Hi Bill, Just as a comment from me to you, your supreme efforts last week, in the face of my uncertainties about when my paper would be etc., made that week very productive and much more fun than it would have been with just the conference. I actually got my draft paper off to the Session Chair yesterday. A copy is attached to this reply - only for our little circle at the moment because it isn't refereed yet, and could contain all sorts of stupidities. My mail program does not show colours, so I am working with simply black and white text. At the conference I ran into Witold Kinsner of University of Manitoba. He would like to be added to our embryonic group. He is at kinsner@umanitoba.ca He is a mixture of very interesting and plus qu'un peu etrange. My comments are all in square brackets. Text that was relevant in introducing the document but not necessary in the reply has been deleted. Subject: Solar forecasting - notes from the "micro-Workshop" I also need: * and other thoughts, issues, questions that you would like expressed (post-workshop or second thoughts) [See Appendix 1] * your EIC-CC presentation from last week [This is attached] * wasn't there was a more detailed graph of pandemics than appears in the Cdn J Infectious Diseases? Can it be distributed? [I think what we should do is look at the data and decide what sort of graph we should make. It would be worth the effort. In the immediate situation I think not saying too much is good, and see what Last et al. think we need to look at. What sort of additional stuff is needed? The Phase graph says it all without adding redundancy, and in the end the analysis stands or falls in that graph. If we are discussing the pandemics stuff, I think it would be good to drag in the coauthors on that paper: Rick Mathias (epidemiologist): richard.mathias@ubc.ca Dave Surkan (GP) : surkd@shaw.ca Dave suggested an immunologist would be a good addition, since suppression of the immune system might be a factor, rather than just antigenic shifts. ] * possibly contacts of the last Canadian Solar Workshop, Kingston Nov05 - here I'm thinking that the attendees would appreciate being aware of the recent micro-Workshop and the desire to advance a "Canadian Solar Forecasting Initiative". An added advantage is that it "prepares" them for future communications, planning and discussions. Or as another way of putting it, if you don't notify your friends you might have to buy them a beer. [I agree. Actually, thinking about it, the mailing list for the last Canadian Solar Workshop is in the hands of Queen's U. I will try to get it. If not, we can use the one David and I put to gether for the first one.] I'm also putting together comments from Valdes (I haven't started quite yet). By the way, that book by Hoyt and Schatten "The role of the sun in climate change" is fantastic - a very cautious, thorough presentation of the history, issues, analysis, and pitfalls of solar activity and its effects. No doubt there will be errors and differences of opinions, but I'm impressed by the book and it s still very timely. I won't have time to complete it any time soon, but I've made it through two-thirds of the book and I've skimmed the rest. [I have asked our library to get that book (and the other one). Today I hope to get to serious grips with the Referee's comments on the Maunder Paper. I think those too will be enlightening by way of guidance, and also what the prejudices are like out there.] There are some really interesting comparisons between the book and Valdes' phase changes, and the comment or suspicion that the basic behaviour and effects of the sun can differ in different phases relates well to many historical results. Additionally, that gives another perspective on using historical results - it may be counter-productive to "average out" correlations in different phases (eg climate, pandemics etc). [I think the Valdes contribution to all this is critical. I think though that we might have to do two things.... (1) encourage Julio to do the additional calculations, or help him get the resources or other support needed, and (2) Sunspot number is a continuous (that is no missing values, not that it is necessarily statistically continuous) time series of data. It is reasonably statistically continuous though, and does not need the full power of Julio's model, and that if we can simplify it, so that it addresses solving the single-dimensional non-linear autoregression model, N_n = F(N_(n-1), N_(n-2)......... N_(n-n)) we might be able to get a model that will run in some way on simpler computer systems, and then we can all get on the act. I have a dual-processor 450 MHz system with 1GB and a 60GB disc, loaded with Scientific Linux, and if the machine needs to run for two weeks, it doesn't matter.] [I agree with broadening the audience to define the main interests. Hopefully things won't get too diversified. Looking at what is happening in the more astronomical parts of NRC, and the stated priority of environmental and global climate change issues, I think we fit into that bandwagon. It looks as though, in addition to some science, what we have particular expertise at is forecasting and statistical analysis of time series.] [My comments in the body of the minutes document are only comments, not corrections.] **************** Summary A micro-workshop on solar forecasting was held at the University of Ottawa on Thursday, 11May06. The objective was to bring together local scientists who might have an interest in solar forecasting and its application, and to introduce them to an effort to put together a Canadian initiative in this area. Three presentations were given: · Ken Tapping presented "Pandemics and Solar Activity" by K.F. Tapping, R.G. Mathias, D.L. Surkan - The risk of influenze pandemics is two to four times higher during certain phases of solar activity. This has been commented on elsewhere, . The reasons for this association are unknown, and epidemiologist ?name? cautioned that pandemics are very unpredictable, and its not clear how to make this a useful tool for practitioners (this echoes comments made before the workshop by Dr. John Last, emeritus of UOttawa) [The factor of four came from a different analysis, which having looked at it again, would rather not show it around. I think, if appropriate, we should not look at this some more, particularly in the light of it being an immune system suppression issue. We need an immunologist. I will ask Rick and Dave about candidates.] · Ken Tapping "?Modeling variations in solar irradiance" - Tapping showed a very solid correlation between F10.7 (10 cm wavelength) solar output and sunspot group numbers from 1949 to the present. The results of applying this to historical records of sunspot activity since 1400 were shown, with preliminary estimates of the error bar over time. [1600 AD, not 1400, although I wish it were.] · Julio Valdes "Time dependent neural network models for the detection of changes of state in Earth and planetary processes" - Valdes' presentation was based on a paper that he did with Graham Bonham-Carter of NRCan in Ottawa (recently retired), whereby sunspot and Greenland ice-core time series since 1700 were analysed independently with the intent of detecting phase or state changes in each of the series. They used an extremely powerful ARMA-like approach based on heterogeneous neural networks double-wrapped by genetic algorithms. Discussions touched on potential solar influences on climate, utilities, pandemics, and geomagnetics. Some of the pitfalls and complexity of modeling and forecasting were also introduced. Overall, optimism was expressed that there is a great opportunity to significantly advance the quality and greatly extend the time horizon of solar forecasts and the effects of solar activity. [I agree a million percent with this. In the initial phases I think we will need to focus on a couple of key requirements and stake our claims in those areas.] There is a clear distinction in the timescales of interest to different researchers. The solar physicists are really focusing on the annual data from ?1400? to present, as their interest is in understanding solar phenomena using "direct" datasets (sunspots, and from 1949 to present). Geologists are interested in a whole range of timescales [1600. If we imagine two components to the solar variability model: M(t).S(t), where S(t) is the modulation of irradiance by the varying distribution of photospheric structures with differing emissivities over the solar cycle, and M(t) is some process modulating the amplitude of the activity cycle as perceived at the Sun's "surface". In that case we can, from a mathematical point of view treat S(t) like a carrier and M(t) as the modulator. The modulator will reflect the trends in solar activity but not the ~11-year rhythm. That could bring us better in line with the longer-term activity records of geophysical origin. Since the sunspot cycle is already positive only, demodulation to get M(t) is simply a filtering process.] Potential participants in this initiative are actively being sought. The intent is to publish a number of initial papers and analysis to establish the credibility of efforts, and to slowly build up activity and support, including funding for graduate students and projects. Participants are very concerned that a solid scientific basis be built, and that the hype and politics of some areas of application (such as Climate Change) be avoided. [Yes. But we need to make sure we don't lose focus.] Action Items Ken Tapping - Will be completing an invited paper based on the content of his presentation at the Engineering Institute of Canada's Climate Change Conference in Ottawa, 10-12May06. Only selected papers will be published in a post-conference book. [Submitted version attached. One of the things I have to look at in detail is what the referee did to the Maunder Paper. This could be relevant.] Bill Howell - Write and distribute notes, distribute presentations or papers from the micro-workshop. Follow-up with selected individuals who could not make it to the meeting. Gather additional contact names from participants. Dave Boteler & Ken Tapping- Will be looking to establish a basis for standard calibrations of satellite [Irradiance] data, possibly through involvement of the Canadian Space Agency. Julio Valdes - didn't you say he was supposed to do some modelling, writing or something? [Yes. The sunspot number analysis he did.... we are supposed to be reformatting that to fit the interests of solar physicists rather than mathematicians, and also predicting the next two or three cycles. I sent him a Draft 0 in January. He is supposed to make this into Draft 1.] Everyone who wishes to join in this effort: · Notify Ken Tapping of your interest and potential contributions - we are looking for solid contributions to build credibility. · Plan to attend the 3rd annual "Canadian Solar Workshop", which will be held in Montréal in October or November 2006. ?Paul Charbonneau? is the contact. Previous Canadian Solar Workshops were held Nov04 in Ottawa, and Nov05 at QueensU. Hopefully a website will provide access to papers from past events plus material for the next workshop. Background to the micro-Workshop: Origins and Objectives of the Canadian Solar Forecasting Initiative Dr. Ken Tapping explained that for a couple of years discussions have been leading towards the formation of a core group to further research on solar activity forecasting and its relevance to a very broad range of terrestrial processes, including climate, agriculture, electrical and pipeline utilities, satellites, communications, and even potentially areas like major pandemics. In addition to Ken Tapping (solar physicist), core members of these discussions have been Dr. Paul Charbonneau at the Université de Montréal (solar physicist), Dr. David Boteler of Natural Resources Canada in Ottawa, Dr. Dave Thompson of Queen's University, and Dr. Julio Valdes of NRC in Ottawa. For many years solar forecasts have been provided to industry and government organizations so that they can prepare facilities and operations to minimize the risk of damage and danger to life. However, as with weather forecasts, this service has been very short term, essentially on the scale of a few hours to days, and it is felt that longer term forecasts on the scale of a complete solar cycle of ~22 years (which is two Schwabe half cycles, between which the solar magnetic poles reverse) and perhaps longer are both do-able and potentially extremely useful for large-scale, longer term planning activities such as required for utility, civil engineering for floods etc, and urban planning. An immediate forecast horizon of 30 years is the objective coming out of discussions between Ken Tapping, Paul Charbonneau, and Julio Valdes. Furthermore, it is felt that the importance of solar activity to all of these areas has often been seriously underestimated or completely ignored, and perhaps at times it has also been exaggerated. To some extent this is understandable because: · there are relatively few solar physicists, and they are fairly quiet, well-behaved people who like to have solid results and understanding before they broadcast ; · solar dynamics are typically characterized as chaotic on the scale of hours to a year, and irregular on longer timescales. This is due to multiple hidden and poorly understood processes occurring within the sun and other similar stars; · there are complex interactions between solar activity and: · other astronomical phenomena (orbit around the sun, axis precession and tilt, cosmic radiative flux and cloud formation etc etc). · geological phenomena (weather, geomagnetics, volcanic eruptions (particulates, CO2, CH4, NOx etc), CO2 solubilization or chemical reaction in the seas and release from subduction zones) · biological processes (disease outbreaks, migrations, productivity, CO2 response and possibly regulation effects) · anthropogenic effects (GHG emissions, particulates, other pollutants, local heat loads such as urban warming). · there is evidence that the sun's "behaviour or phase" changes in character significantly, sometimes with the influence of key variables changing in sign with respect to dependent variables. It would therefore be a mistake to glibly "average over" dissimilar phases or states of solar activity, without paying attention to whether this masks important effects. · solar impacts can be specific to a region on Earth, sometimes having opposite effects in different regions. By solar activity, we are primarily referring to the: · structure and "behaviour" of measurable systems within and around the sun (sunspots, faculae, magnetic poles etc); · irradiance of the Earth by the sun, including its spectral composition and variability; · other variables such as solar winds, although these variables are not being used in current projects. Influenza and other Pandemics and Outbreaks Tapping presented the results showing a significant correlation between major influenza pandemics (worldwide outbreaks) and sunspot activity since 1700 (see references). Pandemic risks increase 2 to fourfold (very roughly) during a specific phase of the sun. He was careful to state that a solid mechanism to explain the association has NOT been established. In earlier emails with Dr. John Last of the University of Ottawa and ?Dr, Mathias at UBC?, increased mutation rates, large climate changes (especially much cooler, wetter weather) and possible effects on the immune system were mentioned as possible lines of investigation. It is clear that solar activity cannot be the only determinant of outbreaks, and may not even be a major determinant. However, given that we are speaking about many human lives, it is important to draw attention to the association. Dr. Raina Fyson commented that the factors that influence outbreaks are quite complex ?and include human behaviours, etc?. There is no way of predicting them, and spurious associations can arise in datasets. She agreed with comments made before the workshop by Last, to the effect that even knowing the correlation itself will not be a critical tool for epidemiologists, as they must monitor real outbreaks, their progress, and the effectiveness of responses and controls such as treatments, quarantines of infected people, and vaccination programs. Perhaps the value of the association between solar activity and pandemics may be that it would help to justify increased preventive resources and programs during high-risk periods. Perhaps it would also help if correlations between solar activity and climate were more solid and widely recognized. This could be a key outcome of solar forecasting and modeling. Fyson's recommendations · verify trends in cholera and malaria, as they are climate-sensitive diseases · check the activities of the American Society for Microbiology (Note: In previous email exchanges . Melanoma of the skin hasn't been mentioned, but there wasn't much time for discussion/reaction.) %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% APPENDIX 1 ========== How this evolves will depend upon how and who responds. We will need to spread the focus enough to keep the newcomers in there, while not letting this broadening stop things happening. At the moment we have a small list of immediate objectives in paper production. These are necessary flags for staking our claims. If we are thinking of evolving in the direction of getting money, we will need to define clear goals and objectives. I don't see this as incredibly difficult. Immediate related jobs/objectives: 1. Get that Maunder Paper revised and resubmitted. This is not in the Forecasting court but it will be a reference in subsequent work and a diversion until it is out of the way (Boteler, Charbonneau, Tapping). 2. Paper using F10.7 to examine the internal consistency of the integrated irradiance database. To produce/propose a modified database (Boteler, Tapping +). 3. Get that paper presenting Julio's work and the sunspot prediction out there (Valdes, Boteler, Tapping, +). 4. Re-examine the influenza pandemics issue. Is there more that we can do. If so.... do it (??). I think a key issue is communication. We need to develop a comprehensive e-mail list and use it to make sure ideas and things circulate. Therefore: 1. Produce a generic writeup on the initiative and encouraging input and participation, emphasizing that this is both science and something that could be very relevant to Canada. 2. Make sure all this gets out there to at least those who attend CSW's. 3. Make a presentation on this at the next CSW, perhaps with a further information sheet. 4. We need to decide the next course of action, which could be growing an appropriate funding proposal. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Regards, Ken