Daily X-ray background measurements are available for the last seven days at the Space Weather Prediction Center SWPC. The full dataset can be found here. No scaling factor is necessary for this data, as it is in physical units. To determine the daily background, hourly and 8-hour minima of the 1-minute averages are used. The 1-8 Å (long) and 0.5-4 Å (short) background fluxes are shown below, from February 2017 until October 2022. The daily average flux can also be toggled on.

To select an event of interest from a long-duration spectrogram or pixel data file, one may want to reduce the file to a smaller size containing only that time period. Similarly, one might want to combine observations to cover a longer time period.

Image-to-image translation has been sucessfully demonstrated for many cases, such as colorizing black-and-white images or filling in an object from its edges. Tensorflow has a detailed tutorial using Isola et al’s Pix2Pix. It uses a generator with a U-Net-based architecture and a discriminator using a convolutional PatchGAN classifier.

Previously, the reprojection of AIA maps to the Solar Orbiter reference frame has been shown. The method of reprojecting STIX images, reconstructed through one of many algorithms such as back projection, forward fit, CLEAN, or expectation maximization, is very similar.

NASA’s X-ray Spectral Fitting Package lacks one of the key solar flare emission models, that of thick-target bremsstrahlung. In sswidl’s OSPEX, this model is implemented as thick2.pro. Using sunxspex’s emission module, the thick-target model can be adapted for use with XSPEC or PyXspec.

Deep Extreme Cut DEXTR uses extreme points of an object in an image as input to perform image segmentation. Here are some results of using the model trained on PASCAL VOC to segment HMI and AIA images (after conversion to RGB jpg).

Pre-trained model applied to HMI data

STIX spectrogram FITS files are not natively compatible with NASA’s X-ray Spectral Fitting Package. However, work is ongoing such that this powerful, free software can be used with STIX spectra, should sswidl’s OSPEX prove insufficient (ex., OSPEX can only fit with chi-squared minimization and one spectrum at a time).

Reprojecting SunPy maps from one reference frame to another is most commonly done with full-disk images. The built-in implementation of reproject_to() is also capable of reprojecting submaps or cutouts; however, unless the output WCS are carefully adjusted this is just as slow as reprojecting a full-disk image.

The as-flown attitude of an orbiting spacecraft affects its local coordinate system. A roll angle correction must be applied to imaging data in order to recover the proper coordinates. Spacecraft attitude data, both predicted and as-flown, are stored in the SPICE kernels.

The SPICE kernels used to get spacecraft or other oribiting body positions does not contain information for the SDO; therefore it is common to assume an Earth observer when considering coordinates associated with SDO/AIA data. Exact SDO location information is available through the FITS file headers. It’s inconvenient to download a large, fulldisk SDO image just to get its header, which is the way almost everyone knows how to access these keywords. There are lesser-known and/or used ways of getting header information alone, through dmrs or JSOC records queries.

Solar image data has its own special set of colorscales associated with it. Certain instruments and wavelengths are by default presented in these scales, making their identification quick by the trained eye. SunPy has implemented most of these colormaps in its visualization.colormaps module. To use them in PlotLy, simply convert the LinearSegmentedColorscales to rgb values and add intervals to use them as PlotLy colorscales.

Solar Orbiter’s STIX has observed over 1000 flare candidates as of September 2021. Of those, 358 are given locations by the Coarse Flare Locator. Joint visibility with Earth-orbiting SDO can be determined using ephemeris information from Solar Orbiter, contained in the Spice kernels, and the Earth. Code is here.