000004360 001__ 4360
000004360 005__ 20191219230959.0
000004360 037__ $$aPOSTER-2019-0119
000004360 100__ $$aBechet, Sabrina
000004360 245__ $$aData homogenization for a network of ground-based synoptic imaging telescopes
000004360 260__ $$c2019
000004360 269__ $$c2019-09-19
000004360 520__ $$aGround-based synoptic imaging telescopes are used nowadays for continuous whole-disk monitoring of solar activity as a lightweight patrol for space weather forecasts. However, one single station has limited time coverage due to the night-day cycle and variable observing conditions. This can be mitigated by considering a network of stations at different geographical locations. However, before such multi-station data can be merged, we need to homogenize images from different instruments (different optical set-up, bandpasses, CCD), or from identical instruments at different sites (different observing conditions, slightly different spectral bandpass, ..) such as the planned SOLARNET-SPRING network. This is a first mandatory step towards more advanced products such as synoptic maps and solar feature or event detection. We present the ongoing development of such homogenization algorithms for full disk images at three different wavelengths (white-light, Ca II K and H-alpha). In particular we illustrate the correction of geometrical differences (disk re-centering, intensity normalization) as well as radial and non-radial photometric inhomogeneities (limb darkening, atmospheric transparency, stray-light) on synoptic images taken at the Royal Observatory of Belgium (ROB) and at the Kanzelhöhe Observatory (KSO). 
000004360 594__ $$aSTCE
000004360 6531_ $$aUSET
000004360 6531_ $$ahomogenization
000004360 6531_ $$asynoptic images
000004360 700__ $$aClette, Frédéric
000004360 773__ $$tML-Helio, Machine Learning in Heliophysics
000004360 8560_ $$fsabrina.bechet@observatoire.be
000004360 85642 $$ahttps://ml-helio.github.io/
000004360 980__ $$aCPOSTER