Managed by Triad National Security, LLC for the U.S. Department of Energy’s NNSA

Mathematics and Los Alamos National Laboratory: Advances and Collaboration

Carrie A. Manore, PhDA-1, Information Systems and Modeling

cmanore@lanl.gov

Sarah Michalak, PhDCCS-6, Statistical Sciences

Thanks to: Dr. Scott Doebling and Dr. Abigail Hunter

Los Alamos National Laboratory: History and MissionAt Los Alamos, we deliver science and technology to protect our nation and promote world stability

• Our mission began by applying science and technology to address an international crisis

• Now we work to assess & reduce global nuclear danger

• Responsible for nuclear stockpile stewardship

• Theory, simulation & computing, experiments, engineering, and manufacturing

Simulation & Computing

Experiments

Theory

Los Alamos National Laboratory: History and Mission

Our Mission:

• Nuclear deterrence and stockpile stewardship

• Protecting against nuclear threats

• Emerging threats and opportunities

• Energy security solutions

Los Alamos National Laboratory: History and Mission

Our Mission:

Dual-axis radiographic hydrodynamic test facility (DARHT)

Proton radiography at Los Alamos Neutron Science Center

Chemistry and metallurgy research

Supercomputing and computational physics

Materials and explosives research and fabrication

Los Alamos National Laboratory: History and Mission

Our Mission:

• Nuclear deterrence and stockpile stewardship

• Protecting against nuclear threats

• Emerging threats and opportunities

• Energy security solutions

Los Alamos National Laboratory: History and Mission

Our Mission:

Rapid Response: respond to and successfully resolve nuclear and radiological threats worldwide

Nuclear Safeguards: ensure non-proliferation

Space Systems: instruments, modeling & simulation

Radiation detectors

Los Alamos National Laboratory: History and Mission

Our Mission:

• Nuclear deterrence and stockpile stewardship

• Protecting against nuclear threats

• Emerging threats and opportunities

• Energy security solutions

Los Alamos National Laboratory: History and Mission

Our Mission:

Cyber: security, quantum computing

Space: remote sensing

New technology

Other emerging threats: disease, climate change, conflict, fire spread

Prediction, prevention, and mitigation

Fire growth and spread

Los Alamos National Laboratory: History and Mission

Our Mission:

• Nuclear deterrence and stockpile stewardship

• Protecting against nuclear threats

• Emerging threats and opportunities

• Energy security solutions

Los Alamos National Laboratory: History and Mission

Our Mission:

Safe and sustainable nuclear energy

Mitigating impacts of global energy demand growth

Materials and concepts for clean energy

Case Study 1: Forecasting Infectious Diseases

Collaborators include:

Katherine Kempfert (University of California Berkeley)Kaitlyn Martinez (Colorado School of Mines)Daniel Romero-Alvarez (University of Kansas)Jessica Conrad (University of Michigan)Margaret Grogan (University of Tennessee, Knoxville)

The Los Alamos Staff Team

Sara Del Valle A-1

Geoff Fairchild A-1

Amanda Ziemann ISR-3 Nidhi Parikh A-1

Carrie Manore A-1 Jeanne Fair B-10

Adam Atchley EES-16

Kim Kaufeld CCS-6

Ethan Romero-Severson T-6

Chonggang Xu EES-14

Jon Schwenk EES-14

Students and Postdocs

Jessica Conrad, Public Health & Comp Math; UMich

Katherine Kempfert, Statistics Berkeley

Morgan Gorris, Earth Systems Science UC Irvine

Jamil Gafur, Computer Science postbac XCP-8

Kaitlyn Martinez, Math & Stats Colorado School of Mines

Devin Goodsmanprevious postdoc, scientist Alberta F&W Canada

Andrew Bartlow B-10

Deborah Shutt A-1

Daniel Romero-Alvarez, Ecology and Evolutionary Biology, Univ Kansas

Amir Siraj, University of Notre Dame

Forecasting Dengue (Mosquito-borne Disease)

Clinical Surveillance Data

Satellite Imagery

Climatological Data

Demographic Data

Google Search QueriesRisk Map & Forecasts

+Predictive Models

Real-time, voluminous, extremely noisy data

Can we forecast events using big, heterogeneous data streams?

Predict: Clinical Surveillance Data

■ Dengue case data from Brazilian Ministry of Health

■ Cases of Dengue from 2010 to 2016 for 5,570 municipalities‾ Weekly resolution

‾ Seasonal variation

‾ Regional differences

■ Likely reporting bias

■ “Gold standard” Percent of the total 2015 reported dengue cases in each “meso-region” by week.

Use: Satellite Imagery – Computed Values

■ 2010 – 2016 (~360 weeks) for each municipality

■ 4 satellites (Landsats 5, 7, 8, and Sentinel-2)

■ 4 indices– Vegetation (Normalized Difference Vegetation Index -

NDVI); Water content in leaves & bodies (Normalized Difference Water Index NDWI); Burn (Normalized Burn ratio – NBR); percent Cloudy pixels

Change detection & classification techniques

at scale

Soil

Vegetation

Water

Use: Weather Stations

■ Weather stations in Brazil from NOAA’s “Global Surface Summary of the Day GSOD” data set (613 stations)

■ Collected:⎼ Daily temperature (mean, min, max)⎼ Daily precipitation

■ Missing data and errors common■ Use interpolation techniques to fill

gaps

Use: Demographic Information

■ 2010 Census data consisting of over 200 variables at municipality

■ Example factors predictive of dengue– % Rural population– % with inadequate water supply and

sewage– % Poverty– % Population with garbage collection

service

Use: Google Health Trends

■ Google Health Trends time series matching available incidence data

■ Identified keywords most associated with dengue

■ There is a ~2 week lag in reporting for case counts, Google searches is real time, so gives lead

■ Google Health Trends predictive with lead time for 12 out of 26 states (R2 > 0.75

Data Fusion

Fuse data to:-5 spatial scales-weekly time scale

Forecasting: 2 Weeks Ahead Using All Data Sources

Preliminary work. Statistical time-series models built on 5 years of data using satellite indices, weather, and Google Health trends provided accurate weekly predictions for the 2 year validation data (two states). Our predictions (blue) , actual cases (red).

Seasonal Autoregressive Integrated Moving Average with Exogenous Variables

Case Study 2: Materials Physics Modeling

Collaboration with University of California, Santa Barbara

Los Alamos National Laboratory

• Lead UCSB Professor: Dr. Irene Beyerlein• Lead LANL Scientist: Dr. Abigail Hunter• Ph.D. Students: Claire Weaver, Lauren Smith– Students spend summers visiting Los Alamos• Dr. Hunter visits UCSB for ~2 months in during the Spring quarter

Aerial view of LANL – www.lanl.govHenley gate at UCSB –labs.me.ucsb.edu/beyerlein/irene/

www.lanl.gov

Project: Phase Field Dislocation Dynamics (PFDD)

How do we design stronger and lighter materials?

How do metals deform?How do we understand and predict material strength?

How do we design stronger materials?Hull and Bacon, Introduction to Dislocations

deloreanipsum.comMeters

Microns

Nanometers

Macroscale/Continuum

Mesoscale

Microscale

Grains

Grain boundaries

wikipedia.org Dislocations

wikipedia.org

PFDD

Dislocations

Grains

I.J. Beyerlein and A. Hunter, Phil. Trans. R. Soc. A 374 (2016).A. Hunter, B. Leu, and I.J. Beyerlein, J Mater. Sci. 53 (2018).

Hexagonal Close Packed (hcp) Metals – Magnesium

• High strength-to-weight ratio– Lighter than Al and steel – “light-weighting”• 8th most abundant element and recyclable• Resistant to:– Radiation, Fatigue, Corrosion, High Temperatures…• Low intrinsic ductility and poor formability– Severely limits applicability due to difficulties in

fabrication• How do we improve ductility? How do Mg-

alloys deform?– Solutions require studies of dislocation energetics,

nucleation, multiplication, annihilation processes, and interactions with microstructure and interfaces

Figures courtesy of Claire Weaver (UCSB).

Schematics showing the complex hcp unit cell and some of the different slip modes that must be considered in simulations.

!

"

#

11%20

1%100

0001 !

"

#

11%201%100

0001

!

"

#

11%20

0001

1%100!

"

#

11%20

0001

1%100

!

"

#

11%23

1%100

%1%122 n

!

"

#

11%23

1%100

%1%122 n

a b

c d

!

"

#!

"

#

11%20

1%100

0001 !

"

#

11%201%100

0001

!

"

#

11%20

0001

1%100!

"

#

11%20

0001

1%100

!

"

#

11%23

1%100

%1%122 n

!

"

#

11%23

1%100

%1%122 n

a b

c d

!

"

#

!

"

#

11%20

1%100

0001 !

"

#

11%201%100

0001

!

"

#

11%20

0001

1%100!

"

#

11%20

0001

1%100

!

"

#

11%23

1%100

%1%122 n

!

"

#

11%23

1%100

%1%122 n

a b

c d

!

"

#

Simulation results showing dislocation dissociation on the prismatic plane in three hcp metals

C. Albrecht-Weaver, A. Hunter, A. Kumar, and I.J. Beyerlein, in preparation (2019).

Multiple Principal Element Alloys (MPEAs)

• MPEAs are material systems with three or more alloying components that are combined in approximately equal proportions (equimolar).–Have been show to have high strength and good

fracture toughness particularly at high temperatures.• Exciting candidate for structural materials under extreme

conditions.

• In contrast, conventional alloys are composed of one primary element with small amounts of other elements to enhance properties.– In conventional alloys, the material properties are

often similar to the primary element. This is not the case in MPEAs, in which there is no primary element.

Figures courtesy of Lauren Smith (UCSB).

Variations in the lattice energy due to the different chemical compositions present in MPEAs.

Dislocation nucleation via Frank-Read source in an MPEA. The loop is not symmetric due to the varying chemical composition of the MPEA.

L. Smith, Y. Su, S. Xu, A. Hunter, and I.J. Beyerlein, in preparation (2019).

Computational Science & Engineering: Opportunities

Los Alamos is the birthplace of computational physics

Numerical methods for multi-material compressible flow

Richtmyer-M1eshkov instability

Laser-driven reshock simulations

Inertial Confinement FusionReactive high

explosives modeling

High performance computing & visualization

1

Verification, Validation & Uncertainty Quantification Machine Learning

for inter-atomic potential calculations

• 40 miles to Santa Fe, 100 miles to Albuquerque• 7000 feet above sea level: Low pollution and four seasons• Abundant outdoor activities in nearby mountains, mesas,

forests, and rivers• High quality of life with moderate cost of living• Rich Spanish & Native American cultural history

Where We Are: Northern New Mexico

Albuquerque

Santa Fe

Los Alamos

60 miles

40 miles

1

Diverse teams of 12,000 employees at Los Alamos work collaboratively to solve national security challenges

• 4000 Scientists and Engineers

– 2200 PhD-level

– 145 R&D100 awards, 34 EO Lawrence awards,9 Presidential Early Career awards

• 400 Postdoctoral researchers

• 1500 summer students

• $2.8 Billion budget

• 36 square miles of facilities

Physics Verification & Analysis Group

1

Data Analytics Team, Information Systems and Modeling Group

Student and Career Opportunities

• Student Programs– Undergraduate internships in science & engineering:– Post-Bachelor / Post-Masters: • Can be entry level or “gap year” between degree programs

– Post-Doctoral:• The most common entry level for PhD. Minimum pay $74k/yr

• Scientific & Engineering Staff– Permanent positions with pay depending upon degree and

experience, often pipelined via one or more of the student programs

– Often requires the ability to obtain a security clearance, which normally requires US citizenship

• We work hard to find opportunities for dual-career couples in science and engineering

1

Explore the opportunities for you at Los Alamos

lanl.jobsor

jobs.lanl.gov

1

www.lanl.gov