts. In fact,
in 2017 Congressional Research Service reported on forecasting
hurricanes and found that the biggest challenge facing the
National Hurricane Center is how to improve the ability to
predict hurricane intensity. Being able to better predict how
strong a hurricane will become, as well as when and where it
will peak in intensity, is key for the district I represent
along the west coast of Florida, which, as you know, is
incredibly vulnerable to hurricanes year in and year out.
So, Dr. Jacobs, given how costly hurricanes are to the
United States, and in particular to Florida, how will EPIC
improve hurricane intensity prediction, if it will?
Dr. Jacobs. My expectation is that it will. The focus
primarily for hurricane intensity is largely centered around
two things: Two-way coupled modeling, with an emphasis on sea
surface temperature, because that's essentially the fuel; and
the physics in the model. So there's a deficiency in the model
physics that needs a lot of research and improvement. EPIC will
essentially be the external sandbox where stakeholders can test
their improvements to both the physics, as well as the two-way
coupling of the models, and then, additionally, new
observations, new observing systems.
So there's a lot of new observing systems coming online,
not necessarily NOAA assets, but industry assets, and even
academic devices, that we don't have the internal bandwidth to
test the impact of those obs in our system, but we can test the
impact of those observing systems in the proposed EPIC sandbox.
Mr. Crist. Great. Would any of the other witnesses care to
comment on how EPIC will improve hurricane prediction? Dr.
Neilley?
Dr. Neilley. Thank you, Mr. Crist. I think one of the
critical ways in which EPIC can improve hurricane forecasting
is by marriage of the Unified Forecast System with next-
generation data assimilation techniques, particularly the types
of technologies that are being developed in Dr. Auligne's
group. There has been numerous scientific evidence that data
assimilation, taking the observations that Acting Administrator
Jacobs mentioned, and using them to initialize the model, can
be one of the most important aspects of getting the hurricane
forecast right. EPIC, if crafted correctly, is the cauldron in
which entices all of these scientific capabilities to come
together and be married to improve our weather forecasting
capabilities.
Mr. Crist. Great. Thank you very much. Any others? Yes.
Dr. Auligne. Let me tag along to this response. So there's
multiple evidence showing that, if we're looking at the quality
of the forecasting, the skill of the initial conditions and the
actual model are equally important, and data assimilation is
handling the initial conditions for the model, so we're
actually blending together observation and previous model
forecasts to actually optimize these initial conditions, which
due to the chaotic nature of the atmosphere, are propagating
and amplifying in the forecasts, so it's actually a critical
component of forecast accuracy.
Mr. Crist. Great, thank you. I wanted to ask--I'm running
out of time. I wanted to ask one additional question, if I
might. In addition to weather forecasting, NOAA's also
responsible for researching and modeling other types of
environmental concerns, such as red tide and algae bloom,
which, as you know, is of great interest in the Sunshine State.
Dr. Jacobs, can you discuss how EPIC will help improve
forecasting for red tide and other harmful algae blooms?
Dr. Jacobs. So, as I was explaining earlier, a lot of these
harmful algal bloom models depend on the atmospheric model
forecast of precipitation and runoff to determine when there'll
be triggers. Ultimately we are going to put in all of these
secondary downstream models, so to speak, into the EPIC
program. In conjunction with this, if we have an external
repository for the modeling system, as well as the code that
we're running on the cloud, we need to have an archive and
repository for observations, and so building our observation
system in the cloud is going to be essential both for
initializing and verifying the models. And that's largely what
NOAA's Big Data Project is focused on.
Mr. Crist. Thank you very much, Dr. Jacobs. I would now
like to recognize the Chair, Madam Fletcher.
Chairwoman Fletcher. Thank you, Mr. Crist, for sitting in
the Chair. I believe my first order of business will be to
recognize Mr. Murphy for 5 minutes.
Mr. Murphy. Thank you, Madam Chair. Thank you, gentlemen,
for coming this afternoon. My area of expertise for the weather
is looking up and seeing whether I need an umbrella or not, so
thank you for giving me a heads up on that. Just one actually
really kind of rudimentary question. Can someone just explain
to me, in terms that I might understand, the difference between
the European models and the American models? In other words,
when I look at hurricanes, I live in eastern North Carolina, so
we love to do the hurricane watch, and we see the American
model is doing one thing, and then the European model is doing
another thing. What's the fundamental difference between those
two models?
Dr. Jacobs. So I'm going to try to answer this real fast,
and then hand it over to the rest of the panel, but the primary
fundamental difference is how the European center does data
assimilation, and that's basically how they generate the
initial conditions in the model. They use a true what we call
four dimensional variational data assimilation, and right now
the NOAA modeling system uses a four dimensional ensemble
variational assimilation, but it's not a--they don't truly vary
time as the fourth dimension.
Mr. Murphy. Is he being truthful?
Dr. Auligne. Yes. So the principles--the fundamental
equations are the same. The way that these models are
implemented are quite different, like two car models can be
quite different, although the principle of the car is the same.
So, in terms of data assimilation, there's definitely a lot of
emphasis in Europe on the algorithms, and the use of additional
instruments, and more data, so that translates into actually
additional forecast skills.
Mr. Murphy. All right. Thank you. One other question. In
medicine, we look forward to every new advancement, and what
things are going to happen. But I will ask this question, and
this is, you know, an honest, hard question. What is going to
be our ROI (return on investment) on this? In other words,
where are we now, stagnant now, and then if EPIC is fully
instituted, how much more advanced do we truly, honestly,
expect to be? In other words, what can we gain from this in
implementing this system? Yes, sir, Dr. Mass?
Dr. Mass. Well, there's no reason to think the European
Center is as good as we can be. American research capabilities
far exceed that of Europe as an aggregate, so our skill can be
better than that. I think we could catch up within years if we
just got the data assimilation right, and then, over a longer
period of time, improve physics. I think we could be ahead of
them in 3 to 4 years if we really put the energies into it.
Mr. Murphy. All right. Just a follow-up question, because
I'm a function over form kind of guy, when we have hurricanes
coming toward the east coast, everybody acts the same. They buy
their water, they get their bread. How is this going to make it
any different? Again, I'm just looking, you know, I believe in
research, I believe in the advancement of knowledge by all
means, but how is it going to change the lives of the average
American, say on the east coast, with hurricanes?
Dr. Neilley. All people make decisions in all types of
weather events, whether or not it's a hurricane, or a more
mundane thunderstorm in the afternoon. Anytime you can make
better decisions because you have better information, you're
better off. There are estimates that the national economy is on
the order of a trillion dollars dependent on weather, and by
incrementally improving our weather forecast, if we can make
that dependency down by, say, just 1 percent, that alone is a
$10-billion payoff for our economy. I think that's the ROI that
you asked for.
Mr. Murphy. All right. Thank you. Chairwoman, I yield back
my time.
Chairwoman Fletcher. Thank you very much. I will now yield
myself 5 minutes for questions, and I apologize for just
arriving, as I believe my colleague Mr. Babin did. We were
voting in another Committee. But I'm really glad to be here to
see all of you, and I appreciated your written testimony. I'm
sorry I missed your initial comments, but I do have a couple of
questions that I think haven't been covered yet.
It's clear, from the written testimony, the recommendations
from the Environmental Information Services Working Group, that
a strong, accountable, and vision-oriented leadership and
management is needed to ensure EPIC's success. To my
understanding, there isn't a clear plan for that leadership, or
management, or governance at this point, and so I want to ask
all of you, what are your thoughts about who should lead EPIC
and how it should be structured? And, Dr. Jacobs, I'd like to
start with you.
Dr. Jacobs. So I'm going to answer this question at a very
high level, because, really, the point of EPIC is to have it
governed by the weather enterprising community, so I'd be
interested to hear what the weather enterprising community had
to say. The request for proposals is going to have some
guardrails, but part of what we're going to be asking in the
RFP is also proposal of a governance structure. You know, NOAA
obviously has to be involved, but we want involvement from
private sector and academia.
And ultimately EPIC may end up, you know, it could be, you
know, end up at a university, it could end up at UCAR
(University Corporation for Atmospheric Research), it could end
up in industry, it could end up in some kind of combination of
all of the above. The only things that I really would like are
that its' got to be external to NOAA, NOAA's got to have a seat
at the table, it's got to have an operational outcome for NOAA
forecasting products in mind, and, other than that, a lot of
the governance is going to be part of the proposal of wherever
it ends up.
Chairwoman Fletcher. Thank you, Dr. Jacobs. Dr. Mass?
Dr. Mass. Well, this is going to be a community modeling
system, so the community needs to be there. There needs to be
at least an advisory committee that's in place. There needs to
be some kind of group that encompasses all the people that are
putting money into it, so that'll be there. But there needs to
be leadership. One person has to be responsible. There has to
be a leader, a director of EPIC, somebody who's responsible,
and if it doesn't work out, heads will roll, that person. So
you need responsibility, one point of responsibility.
Chairwoman Fletcher. Thank you. Dr. Neilley?
Dr. Neilley. Thank you. I think the most critical
characteristic of a successful EPIC is the breadth of the
science and diversity of science that takes place in it. In
order to achieve that, EPIC has to be crafted in a way that the
scientific community is enticed to participate, and is not sort
of mandated or directed to participate. It is the place to go
to conduct numerical weather prediction science in the world,
and, as such, it will create, therefore, the best numerical
weather prediction science, and come back to benefit NOAA and
others.
Who should lead EPIC is the institution that is best able
to create that enticing institution that scientists want to go
to, and that's who should lead them.
Chairwoman Fletcher. Thank you, Dr. Neilley. Dr. Auligne?
Dr. Auligne. So I think that one of the main risks for EPIC
is fragmentation. We want, first, to make sure we have a clear
focus, and clear--as Dr. Mass was saying, clear accountability.
We need to make sure we can define and measure success, and can
have somebody accountable for it. Then we need to have these
clear connections with the community, and clear connection with
the government as well. So it can't be completely inside the
government, because we're not trying to replace NOAA. We're not
trying to replace the research and the R&D (research and
development) in NOAA. We're trying to supplement, and really
help the government with more agility, and more connection to
the community. So that's basically what I think is required for
the institution that would lead it.
Chairwoman Fletcher. OK, thank you. And I want to follow
up, Dr. Mass, on your comment about leadership, there needs to
be clear leadership. Do you have ideas, or a vision, or a
thought, of what that leadership should look like, not a
specific person, obviously, but when we talk about what is the
structure, and I'm going to circle back to Dr. Jacobs in a
second, but what does that look like to you, or what should it
look like, in terms of that leadership?
Dr. Mass. Well, we can see that. We can see our
competition, the European Center. They do have a leader, a
scientific leader, that oversees the whole program. That's the
responsible person. But they do have an advisory board that's
there as well that represents all the various countries that
are involved, and they have scientific advisory committees. So
they give us somewhat of a pattern of what we could follow
that's been highly successful.
Chairwoman Fletcher. OK. Thank you. And, Dr. Jacobs, you
mentioned in your testimony about the RFP, so I just wanted to
get a follow up on that as well to know your thoughts about
sort of a dedicated staff and leadership team, and if that is
something planned, and if so, when it will be announced?
Dr. Jacobs. So that's largely going to come out of the
responses from the RFP, wherever the award goes, but, to Dr.
Mass' point, I think we'll probably end up finding a--where we
have some type of board, and then a single-point person who has
autonomy, accountability, and control over the budget. Some of
the things that I've seen have failed in the past were run by
individuals who had complete autonomy, and no budget authority,
so they couldn't actually execute great decisions.
Chairwoman Fletcher. Thank you, that's helpful. I have
managed to go over my time already, so I'm going to yield back,
and I'm going to recognize Dr. Babin for 5 minutes. Thank you
very much.
Mr. Babin. Thank you, Madam Chair, appreciate it, and
appreciate all of you expert witnesses for being here. Dr.
Jacobs, as you know, the 36th District of Texas, over in the
southeast portion, was hit especially hard by Hurricane Harvey
in 2017. Additionally, Tropical Storm Imelda dumped a record
amount of rainfall in my district just a few months ago, in
September. This storm came out of nowhere, with a severity that
surprised everyone, and left most of my constituents without
time to prepare for it.
Let's jump ahead, hypothetically, just a few years into the
future, where EPIC has been successfully implemented, and is
operating. Can you walk us through the processes of a fully
functioning EPIC as the storm approaches, and how the days
leading up to, and immediately following a storm like Harvey or
Imelda, and how that would be different?
Dr. Jacobs. Sure. Thank you very much for the question.
Mr. Babin. Yes, sir.
Dr. Jacobs. So the fully successful EPIC will happen,
pardon the pun here, but way upstream from these precipitation
events. So what would happen is--when you look at the National
Water Center, which is a fantastic center, they've got the
National Water Model, it's, you know, it's a very successful
program, but if there is one weak link in the National Water
Model, it's that we have to forecast properly where the rain's
going to fall, otherwise we don't know where the runoff is
going to go.
And so having EPIC be the center where we actually feed in
inputs to improve the forecast of prediction of rain will then
subsequently improve the prediction in the hydrological models.
So a lot of this will happen months to years in advance, but
you will see the improvements of those actual forecasts find
their way down into, you know, things like, you know, not just
hydrological models, but also biological and ecological models
as well.
Mr. Babin. Sure. OK. And I also serve as the Ranking Member
of the Space Subcommittee here, with the privilege of
representing Johnson Space Center. I know the impacts that one
government facility can have on an entire region. As it stands,
EPIC will be a virtual center that will have tremendous
benefits by operating in the cloud, both in terms of cost and
innovation. Again, looking ahead to years down the road, when
EPIC will be running smoothly, and surely will be the gold
standard at that time, is there a scenario where a physical
center, rather than a virtual center, could be beneficial to
EPIC's mission?
Dr. Jacobs. Absolutely. Thank you for allowing me to
clarify this. I have often referred to EPIC as a virtual center
because EPIC, when it's listed in the budget proposal, was--as
$15 million, and I didn't want anyone in the budgeting process
to think I was planning to actually build a brick and mortar
center for only $15 million. So, when we put out the RFP, EPIC
will have to live somewhere. There will have to be people in
seats, at computers, in some type of facility, whether it's a
university or other, you know, other facility, there will have
to be a physical center. In addition to that, if we have the
opportunity to expand this program, it's entirely probable that
we, you know, we would need an additional physical center for
this somewhere.
Mr. Babin. Right. Yes, sir. Dr. Mass?
Dr. Mass. Well, even if we're very successful creating the
best weather prediction system in the world, we're still going
to need the computers. So if we don't have vastly increased
computer resources, we're not going to be able to deliver the
forecasts that we really want to. That's really important.
Mr. Babin. I understand. Thank you. Well said. You know,
I'll yield back, Madam Chair. Thank you very much.
Chairwoman Fletcher. Thank you, Mr. Babin, and thank you to
all of our witnesses. I know we just rushed in, and
unfortunately, we have to go back and vote in our other
Committee. I'm sorry to say, but since we have all come and
gone from the hearing, I really appreciate all of you coming
in, testifying, sharing your thoughts. I think this is really
exciting to see what's happening, and I'm grateful for all of
you participating today.
Before we bring the hearing to a close, I also want to
mention that the record will remain open for 2 weeks for
additional statements from Members, and any additional
questions that the Committee Members may have for the
witnesses. But, for now, the witnesses are excused, and the
hearing is now adjourned.
[Whereupon, at 2:57 p.m., the Subcommittee was adjourned.]
Appendix I
----------
Answers to Post-Hearing Questions
Answers to Post-Hearing Questions
Responses by Dr. Neil Jacobs
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Responses by Dr. Cliff Mass
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Responses by Dr. Peter P. Neilley
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Responses by Dr. Thomas Auligne
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Appendix II
----------
Additional Material for the Record
Letter submitted by Representative Sean Casten
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]