We present a comprehensive solar flare forecast model with a probability and a statistically significant range of daily peak X-ray flux. For this, we consider μ-corrected total unsigned radial magnetic fluxes from the SOHO/MDI and SDO/HMI, and flare lists from GOES from 1996 to 2021. Our model predicts two types of forecast results when a magnetic flux of an active region (AR) is given. First, using a relationship between magnetic fluxes and flaring rates, a probability of C1.0 or greater flares and a probability of M1.0 or greater flares within a day are predicted respectively. Second, a mean (μ) and standard deviation (σ) of daily peak X-ray fluxes are given from a historical distribution between magnetic fluxes and daily peak X-ray fluxes. Using the mean and standard deviation, we provide the statistical range of possible flare sizes. We verify two forecast results by using various performance metrics and investigate the performance depending on the climatology event rate. Based on the metric values, our model can give a better performance than the climatology forecast. Solar flares are considered to be caused by specific triggers and physical mechanisms that have not yet been precisely identified. In addition, there is another perspective that the size of the flare that will occur due to a trigger is close to random because the flaring loop is in a self-organized critical state. Our model can give the simplest forecasting results considering these two perspectives.

We introduce the two-dimensional spectral observations of solar flares using the Solar Tower Tele-scope of Nanjing University, China. In particular, we introduce three typical events and the methods used to analyze the data. (1) The flare of November 11, 1998, which is a limb flare. We derive the temperature and density within the flaring loop using non-LTE calculations. The results show that the loop top may be hotter and denser than other parts of the loop, which may be a result of magnetic reconnect ion above the loop. (2) The flare of March 10, 2001, which is a white-light flare that shows an emission enhancement at the near infrared continuum. We propose a model of non-thermal electron beam heating plus backwarming to interpret the observations. (3) The flare of September 29, 2002, which shows unusual line asymmetries at one flare kernel. The line asymmetries are caused by an upward moving plasma that is accelerated and heated during the flare development.

We have developed and tested a CCD camera (100 × 100 pixels) system for observing Ha images of the solar flares with time resolution> 25 msec. The 512 × 512 pixels image of CCD camera at 2 Mpixels/sec can be recorded at the rate of more than 5 frame/sec while 100 × 100 pixels area image can be obtained 40 frames/sec. The 100 × 100 pixels image of CCD camera corresponds to 130 × 130 arc - sec2 of the solar disk.

Non-LTE calculations, with the non-thermal ionization effects included, indicated that for electron bombardment, the Hα line is widely broadened and shows a strong central reversal. Significant enhancements at the line wings of Lyα and Lyβ are also predicted at the beginning of the impulsive phase of flares. For the proton bombardment, no strong broadening and no large central reversal are expected. However, due to proton-hydrogen charge exchange, the enhancements at the red wings of Lyα and Lyβ lines at the early impulsive phase of flares are significant. Our results show that the electron beam can also in some cases generate visible and UV continuum emission in white-light flares. However, at the onset phase, a negative flare may appear within several seconds, due to the increase of the H- opacity. Another spectroscopic signature of energetic particles, i.e. the impact polarization of atomic lines, is also mentioned.

The magnetic reconnection mechanism is a primary candidate for "flare" processes in solar coronal regions. Numerical simulations of two-dimensional magnetic reconnection are carried out for four different cases: (1) adiabatic condition with constant resistivity, (2) adiabatic condition with temperature-dependent resistivity, (3) energetics with radiation loss and constant resistivity and (4) energetics with radiation loss and temperature-dependent resistivity. It is found that the thermal instability prompts the magnetic reconnection process, thus increasing the conversion rate of magnetic energy into kinematic energy of the fluid. We demonstrated that the observed microflares can be accounted for by our magnetic reconnection models, when the effects of the radiation loss and the temperature-dependent resistivity are taken into account.