Utilizing Scientific Advances in Operational Systems - PowerPoint Presentation

Slide1

Utilizing Scientific Advances in Operational Systems

Terry Onsager, Tom Berger, and Howard Singer

NOAA, Space Weather Prediction CenterTerry.Onsager@noaa.gov Slide2

Main Points

• Operational Sun-to-Earth Modeling Suite

• Elements of a Research-to-Operations Effort•

Elements of an Operations-to-Research Effort• Challenges in Utilizing Research in OperationsSlide3

Solar /Solar Wind

Components of NOAA’s Numerical Space Weather Modeling Effort

Magnetosphere/

Ionosphere

Atmosphere/

Ionosphere

L1 Satellite Location – ACE and

now

DSCOVRSlide4

SWPC Operational Model SuiteTracking solar storms from “Sun to Mud”

GMU/AFRL WSA/

Enlil

U. Michigan

Geospace

NOAA/CIRES WAM-IPE

USGS/NOAA E-field

Inputs

:

GONG solar magnetic field data

SOHO/LASCO coronagraph CME images from L1

Validation:

DSCOVR solar wind character at L1

GOES magnetometer shock arrival

Inputs

:

DSCOVR solar wind density, temp, speed, mag field at L1

Solar F10.7 radio flux measurements

Validation:

GOES vector magnetic field

USGS magnetometer network

Inputs

:

GFS Tropospheric weather model inputs

GOES Solar EUV flux

COSMIC-2 RO electron density

Geomagnetic storm data from

Geospace

Validation:

GPS receiver network TEC measurements

Inputs

:

USGS lithospheric conductivity model

USGS magnetometer network

Validation:

USGS

geoelectric

field measurements.

Note: all models developed with NASA and/or NSF funding at some level.

Operational

Operational FY16

Operational FY17-19

Operational

FY17Slide5

Four Stages of Utilizing Scientific Advances in Operations

1. Demonstrate model value – Value to customers must exceed cost to transition and to run model operationally

2.

Develop operational software from research model

3.

Implement

model and product

generation on

operational

computer

4. Continuously improve and upgrade operational model

WSA-Enlil

GeospaceSlide6

Establishing Value toward Operational Services is a Necessary Condition

• Demonstrate model value – Value to customers must exceed cost to transition and to run model operationally

Strategic Importance

Operational Significance

Implementation Readiness

Cost to Operate, Maintain, and Improve

• The need for performance metrics has long been well known

•

However, we still do

not know how quantitatively good or bad our current scientific or operational capabilities are (metrics)Slide7

Establishing Operational Value is a Necessary Condition

• Demonstrate model value – Value to customers must exceed cost to transition and to run model operationally

WSA-Enlil

Taktakishvili

et al., 2009

Enil

/Cone Shock Arrival Time Errors:

+/- 5.9 hoursSlide8

Establishing Operational Value is a Necessary Condition

• Demonstrate model value – Value to customers must exceed cost to transition and to run model operationally

Geospace

models evaluated: Regional K and dB/

dt

CCMC, modelers, SWPC, and science community

Univ. of Michigan SWMF

Distribution of observed mid-latitude K values for modeled K values of 4, 6, and 8

Observed K (Newport Station)

K = 4

K = 6

K = 8

Modeled KSlide9

Operational Evolution – “Operations-to-Research”

• Forecasters and customers gain experience and provide feedback

•

Scientific advances improve model quality

• However, mechanisms do not exist to enable scientists to use, evaluate, and participate in the improvement of operational

models

Continuous improvement of operational models is an unmet needSlide10

Operational Evolution – “Operations-to-Research”

National Space Weather Action Plan (2015):

Action 5.6.2: DOC

and DOD, in collaboration with NASA and NSF,

will develop

a plan (which may include a center) that will

ensure the improvement, testing, and maintenance of operational forecasting models

.

Operations-to-Research Workshop held August 16-17, 2016

Building a culture of research-operations coordination was recognized as a key factorSlide11

- Agencies recognize the need to address both space weather research and operations - There have been many success where operational capabilities have been significantly enhanced;

- The major user needs for operational forecasts and specifications are well known; - The need to improve our operational products and services from our research understanding is well known;

- The need for performance metrics is well known.

Areas with Good AwarenessSlide12

Areas in Need of Improvement

-

We

do not know how to predict

what we know

needs to be predicted (i.e., solar eruptions, IMF at 1 AU, etc

.)

-

We do not have quantitative scientific or operational metrics

-

Funding to address topics of operational relevance is lacking

- Scientists are unable to use, evaluate, and participate in the improvement of operational models, e.g., community models - The research environment fosters the advancement of fundamental scientific understanding, not the improvement of operational products.Slide13

-

Research to improve the accuracy of forecasts and warnings

- Collaborative between operational forecasters and academic institutions- Applied research of interest to operational meteorology community- Apply scientific knowledge to operational products and services

Applied Research in MeteorologySlide14

• Elements of using scientific tools in operations:

- Demonstrate operational value; - Develop operational software; - Implement on operational computer;

- Continuously improve operational capability. • Performance metrics are important to determine operational value and to measure improvement. • A key challenge is to have scientists actively participate in the use, evaluation, and improvement of operational products. • If community models are required, how can we establish a path to develop them?

Summary