overview

1 May 2008 Space Weather Workshop 2008

Predicting the Space Environment:Recent results from CISM and Where do we go from here?

W. Jeffrey HughesCenter for Integrated Space Weather Modeling,

and Boston University


Overview• What is CISM – a very brief introduction.

• Recent modeling advances

• Where do we go from here? How do we improve 1-day predictions of: – Solar flares– Solar protons– Geomagnetic storms– Ionospheric disturbances


Center for Integrated Space Weather ModelingFunded in August 2002, initially for 5 years, under the NSF Science and

Technology Center program.

NSF has extended funding for the next 5 years (2007-12).

GOAL: Develop a suite of physics-based numerical simulation models that describes the space environment from the Sun to the Earth.

USES:

• Scientific tool for increased understanding of the complex space environment.

• Specification and forecast tool for space weather prediction.

• Education tool for teaching about the space environment.

MANDATE: in the National Space Weather Program plan


Modeling the May 12, 1997 CME

Uses:• MAS coronal MHD model with full energy equation• ENLIL heliospheric MHD model

Method:• Start with a potential magnetic field• Relax (for ~ 3 days) to develop a solar wind solution• Energize the active region via shearing flows or emerging flux• Apply converging motions towards the neutral line (with diffusion) to cancel flux at the neutral line.

Jon Linker, Roberto Lionello, Zoran Mikic, Pete Riley, Viacheslav Titov (SAIC), and Dusan Odstrcil (U. Colorado)


May 12, 1997 CME: Comparison of the Simulated Pre-event Corona with Observations

-2 -1.2 -.4 .4 1.2 2 2.8 0 1 2 3 4 Log10(DN/s)

EIT 171ÅEIT 195Å

EIT 284Å SXT (composite)

-1 -.1 .8 1.7 2.6 3.5

EIT 195Å

EIT 195Å EIT 284Å SXT (composite)


Evolution of magnetic field during May 12 eruption


• CME flux rope starts out connected to active region.• As CME propagates, it reconnects with nearby open field andloses connection with the active region


Coupling coronal model to ENLIL heliospheric model shows propagation of magnetic cloud to Earth, reaching Earth in 3 days 20 hours.

Heliospheric magnetic field evolution during May 12 eruption.

Inner boundary of ENLIL at 20 RSun

Earth(not to scale)


Lessons learned from modeling the May 12, 1997 CME

Modeling CMEs with this fidelity has many challenging aspects:• Must capture many scales (e.g. transition region, active region)• Choose heating to match observed emission• Best way to energize active region - No vector magnetograms of sufficient quality

At the beginning of CISM, we thought we knew how to do all of this: • This case has made us rethink nearly all of our previous assumptions• We have developed significant new capabilities


WSA-ENLIL: 2008 January 6-15 – Flow Velocity


WSA-ENLIL-Cone Model of CME’s: 2007 January 24-25


How does solar wind plasma enter the magnetosphere? – using Multi-Fluid LFM Peter Damiano, John Lyon, Bill Lotko (Dartmouth Coll.)

• Plasma is split into two identical proton fluids (Blue and Red)– Initially 99.99% Blue fluid and 0.01% Red– At start of inflow, solar wind is flipped the other way

(0.01% Blue, 99.99% Red)– Watch how the Red fluid enters the magnetosphere.

• Standard solar wind conditions– 400 kms– N = 5/cc– IMF, Bz = -5 nT


View of Earth’s equatorial plane


View of Earth’s noon-midnight meridian plane


Validating global ionospheric models

J. Lei, Wenbin Wang, A. G. Burns, S. C. Solomon, M. Wiltberger, and others (NCAR/HAO)

Comparing model predictions with observations is the only way to see how well we can do. – Climatology– A particular event


Ionospheric Climatology Comparison


Comparison of CMIT model with Millstone Hill Incoherent Scatter Radar observations.

December 2006 Storm


Where do we go from here?

“Focus on the ability to predict one day in advance our four most important space weather impact areas:

• Solar flares

• Solar protons

• Large geomagnetic storms

• Ionospheric disturbances”

Terry Onsager


Solar Flares• Solar Flares are the result of rapid reconfiguration of

unstable coronal magnetic fields. • They are associated with magnetically complex active

regions.• Currently measurements of photospheric/coronal

magnetic fields are insufficient in spatial resolution and accuracy/reliability to allow real time modeling – even post-event modeling is very difficult.

• Forecasting of flares 1-day in advance must rely for some time on empirical probability estimates based on the observed characteristics of complex active regions.


Overview of Solar X-rays and Solar Protons: December 2006 storms

X-rays(flares)

Proton’s

X flareM flareC flare

From: Uccellini et al, NOAA/NCEP Memo on NOAA Region 10930


Solar Protons• Solar protons are generated by flares and by shock

waves in the corona and solar wind.• They travel easily along magnetic fields, so arrive

promptly when the generation site is connected magnetically to Earth.

• The most energetic particles are generated closer to the Sun.

• Solar Proton events often last several days. • Forecasting onset of solar proton events is similar to

forecasting flares – with added information from knowing magnetic connectivity, which can be modeled.

• Following onset of an event, the subsequent evolution of the event can be modeled.


Large Geomagnetic Storms• Large geomagnetic storms are caused by

coronal mass ejections (CME’s) that impact Earth (medium-sized storms can also be caused by high speed solar wind streams).

• CME’s take 2-3 days to reach earth, therefore a 1-day forecast is possible if they can be accurately identified and characterized soon after they are ejected.

• Currently estimates of size and speed are possible. Magnetic structure of a CME is beyond current real-time observations/models.


The properties and structure of the preexisting solar wind are important to accurately predict CME arrival time.

The arrival time of a CME as predicted by the cone model varies ±1 day depending on the speed of the solar wind into which it is ejected.(from Case et al., 2008)


Large Geomagnetic Storms

• Predicting onset and approximate strength of a geomagnetic storm 1-day in advance is doable

• Predicting the temporal and spatial details of a storm requires knowledge of the magnetic structure of the CME – observing and modeling this is not currently possible in real-time.

• Modeling on a large range of scales is difficult:

1 AU ≈ 200 RSun ≈ 2 x 104 REarth


Ionospheric Disturbances

Ionospheric disturbances can be caused by:

• Ionospheric response to geomagnetic storms

• Direct ionospheric response to solar radiation (X-ray flares, solar protons)

• Internal instabilities (e.g. equatorial Spread-F)

The thermosphere is very weakly ionized – 1 part in a million to 1 part in a trillion depending on altitude.


One-day forecasts of Ionospheric Disturbances

• Ionospheric memory is short (several hours), so for a 1-day forecast, data assimilation doesn’t help.

• Details of high latitude (magnetospheric) forcing requires detailed knowledge of the solar wind 1-day in advance.

• Predicting general conditions, e.g. a geomagnetic storm, certainly is possible.

• Predicting non-storm low and mid latitude conditions is possible with a thermospheric circulation model such as TIEGCM.


Conclusions• Model capabilities are advancing quickly, but the

remaining problems are significant. • Models can only be as good as the observations that

drive them. • Models can provide guidance to probabilistic forecasts.• For a 1-day forecast, current observational/modeling

capabilities suggest most is to be gained from a time dependent heliospheric model with CME propagation.

• Currently Geospace models are best for short-term

(1 hour) forecasts and nowcasts.


Using CISM Models as Education tools: The CISM Space Weather Summer School

Students and faculty working at the CISM Summer School – a two-week school held each year. This year:July 21 - Aug 1, 2008


Models Currently at CCMC– Component and baseline models:

MAS ENLIL WSA PFSS

– Coupled models: CORHEL with Cone, CMIT (ad hoc or InterComm coupled)

Delivering Models to Researchers via CCMC


Delivering Models to Operations via NOAA/SEC Moving research models into operational space weather forecasting through our partnership with the NOAA Space Environment Center

Models currently being transitioned:Solar Heliosphere: WSA-ENLIL Geospace: CMIT 2.0

The Operations Room, NOAA/Space Environment Center, Boulder Colorado


CISM Mission (2002)

• Introduce Sun-to-Earth Community Models into space physics

• Introduce physics-based numerical models into space weather operational prediction and forecasting

• Imbue the notion that sun-earth science is a single unified discipline

overview

Documents

enlil heliospheric model

space weather prediction

complex space environment

solar wind plasma

solar wind solution

transition region

potential magnetic field

blue fluid