Data-driven Simulation And Result Evaluation Of The Coronal And Heliospheric Ambient Solar Wind

Three dimensional magnetohydrodynamics(MHD)solar wind model is a critical tool for ambient solar wind study and prediction.It is able to obtain the distribution of physical parameters for the ambient solar wind in the entire corona and interplanetary space,thus makes up the limited coverage of spacecraft in-situ data in time and space.What's more,it is also a key infrastructure for operational applications to make the transition from empirical solar wind forecast to numerical solar wind forecast.In this dissertation,a data-driven ambient solar wind model is established based on a flexible corona-interplanetary coupling approach.The MHD data-driven simulation of the ambient solar wind may be carried out in the interplanetary space alone,with the help of an empirical coronal model.The MHD data-driven simulation from the sun to 1AU is also realizable by the coupling of the MHD coronal model and the MHD interplanetary model.Employing the two working modes of the model respectively,the evolution of solar wind during Year2008 is simulated.The evolution of the large scale structures is analyzed,and the simulation result at L1 point is evaluated quantitatively.The model developed in this dissertation is based on the CESE-HLL 3D MHD solar wind model.Here,a variety of improvements have been made to the model,including(1)the adjustment of the coverage of the near-sun domain and the off-sun domain,which enables both domains to work in independent mode and coupling mode;(2)the realization of the projected normal characteristic boundary condition for the boundary treatment of the coronal model,which makes it possible for the coronal model to carry out data-driven simulation;(3)the implementation of the data-driven method for the interplanetary model by using output from an empirical coronal model;(4)an update of the parallel computation realization for the coronal model,which enable the parallelization in ? and ? direction,which improves the parallel efficiency and enables simulation with higher grid resolution and larger computational scale;(5)replacing the HLL solver of the interplanetary model with the less diffusive HLLD solver,which improves the performance of the numerical method.For the development of ambient solar wind model,one critical issue is the quantitative evaluation of model's ability.By comparing model results with observation and calculating statistical metrics,it is able to objectively measure the ability of the solar wind model in reproducing solar wind observation.By analyzing these metrics,model researchers may understand the strengths and weaknesses of the model,and find possible means to eliminating or reducing these weaknesses.In this dissertation,a set of automatic evaluation methods are established for the solar wind model,which have the following functions:(1)generate the time series of simulation results at the position of a certain spacecraft,following preset orbital configurations and temporal resolution.This process is carried out automatically as the model is running and no post-processing is needed;(2)calculate the statistical metrics,such as the mean square error and the correlation coefficient between observation data and simulation results;(3)generate the comparison of parameter distributions between the simulation results and observation data;(4)automatically identify the large scale changes of the interplanetary magnetic field polarity and evaluate the quality of the interplanetary magnetic field simulation result;(5)automatically identify the stream interaction regions and evaluate the quality of its simulation result.Using the empirical corona-MHD interplanetary coupling mode,the data-driven simulation of the interplanetary ambient solar wind during Year 2008 is carried out,which employs GONG's 6-hour photospheric magnetic synoptic maps as input.The evaluation results indicate that our model satisfactorily reproduces the large scale structure of the ambient solar wind during Year 2008.The correlation coefficient between observed and simulated speed is higher than 0.6.The strength of simulated interplanetary magnetic field matches observation well.All interplanetary magnetic field reverses and 82.76% of the stream interaction regions are captured by our model.The errors in predicting the arrival time of these two structures are about 1day.Using the MHD-coronal-MHD-interplanetary coupling mode,the data-driven simulation of the ambient solar wind from solar surface to the Earth's orbit is carried out for Year 2008,which employs GONG's 6-hour photospheric magnetic synoptic maps as input.The simulation results are analyzed and quantitatively evaluated by comparing the modeled output with remote-sensing EUV and pB observation of the corona and in-situ observation at the Sun-Earth L1 point.The analysis demonstrates that our model reproduces the main pattern and the evolutionary feature of large scale coronal structures,including the shapes and distributions of the coronal holes,and the positions and shapes of helmet streamer and pseudostreamers.From the simulation result,we find that the height of the pseudostreamer X point is positively correlated with the distance of the coronal holes connected by the pseudostreamer.During Year 2008,the helmet streamer belt is found to have a net southward displacement from the equator while the pseudostreamer belts are biased to the northern hemisphere.Both helmet streamer belt and pseudostreamer belts exhibit a general trend of flattening towards the equator during most time of the year.The evaluation of the modeled results at L1 point shows that the general structures can be generated by the model,and the speed is the best among the solar wind parameters reproduced,which has a correlation coefficient of 0.7.The model captures 96.30% of the interplanetary magnetic field reverses and 74.19% of the stream interaction regions.The errors in predicting the arrival time of these two structures are about 1 day.