c11bb51b0933bbd024674b79bf9773a1.ppt
- Количество слайдов: 16
Summary of: Modeling the May 12 1997 event using HAFv 2 and HHMS 3 D MHD models: Capabilities and Possibilities for Space Weather Prediction • • To show usefulness of models beyond the academic, they must be tested in the operational environment. Both the specification of inputs (from data) and understanding of physics of the models are important for this task. Using the improved input velocity, both HAFv 2 and HHMS models would have given good predictions (transit times to within 3 -8 hours. The original was 31 hours early!). Using an improved ambient state further improves agreement of the observations to the computed shock travel time and simulated solar wind parameters. SHINE, Zermat, Utah July 31 -August 4, 2006
Modeling the May 12 1997 event using HAFv 2 and 3 D MHD models: Capabilities and Possibilities for Space Weather Prediction Zdenka Smith 1, Tom Detman 1, 2, Wei Sun 3, Murray Dryer 1, 2, Charles S. Deehr 3 and Ghee Fry 2. 1: NOAA / SEC, 325 Broadway, Boulder, CO, 80305, USA 2: Exploration Physics Inc, Huntsville, AL, 35806, USA 3: Geophysical Institute, University of Alaska, Fairbanks, AK 99775 SHINE, Zermat, Utah July 31 -August 4, 2006
Abstract: The accuracy of predicted Sun to Earth transit times for traveling interplanetary disturbances depends both on how well the physics is modeled and how well we are able to determine the proper inputs from available observations. We consider two solar wind prediction models: the 3 D kinematic HAFv. 2 and the 3 D MHD HHMS. Looking back at the early Fearless Forecasts (predictions made using inputs available in near-real-time with HAFv. 2), we show the results have improved over time due to improvement in both the models and understanding of the inputs. It is now well known that the evolution of interplanetary disturbances (often led by shocks) can be complicated by their interactions with other shocks and/or inhomogeneities in the solar wind. The May 12 1997 event provides an excellent test case because it was a “well isolated” case with the observation of a large flare (in optical and X-ray), a large halo CME, and metric type II radio bursts. By “well isolated” we mean that it occurred during a time when the Sun had been quiet, and was quiet afterward. We apply both models to this May 12 th event. SHINE, Zermat, Utah July 31 -August 4, 2006
1. THE HAFv. 2 MODEL The HAFv. 2 (Hakamada-Akasofu-Fry version 2) is a 3 -D model that simulates the solar wind in the heliosphere, both background and transients, using solar data for input (Fry et al. , 2003, 2004). It is a “modified kinematic” model: kinematic in that it kinematically models the non-uniform flow and solar wind magnetic field, plasma speed and density as fluid parcels projected outward from a Sun-centered spherical surface. It is “modified” in that it adjusts the flow for stream-stream interactions as faster plasma catches up with the slower. The non-uniform, background solar wind (illustrated in Figure 1) conditions are initialized by daily photospheric, line-of-sight magnetograms that are routinely measured and combined into daily synoptic maps, following the method of Arge and Pizzo, 2000, and Arge et al. 2004. The maps serve as boundary conditions in the spherical-harmonic expansion of a potential field model that specifies coronal conditions on a “source surface”, usually at 2. 5 Rs. This surface is assumed to be the location where coronal field lines become radial as the solar wind kinetic energy dominates over the magnetic field energy. Transient events are modeled by superposing pulses onto the background: These are discussed in Section 3. An example, using Halloween 2003, is shown in Fig. 2. SHINE, Zermat, Utah July 31 -August 4, 2006
Figure 2: Example of: predicted IMF, in the ecliptic plane out to 10 AU distorted both by transient events and the varying background. (red lines indicate field lines directed outward from the Sun and blue lines show the inward-directed field lines ). Figure 1: Example of background solar wind modeled by the HAFv. 2. The heliosphere current sheet (green surface) is included in the model. The inner planets, in the ecliptic plane, are also shown. From Fry et al. (2006), with thanks to Ed Hoch of the Geophysical Institute, University of Alaska, Fairbanks, Alaska SHINE, Zermat, Utah July 31 -August 4, 2006
2. THE HHMS: The Hybrid Heliospheric Model System (HHMS) is a Sun to Earth system of coupled models whose main goal is to create a real-time 3 D MHD based system to aid in the operational forecasting of geomagnetic activity It consists of two physics based models, (the Source Surface Current Sheet model and the Interplanetary Global Model-Vectorized) and the two empirical models (the Interface Module and the Linear Prediction Filter) (Detman et al. , 2006). Analogously to HAFv 2, the HHMS simulates both the background and transients in the solar wind, using solar data for input. The HHMS system is driven by a sequence of photospheric magnetic maps composed from daily magnetograms. An empirical relationship between magnetic flux tube expansion factor and solar wind speed at 0. 1 AU is a key element of the system. This model also gives a predicted time series of solar wind MHD parameters at the location of Earth (or any other location) for easy verification against Omni, Wind or ACE satellite data. The predicted solar wind at Earth is used as input to the second, data based, empirical model to predict the geomagnetic Ap index. An example (using simulations of background only) is shown in Figure 3 SHINE, Zermat, Utah July 31 -August 4, 2006
Figure 3 Comparison of the predicted (red) and observed (blue) Ap geomagnetic storm index. The green is the difference. SHINE, Zermat, Utah July 31 -August 4, 2006
3. INPUTS for TRANSIENT EVENTS Both models simulate transient events which are introduced by superposing onto the background, input pulses at the flare location. The pulses are characterized by speed, Vs, (coronal shock velocity from metric type II and/or halo/partial halo CMEs) and duration (from duration of soft X-ray flares) (Fry et al. , 2003 and Smith et al. , 2004). There is often more than one estimate of shock speed reported in near real time The data available for the May 12 th event are shown in next four Figures 4 -7 (from http: //cdaw. gsfc. nasa. gov/SHINE_Campaign/19970512/presentations/May_12_1997_summary. ppt) Fig. 4 GOES 1 -8Å X-ray Fig. 5 SOHO/EIT Flare, 195Å Fig. 6 CME onset 0430 -0500 UT, SHINE, Zermat, Utah July 31 -August 4, 2006 VCME =600 km/sec, (Plunkett et al, '98)
Type II radio burst from WIND SHINE, Zermat, Utah July 31 -August 4, 2006
INPUTS for the May 12, 1997 Event The Fearless Forcast (FF) for this event (FF#3) used data collected in near real time (and issued in the Edited Events 1). Additional shock speed estimates** were available from another radio-observatory 2 and from the LASCO CME 3. These differed substantially, so we reran the models using these additional speed estimate, given below as 3 x and 3 y. NNNN YYYY MMD D HHM M LAT LON VS TAU Vsw FF# year Mo. Dy UT deg km/sec hr. mn km/sec 3 1997 512 516 N 21 W 08 1400 230 300 3 x 1997 512 516 N 21 W 08 700 230 300 3 y 1997 512 516 N 21 W 08 600 230 300 From the Fearless Forecasts, (FF), this was the 3 rd event. the first line is in the form used in the FF: it gives the date, time and: FF# this was the third FF event: 3 x and 3 y give the retrospective runs with the two velocities given in the Geophysical Data Reports. LAT/LON =helio-latitude/longitude of the source Vs = shock velocity TAU = duration of the solar event Vsw = background solar wind speed, this has been replaced with the SHINE, Zermat, Utah automated background solar wind from Arge at al? ? ? • **Could you run HAFv 2 *with Vs=700 km/se and with Vs=600 km/sec? (700=type II speed from Hiraiso and 600= Vcme, also from July 31 -August 4, 2006 subsequent analysis from check Nat’al Geophys Data reports 1997? )
4. Simulated Shock Arrivals, Comparison to Data • SHINE, Zermat, Utah July 31 -August 4, 2006
Figure 8: Simulation with HAFv 2, using input obtained in near real time but with the updated estimates of input velocity, Vs Vs=600 km/sec (run #2) Vs=700 km/sec (run #1) SHINE, Zermat, Utah July 31 -August 4, 2006
Figure 9 a. Simulation with HHMS, Vs=600 km/sec (run #5) Inputs are those obtained in near -real-time but with the updated estimates of input velocity, and with an improved ambient state. BC= Boundary condition near Sun (0. 1 AU) SHINE, Zermat, Utah July 31 -August 4, 2006
Figure 9 b. Simulation with HHMS, Vs=700 km/sec (run #4) Inputs are those obtained in near -real-time but with the updated estimates of input velocity, and with an improved ambient state. BC= Boundary condition near Sun (0. 1 AU) SHINE, Zermat, Utah July 31 -August 4, 2006
• 5. SUMMARY • To show usefulness of models beyond the academic, they must be tested in the operational environment. • Both the specification of inputs (from data) and understanding of physics of the models are important for this task. • Using the improved input velocity, both HAFv 2 and HHMS models would have given good predictions (transit times to within 3 -8 hours. The original was 31 hrs early!). • Using an improved ambient state further improves agreement of the observations to the computed shock travel time and simulated solar wind parameters. SHINE, Zermat, Utah July 31 -August 4, 2006
References Arge, C. N. and V. I. Pizzo (2000), Improvement in the condition of solar wind predictions using near-real time solar magnetic field updates, J. Geophys. Res. 105, 10, 465 -10, 479. Arge, C. N. , J. G. Luhmann, D. Odstrcil, C. J. Schrijver and Y. Li, Stream structure and coronal sources of the solar wind during the May 12 th, 1997 CME, 2004, J. Atmos. Solar Terr. Phys. , 66, 1295 -1309. Detman, T. , Z. Smith, M. Dryer, C. D. Fry, C. N. Arge, and V. Pizzo (2006), A hybrid heliospheric modeling system: Background solar wind, 2006, J. Geophys. Res. , 111, A 07102, doi: 10. 1029/2005 JA 011430. Fry, C. D. , M. Dryer, W, Sun, T. R. Detman, Z. Smith, C. S, Deehr. , C. -C. Wu, S. -I. Akasofu, and D. Berdichevski, (2004), Solar observation-based model for multi-day predictions of interplanetary shock and CME arrivals at Earth, IEEE Trans. Plasma Sci. , 32(4), Part I of III, 1489 -1497. Fry, C. D. , M. Dryer, Z. Smith, W. Sun, C. S. Deehr, S-I. Akasofu (2003) J. Geophys. Res. , 108, A 2, 1070, doi: 10. 1029/ 2002 JA 009474. Fry, C. D. , T. R. Detman, M. Dryer, Z. Smith, W. Sun, C. S. Deehr, S. -I. Akasofu, C. -C. Wu and S. Mc. Kenna. Lawlor, (2006) Real-Time Solar Wind Forecasting: Capabilities and Challenges, J. Atmos. Solar Terr. Phys. , in press. Mc. Kenna-Lawlor, S. , M. Dryer, M. D. Kartalev, Z. Smith, C. D. Fry, W. Sun, C. S. Deehr, K. Kecskemety, and K. Kudela, (2006) Real-time Predictions of the Arrival at the Earth of Flare-generated Shocks during Solar Cycle 23, J. Geophys. Res. , in press. Smith, Z. , T. R. Detman, M. Dryer, C. D. Fry, C. -C. Wu, W. Sun, C. S. Deehr (2004) A verification method for space weather forecasting models using solar data to predict arrivals of interplanetary shocks at Earth, IEEE Trans. Plasma Sci. , 32(4), Part I of III, 1498 -1505. Smith, Z. , W. Murtagh, T. R. Detman, M. Dryer, C. D. Fry, C. -C. Wu, (2003)Study of solar-based inputs into space weather models that predict interplanetary shock-arrivals at Earth, Proceedings of ISCS, Tatranska Lomnica, Slovakia, 23 -28 June 2003, ESA, 2003 • SHINE, Zermat, Utah July 31 -August 4, 2006