COMPOSITE OF PLAGE AREAS OVER COMPOSITE OF PLAGE AREAS OVER COMPOSITE OF PLAGE AREAS OVER COMPOSITE OF PLAGE AREAS OVER THE ENTIRE 20TH CENTURY THE ENTIRE 20TH CENTURY THE ENTIRE 20TH CENTURY THE ENTIRE 20TH CENTURY Theodosios Chatzistergos, Ilaria Ermolli, Sami K. Solanki, and Natalie A. Krivova Space Climate Symposium 7, Canton Orford, 09/07/2019 1
Total solar irradiance TSI: spectrally integrated solar radiative flux at 1 AU Space-based measurements of TSI Measurements differ in β’ Absolute scale β’ Magnitude of variations Composites of TSI Composites show different long-term trends Solanki and Fligge, 2002 2
Ca II K as proxy to magnetic field plage network SDO/HMI magnetogram Rome/PSPT Ca II K π· = π 1 + π 2 [log( πΆ πππ /π)] π 3 3 Chatzistergos et al., 2019a
Available plage area series Arcetri Results vary due to the use of: Kodaikanal McMath-Hulbert β’ Different archives Mt Wilson β’ β’ Different data sample Different data sample Rome/PSPT β’ β’ Different normalisation area Different normalisation area Sacramento Peak β’ β’ Unaccounted disc ellipticity Unaccounted disc ellipticity β’ β’ Different segmentation methods Different segmentation methods β’ β’ Not accurate removal of CLV Not accurate removal of CLV and artefacts and artefacts Kodaikanal Kuriyan et al., 1982 Mt Wilson Foukal 1996 San Fernando Worden et al., 1998 Preminger et al., 2001 Ermolli et al., 2009a Ermolli et al., 2009b Tlatov et al., 2009 Chatterjee et al., 2016 Priyal et al., 2017 Singh et al., 2018 NSGD 4
Our corrections and calibration Automatic process to: Arcetri 16/05/1957 Kodaikanal 23/09/1913 β’ Determine solar disc radius β’ Identify QS CLV and artefacts β’ Relate QS CLV from historical data to QS CLV from modern data to construct CC β’ Calibrate images and compensate for limb-darkening Overexposure Characteristic curve 1 π Solarisation = log 10 (CC) Linear part Depends on: β’ Plateβs emulsion β’ Observation Underexposure conditions β’ Development Fog level β’ Storage β’ Digitization = log 10 (π½π’) Mt Wilson 04/03/1982 Kyoto 17/04/1956 5 Chatzistergos et al., 2018
Methodβs accuracy Added randomly with different severities on 2000 Rome/PSPT images PSPT observations Artefacts CC Synthetic data CLV Relative differences between original and Relative difference in disc fractions between calibrated contrast images as a function of original and calibrated contrast images as a time. function of time. Our calibration Uncalibrated, imposed CLV Adapted segmentation Uncalibrated, our CLV to minimise residuals Uncalibrated, Priyal CLV Uncalibrated, Worden CLV Error in plage areas less than 0.01 N No need to arbitrarily adjust segmentation No dependence of error on activity parameters! 6 Chatzistergos et al., 2019b Chatzistergos et al., 2018
Plage coverage over the 20 th century Meudon Mt Wilson 0.015nm 0.02nm Arcetri Kodaikanal 0.03nm 0.05nm Chatzistergos et al., 2019b 7
Archivesβ inconsistencies Kodaikanal Mt Wilson Change of grating on 21/08/1923 19/04/1969 19/04/1969 16/11/1990 15/03/1921 24/08/1949 16/12/1980 8 Chatzistergos et al., 2019b
Relations between archives Chatzistergos et al., 2019b 9
Plage coverage over the 20 th century First plage area composite over 1893-2018 Chatzistergos et al., 2019b The composite is available at: http://www2.mps.mpg.de/projects/sun-climate/data.html http://vizier.u-strasbg.fr/viz-bin/VizieR?-source=J/A+A/625/A69 10
Summary β’ Historical Ca II K can provide valuable information on solar surface magnetism, needed by irradiance/climate studies β’ We developed a process to photometrically calibrate and segment historical Ca II K SHGs β’ Relatively small errors estimated from synthetic images β’ No need for adapting segmentation parameters β’ Derived plage areas depend on employed bandwidth β’ We produced the first plage area composite spanning the entire 20 th century from 9 Ca II K archives Thank You! 11
Recommend
More recommend
