Session - Key solar observables for assessing long-term changes of the Geospace

Thierry Dudok de Wit, M. Haberreiter, A. Seppälä

Our understanding of long-term changes in the Heliosphere, and their potential impact on the Earth's climate, heavily relies on our ability to reconstruct past solar activity, on time scales of months to centuries. Most observables of past solar activity, such as the sunspot number, geomagnetic records, and cosmogenic isotope concentrations, are indirect, and thus rely on models to infer the state of the Sun. In recent years, major progress has been made in the collection of historic data, in their (re)analysis, thus shedding new light on how the Sun evolves on the long term.

We welcome contributions on the observables that describe the solar forcing of our Geospace, with a particular emphasis on long-term effects, and the impact on climate. Particular attention will be given to recent advances in the collection or understanding of these observables, but also important open issues, and roads for future improvement.

This session will have oral presentations (contributed and invited reviews), but we're also looking forward to going more in depth with an active and well-attended poster session. We welcome posters on important observables for climate research, including more technical and user-oriented issues such as data collection, data processing and the making of composites, how and where the data are available, how stable they are on the long run, etc.

Tuesday November 18, 11:00-13:00, auditorium Roger

Poster Viewing
Tuesday November 18, 11:00-11:20, area in front of auditorium Roger.
The posters will be introduced during the beginning of the oral part in auditorium Roger.


11:00 am Introduction and Poster Summary
11:20 am The New Sunspot Number: a Full Recalibration
  Clette, F1; Svalgaard, L2; Vaquéro, J  M3; Cliver, E  W4
  1Royal Observatory of Belgium; 2W.W. Hansen Experimental Physics Laboratory, Stanford University, Stanford, CA, USA; 3Departamento de Física, Universidad de Extremadura, Mérida, Spain; 4National Solar Observatory, Sunspot, NM, USA
  For any long-term reconstruction of Sun-Earth relations, beyond the last 50 years, the only direct solar observable over multiple centuries is the Sunspot Number, derived from visual sunspot counts. However, different versions of this primary index have been created, including the widely used Group Sunspot Number. Inconsistencies between those series have prompted a join effort initiated in 2011 in the framework of dedicated Sunspot Number Workshops, currently at their 4th edition.  Here, we will review our present refined knowledge of the construction of those fundamental composite series and the latest problems that have been identified over the entire 400-year series. We will describe the nature of various identified defects, the original data used to establish those inhomogeneities, as well as the resulting corrections that have been derived. The main corrections are rather large and lead to a 40% increase of Group Number values before 1880 and a 20% decrease of Sunspot Number values in the second half of the 20th century. Moreover, a variable 20% drift associated to the Locarno pilot station over the last 34 years was corrected by fully recomputing the recent part of the Sunspot Number series produced in Brussels since 1981. After those corrections, the past disagreements between the Sunspot Number and Group Number series are largely eliminated.  This end-to-end revision of the Sunspot Number brings fundamental changes to our current picture of solar activity over past centuries. In particular, the upward rise in activity to a Modern Maximum in the 20th century largely vanishes. We will also discuss the apparent long-term variations of the average number of sunspots per group before concluding on the very high correlation of the Sunspot Number with present modern solar indices. This gives an additional validation of this synthetic numerical index as a quantitative measurement of the past solar magnetic activity. This new fully modernized version of the Sunspot Number will soon be released on the WDC-SILSO Web site (
11:45 am From Maunder Minimum to the recent Grand Solar Maximum
  Lockwood, M1; Owens, M1; Barnard, L1
  1Reading University
  From historic geomagnetic data we can reconstruct annual means of the near-Earth interplanetary magnetic field, the solar wind speed and the open solar flux between 1845 and the present. Results using four different combinations of geomagnetic indices are very similar indeed and a full Monte-Carlo analysis of errors shows that even at the 2-sigma level they are remarkably small for all three parameters.  The open solar flux can be modelled using continuity equations with emergence rate based on sunspot  numbers but allowance for errors is required for the “Waldmeier” and “Wolf” discontinuities.  The results of both the reconstructions and the modelling show the average open solar flux did indeed double between 1900 and the modern era and has been declining rapidly since 1985.  In the Maunder minimum it was close to a quarter of its value in the recent grand maximum.  These results agree well with best estimates from cosmogenic isotope data which exhibit sufficient predictability to allow some "analogue forecasts" of the likely future of space weather conditions.
12:10 pm On the Influence of Energetic Particle Precipitation on the Climate System: the Challenges and Problems in Measuring Energetic Electron Precipitation Into the Atmosphere.
  Clilverd, M1; Rodger, C2
  1British Antarctic Survey; 2University of Otago 
  Significant progress has been made in the understanding of energetic particle precipitation and its effects on the middle and upper atmosphere. Model simulations now include the chemical and energetic forcing from a broad range of energetic particles, including electrons and protons. Studies have shown that significant forcing of the atmosphere can occur during periods of high geomagnetic storm activity, initially driven by processes occurring on the Sun. Energetic electron precipitation occurs at medium and high geomagnetic latitudes from the Van Allen radiation belts, geomagnetic substorms and the aurora. Proton precipitation occurs at high geomagnetic latitudes due to solar proton events and the aurora.   Although there is now more than a solar cycle of electron precipitation measurements from satellite, the observations of electron flux and energy is incomplete, and requires considerable thought in its analysis. In this talk we will identify the challenges and problems that need to be addressed when providing electron precipitation information for use in atmospheric models. These issues include proton contamination of the electron detectors on satellite, inaccurate measurement of electron flux due to limited detector viewing capabilities, and non-uniform precipitation into the atmosphere due to the configuration of the Earth’s magnetic field. These issues change in significance during the course of a geomagnetic storm, and change in significance depending on electron energy.
12:35 pm The Influence of Solar Variability Past, Present and Future, on North Atlantic Climate.
  Knight, J1; Jackson, D1
  1Met Office
  David Jackson (1), Jeff Knight (1), Adam Scaife (1), Sarah Ineson (1), Nick Dunstone(1), Lesley Gray (2), Mike Lockwood (3), and Amanda Maycock (4)  (1) Met Office, Exeter, UK, (2) NCAS-Climate, Atmospheric Oceanic and Planetary Physics, Department of Physics, Oxford University, Oxford, UK, (3) NCAS-Climate, Department of Meteorology, University of Reading, UK, (4) Centre for Atmospheric Science, Department of Chemistry, University of Cambridge, UK  In the last few years progress has been made in the physical understanding of observed links between solar activity and regional climate. Advances in climate modelling have resulted in the explicit inclusion of the Earth’s middle atmosphere in state of the art models, allowing potential effects of the variability of visible and ultraviolet (UV) solar radiation and energetic particles to be tested. Meanwhile, new observations of solar spectral irradiance from the SORCE satellite have raised the possibility that solar cycle UV variability may be larger than previously appreciated. Here, we present results of studies in which we drive a middle-atmosphere resolving climate model with this larger solar UV variability. We find that low (high) solar activity is associated with the negative (positive) phase of the North Atlantic Oscillation (NAO), the principle mode of year-to-year variability in European and eastern North American winter climate. Negative NAO conditions are associated with an increased incidence of easterly winds for Northern and Western Europe and hence colder and snowier winters, while positive NAO conditions are associated with westerly winds and mild, wet and stormy winters. The size of the signal is similar to that seen in observations, and implies solar cycle signals are useful for predictions of climate on seasonal and decadal timescales. The mechanism for the NAO response involves intricate stratospheric dynamical responses to latitude-dependent changes in ozone heating by solar UV, which can only by captured in a climate model which resolves the middle atmosphere. Further work has examined the role of ocean feedbacks in amplifying this tropospheric response. Statistical evidence for such a feedback is available from observations and gives a lag of 3-4 years for the maximum tropospheric response after the maximum solar forcing. This lag does not generally appear to be reproduced by current climate models. We discuss how this observational evidence may be a valuable way of assessing the relative strength of ocean-atmosphere coupling in the present generation of climate models. Lastly, the prolonged solar minimum during the transition between solar cycles 23 & 24, and the relatively low maximum activity of cycle 24, have increased suggestions of an end to a 20th century 'grand solar maximum' and a possible return to Maunder Minimum levels of solar activity during the 21st century. We illustrate the possible regional climate impacts that an associated projected decline in the UV flux would have and how this would relate to regional anthropogenic climate change.


The numbers of the posters on this page may differ from the numbering on the short overview without abstracts.
2   A New "Brussels" Sunspot Number: Full Recomputation of the last 34 Years
    Clette, F1; Lefèvre, L1
    1Royal Observatory of Belgium
    In the framework of a full revision of the multi-secular Sunspot Number series, we identified scaling changes in the most recent period (1981-2014), when the International Sunspot Number was produced in Brussels using the Specola Observatory (Locarno) as pilot station.  We first describe the full archive of individual sunspot counts collected by the worldwide network coordinated by the World Data Center SILSO. Based on this huge archive, we built global statistics of the relative scale of all 262 stations (personal k coefficients), which allowed to diagnose a large variable drift in the scale of the reference Specola observations. We describe the method and present the resulting scaling correction. Comparisons with parallel solar indices confirm the validity of this correction. An analysis of k coefficients of individual observers at the Specola Observatory further indicates that the drifts can be traced to the main reference observer.  As all raw data have been carefully preserved, we carried out a full recomputation of the Sunspot Number over the last 34 years, based on a new alternate reference. The latter is based on station quality criteria derived from our global statistical study. After correction, we observe some changes in the characteristics of the last 3 solar cycles, bringing them in better agreement with other short-term activity indices but also leaving disagreements that may reflect actual changes in the solar magnetic field production. We conclude on the new prospects opened by the entirely new software developed for this study, a versatile tool that forms the base of a new permanent quality control for future years.
3   60 Years of Solar Multi-Wavelength Radio Observations as Proxies for Solar Radiative Forcing
    Dudok de Wit, T1; Bruinsma, S2; Shibasaki, K3
    1LPC2E, CNRS/University of Orléans; 2CNES; 3Nobeyama Solar Radio Observatory
    The paucity of properly calibrated observations of solar activity is a critical problem in determining the climate impact of solar radiative forcing. Total irradiance observations started in the 1980s only, and their long-term trend is still subject to debate. Reconstructions going further back into the past mostly rely on the sunspot number, which suffers from being only loosely connected to the radiative output of the Sun. However, radio observations in the centimetric range have been made since the early 1950s at several wavelengths, and with highly stable receivers, thus providing a unique record of past solar activity. So far only the 10.7 cm has been routinely used, whereas the other wavelengths have been almost totally ignored.  We have collected these daily radio observations at five wavelengths (3.2 to 30 cm) into a single composite dataset. By using blind source separation, we are able to decompose the variability on time scales of months, and below, into three components; these can be readily ascribed to thermal bremsstrahlung, and to gyroresonant emissions. We find that the 30 cm radio flux is a better proxy for the specification of the upper atmosphere than the better known 10.7 cm flux, presumably because it has a stronger bremsstrahlung contribution. This is confirmed by comparing the modelled and the observed neutral density of the thermosphere, using the DTM2013 drag temperature model. When considering long-term variations, we find tiny but significant differences between the various wavelengths, which allow us to constrain the variability of the solar UV forcing since the early 1950s. Particular attention will be given here to the possible existence of long-term trends and their relevance to climate studies.
4   Making of Solar Spectral Irradiance Composites out of Multiple Datasets
    Scholl, M1; Dudok de Wit, T1; Haberreiter, M2; Kopp, G3; Kretzschmar, M1
    1LPC2E/CNRS, Universite d'Orleans; 2PMOD-WRC; 3LASP, University of Boulder; 1LPC2E/CNRS, Universite d'Orleans
    The reconstruction of a single solar spectral or total solar irradiance data set out of disparate and incomplete records is a challenging problem, and also an important one, because there is a need for having a single homogenous record with well-calibrated observations. The same problem also emerges in the making of climate records from proxy data.  Two different projects aim today at building such composite data sets. One is the FP7 SOLID project, which focuses on the reconstruction of the spectrally-resolved irradiance ( The other is an ISSI team, which addresses the total solar irradiance (  Here, we concentrate on the methodology for making such composites and present some preliminary results of the reconstructions. Traditional solutions rely on some averaging of the various observations; they become inadequate when the observations have data gaps, do not agree with each other, and are affected by noise. For that reason, we advocate a different and more general approach, which is a Bayesian one. This approach involves several steps, starting with the filling in of all the data gaps, followed by a multiscale decomposition. These steps will be detailed and a preliminary composite will be compared with model results.
5   Prediction of Amplitudes and Periods of Solar Cycles 24 and 25.
    Basurah, H1; Baqatyan, S2; Sharaf. , A2
    1King Abdulaziz University; 2KAU 
    Smoothed monthly mean Zurich relative sunspot numbers for 1749 to 2012 have been used to estimate amplitudes and periods of the 24 and 25 solar cycles, using the degree of similarity (h) of the solar cycles. We used three parameters; solar minimum (Rmin), ascending times (Ta), and effective duration (De), and found that the peak of cycle 24 to be around  8/2013 with amplitude of about Rmax = 78 ± 11 and to end around  9/2019, while the peak of cycle 25 to be around 8/2023 with amplitude of about Rmax = 116 ± 48 and to end around  12/2029.
6   First Results of Statistical Analysis Applied on Different Solar Spectral Irradiance Datasets Acquired from Different Satellites
    Chehade, W1; Weber, M1; Burrows, J  P1
    1Institut für Umweltphysik (IUP)
    The solar irradiance datasets are crucial input parameters to any atmospheric and climate modelling. The spectral and total solar irradiance (SSI and TSI) change on scales ranging from few days up to the 11-yr sunspot cycle. In the existing solar datasets from different instruments, the magnitude of the variability has large uncertainties. Moreover the irradiance variations are not uniformly distributed over the solar spectrum. The variability of the ultraviolet (UV), visible (VIS), and infrared (IR) wavelengths is most important for climate studies. The first European comprehensive SOLar Irradiance Data exploitation (SOLID) project aims to provide uniform data sets of modelled and observed solar irradiance data from the beginning of the space era to the present for the time period of around thirty years. This will help in reducing the uncertainties in the long term irradiance time series. In the presented work, a continuous wave analysis is applied to the different datasets to remove the long term variability and compared the results to the Mg II index and our first results are shown here.
7   Study of Solar cycle Evolution by a Recurrence Quantification Analysis of Solar cycle Indices
    Stangalini, M1; Ermolli, I1; Consolini, G2; Giorgi, F1
    Study of solar cycle with FFT-based techniques applied to solar activity indices has been exploited to identify the periodicities in the data modulated by the solar cycle, and in relating solar variability to other phenomena, e.g. the variability observed in the heliosphere and near Earth space environment. However, FFT-techniques suffer from severe limitations, the most important one being the inability to account for the non-linear  and non-stationary dynamics of the processes responsible for the analyzed time series. Applying the recurrence quantification analysis (RQA), a method of non-linear time series analysis, on solar activity indices, we study the solar cycle evolution during from 1800 to present. We discuss the main results of RQA and the relations among various descriptors of the solar cycle evolution impressed in the time series of solar activity indices analyzed in our study, by providing information for the physical interpretation of the various descriptors employed to characterize the solar cycle evolution.
8   Analysis of Ground-Based Full-Disk Synoptic Observations of the Solar Atmosphere for Solar Irradiance Reconstructions
    Ermolli, I1; Giorgi, F1; Cristaldi, A1; Stangalini, M1; Romano, P1; Zuccarello, F2
     1INAF Osservatorio Astrofisico di Catania; 2Universita` degli studi di Catania
    We review the availability and use of ground-based full-disk observations for the study of the sources of solar irradiance variations. Daily, photometric, multi-wavelength, full-disk observations of the solar atmosphere provide us with input data to irradiance models, concerning the photometric properties of solar features and the irradiance changes due to the evolution of the observed features on the solar disk. In this frame, we present the results derived from the analysis of the Rome-PSPT archive, which consists of synoptic full-disk observations characterized by a spatial resolution of 1 arcsec per pixel and a photometric accuracy 0.1%, taken since July 1997 at the radiation of the Ca II K line, Blue and Red Continua. The analysis of the Rome-PSPT observations was complemented with the study of a sample of historical observations obtained with the Equatorial Spars of the Rome and Catania Observatories, to the purpose of extending back in time the time series of solar feature measurements for the modeling of solar irradiance variations.  This study has been carried out in the framework of the EU FP7 project "SOLID  First European Comprehensive SOLar Irradiance Data exploitation" (grant n. 313188).
9   Trend Detection and in Solar Irradiance Data: New Insight into the Recent Anomalous UV Observations
    Dudok de Wit, T1; Kretzschmar, M1; Schöll, M1
    1LPC2E, CNRS/University of Orléans
    Detecting and removing instrumental trends is one of the most difficult challenges faced in the analysis of solar irradiance records. This is also crucial for properly assessing the impact of solar radiative forcing on the Earth's environment.  There is no universal recipe for determining such instrumental trends. Cross-calibration and careful monitoring of the instrument characteristics are the preferred way for getting hold of instrumental effects. However, there are many instances in which this information is not available, so that other approaches must be sought. Another classical approach then consists in comparing the observations with independent proxy data, assuming that the latter are better calibrated, or better understood in their long-term evolution.  Here, we bring this approach one step further by assuming that the solar spectral irradiance varies in phase with these proxies; however, no assumption is made on the relationship between the two, which may be nonlinear. By considering epochs at which the reference proxy has reached the same level of activity (and thus the spectral irradiance in principle too), and by inverting a parametric representation of the long-term correction of the spectral irradiance, we are able to determine by how much the latter is drifting away from the proxy. The major advantages of this novel approach are: 1) its high sensitivity, which allows to quantify weak deviations, 2) its ability to provide confidence intervals for the trend, and 3) the ability to correct for a possible phase shift between the proxy and the spectral irradiance.  We use this approach to investigate the presence of trends in various UV records, and in particular in the hotly debated observations from SORCE/SOLSTICE, which exhibit an anomalously strong cycle amplitude. Our approach indeed reveals the existence of a significant trend in SORCE/SOLSTICE. After correcting for this trend, SORCE/SOLSTICE observations can be made to agree again with data from UARS/SUSIM, and others.
10   Error Analysis of a Multi-Instrument Composite Solar Mg II Index
    Weber, M1; Chehade, W1; Machol, J2; Snow, M3; Viereck, R2
    1Universität Bremen; 2National Oceanic and Atmospheric Administration (NOAA); 3University of Colorado
    From a combination of UV satellite sounders measuring a composite solar Mg II index can be created that spans more than thirty years and roughly three Schwabe solar cycles. The Mg II index is a popoular proxy for spectral solar UV variability. It is calculated from daily spectral solar irradiances (SSI) by forming a emission core to wing ratio of the Mg II doublet at 280 nm. The Bremen composite comprises data from the SBUV/SBUV2 series, SUSIM and SOLSTICE aboard UARS, GOME/ERS-1, SOLSTICE/SORCE, OMI/AURA, SCIAMACHY/ENVISAT, and the two GOME-2' aboard Metop-A and B. Missing daily values are filled with scaled F10.7 cm radio flux data. Here we present details on the Bremen composite Mg II index. We also present a first estimate of assoicated errors in the composite timeseries and evaluate the long-term stability of the decadal Mg II index timeseries.
11   Towards the Longterm Reconstruction of the Solar EUV Irradiance for Planetary Science Applications
    Haberreiter, M1; Beer, J2; Muscheler, R3; Delouille, V4; Mampaey, B4; Verbeeck, C4; Schmutz, W5
    1PMOD/WRC; 2EAWAG, Switzerland; 3Lund University; 4ROB; 5PMOD/WRC, Davos
    Variations of the solar irradiance determine the temperature, density, and composition of any planetary atmosphere. For the understanding of the longterm changes of planetary atmospheres it is important to be able to provide realistic variations of the EUV. We present first results of the longterm reconstruction of the solar EUV since 1600. In order to derive the solar cycle variation for the present time we use the segmentation of SOHO/EIT images for the SOHO era. Synthetic spectra, emitted by the identified coronal features, are weighted by the filling factors derived from the segmentation maps, yielding the variations of the EUV for Solar Cycle 23. Moreover, for the reconstruction of the EUV back to the Maunder Minimum cosmogenic isotopes derived from ice cores are employed. Finally, neutron monitor data are used to link the cosmogenic isotope  data with the present level of solar activity. The results will be compared to existing EUV reconstructions and their implication on planetary atmospheres will be discussed.
12   Studying Solar Wind Properties using Pattern Recognition Methods
    Mursula, K1; Holappa, L1; Asikainen, T1
    1Univ. of Oulu
    Traditionally, global indices of geomagnetic activity have been used to study the long-term evolution of the Sun, solar wind and the heliospheric magnetic field. More recently, new pattern recognition methods like the principal component (PC) method, together with local indices of geomagnetic activity from an extended station network, have been applied in order to extract new, accurate information about the main structures of solar wind. It has been shown recently that the second PC component is highly correlated with the relative annual fraction of high-speed streams (HSS) in solar wind. The latitudinal distribution of the second PC mode has a high maximum at auroral latitudes, and a local minimum at subauroral latitudes related to the wider extent and stronger intensity of CME-related substorms. These results allow new, interesting possibilities to study the centennial evolution of the properties of solar wind and the HMF.
13   Solar EUV Modeling Efforts within the FP7 SOLID project
    Haberreiter, M1; Dellouille, V2; Del Zanna, G3; Dammasch, I2; Erhardt, C4; Dominique, M2; Erhardt, C4; Jones, A5; Kretzschmar, M6; Mampaey, B2; P, M4; Schmitdke, G4; Verbeeck, C2; Wieman, S7; Woods, T5; Dudok de Wit, T6; Schmutz, W1
    1PMOD/WRC; 2ROB, Brussels; 3University of Cambridge; 2ROB, Brussels; 4IPM, Freiburg; 5LASP, Boulder; 6LPC2E, Orleans; 7UCA; 
    The aim of the European FP7 Project SOLID is to provide a consistent time series of Solar Spectral Irradiance (SSI) from the X/EUV to the IR spectral range for with a focus on the space era. Here we present latest results for the EUV spectral range. Specifically, we model the SSI variations in the EUV spectral range based on the analysis of images obtained with SOHO/EIT, PROBA2/SWAP, and SDO/AIA. These images are segmented into six brightness classes using the SPoCA tool. Moreover, with the SOLar MODeling code (SOLMOD) we calculate intensity spectra representing the intensity emitted by these regions. We present the latest set of reconstructions and compare it to available data, such as SOHO/SEM, PROBA2/LYRA, ISS/SOLAR/SOLACES, and SDO/EVE. These results are important for the final SSI composite to be provided by SOLID.