 Has provided enough food for a good discussion before coffee? Please, I open up the floor for any questions and comments. If you could maybe briefly introduce yourself? Yes, Martin Morton, University of Liverpool. I work with the Met Office quite a bit in the UK, and I'd just like your response to this question. How confident are you that that particular climate model can accurately predict in the future what the Indian or the Sun will do? Not very. Thank you very much. Because the experience that I have working with them is you need an ensemble of 15 model outputs before you get an even remote action of the Indian or the Sun. The uncertainties associated with that really raise enormous questions about how you can be confident in predicting what might happen. It worries me enormously, I have to say, to see the generalisation coming out, because it implies confidence, which I'm not convinced is warranted. So I'd be interested to see what you could say. Well, I'll quickly comment on that. I thought I had stressed this enough, but I will stress it again, we realise we're only running this with one climate model, but we simply do not have the time and capacity to do anything else at this stage. Across all the models, rainfall is the big sticking point. There is no model that has any good, as I understand, physical understanding of the rainfall patterns. We're certainly not giving this information out with an idea of confidence and accuracy. What we do want to say is that this may represent one possible plausible future climate. I'm not going to say that this is the future climate, I don't think anyone would dare to stand up and say such a thing. But what we do want to do is to try and see what information we can use to come up with some form of plausible future yield potential in this case. But I completely agree with the uncertainty in this kind of modelling. We simply don't have the options to do anything more comprehensive at this stage, because the time and effort required to do just one model for this area has been actually way beyond what we initially anticipated. And it goes back to a lot of the other institutes in the region who have actually done this kind of modelling, but for whatever reason we were unable to share that information with them. I would like to think that this situation would improve in the future because we're not going to be able to do what you just described in terms of ensemble modelling across this type of scale unless we do share this kind of information more freely. Okay, we have one here and then there. I'm a member of the media. I just want to highlight that we're doing all farmers to understand these research needs and fixing their research priorities for the future. I think you... Get a microphone, please. There's a lot of... You will find out about the methods that you're using, whatever you're using in the subsequent presentation, so I'd encourage you to stay away from this. There is a lot of farmer input embedded in the methods that we've used, but it's not directly based on farmer interviews per se. We ask them what the priorities are, but you'll find out that it is using information sources that are based on data collected from farmers. You've seen the example of Alayu Diagne before, and I think we have also, for Asia, a substantial database of household-level information. And so I think it feeds indirectly into the process, but you need to still have a structured methodology that combines that household-level or farmer information or feedback with the more generalized biophysical information of the kind that Andy has shown. So it is, I think, indirectly part of the process. What we have not done is like... Well, sometimes it's being done, you know, conduct like an e-consultation and ask people, what do you think are the ten most important things? Because that is very difficult to make any use of, because you ask ten people the same question, you get ten completely different answers, because each of them has a certain priority, which is not balanced against the other ones. And the other problem that we have with this type of approach is that it's a very biased sample of people who respond to that. So we're trying to combine different sources and what we believe at this stage is a reasonably robust and objective approach, but of course knowing very well that many other stages of feedback will have to come into this at a certain point. Dr. Dutta. Just to supplement Martin's question actually, not a question, but as a comment, just a few days back we were having a meeting about express kind of discussions and the Minister for Power was present in the discussions. And in the Metrological Department was making a presentation and giving even more elaborated than what you said and it was just very nice. I enjoyed your presentation. And then the Minister himself asked a question to the person who was presenting, is monsoon stopped in India? Still in India, monsoon rain, what rain is still going on? When it started, it was said that it was not a monsoon, it was something pre-monsoon. And even still continuing monsoon water and Andhra Pradesh again flooded. So that is the things what Martin just raised the question. I think it is a bit complicated, but at the same time we need to understand these things and to make it as much as possible as Dr. Norman mentioned, that combination of different parameters that we put together. And even we do have in India now, every week there is data coming, details that are coming, still you could not predict it correctly. Thank you. I completely agree and it's rainfall which is the sticking point in all of these models. Downscaling is particularly troublesome. At the moment we are using the most advanced models that are out there and available to us. I am working within the limitations that we have with those. That is what I would like to add today. It is probably well able to say that, remind everyone that this particular piece, the projection of future climate and its implication in yield potential is really only one piece of the whole exercise. We wanted to have it in there so that we have at least general protection where our major areas of change. But I don't think we want to go down and say this pixel is going to behave differently than this pixel. Because we also know very well that the crop simulation models that we have also have inherent uncertainties particularly when it comes to simulating the interaction of CO2 in temperature and crop development and performance. There is another layer of uncertainty which we are well aware of. So I think we don't want to put all our eggs into the basket of these kind of simulations. But we still have some more time for more. There is a young lady there. Wait, wait, wait, wait. So my question is, in the plenary on day one we talked a little bit about that we are producing enough for 9 billion and I'm wondering where the mapping is happening to the distribution issue or the 1.4 billion people where this yield is not reaching and are you starting to work with anybody that integrates the distribution issues and the strategic planning process or which of your working groups is the distribution power? I think the first question was do we have enough food for 9 billion people? That was a comment made by one presenter and I have a serious question on that. The presenter says we have enough food for 9 billion people. I believe that too actually. But David do you want to answer the distribution question? Yeah, again I'd like to make a plug to the second component of this symposium because you'll learn more about how that's handled there. But basically what we're doing is in terms of distribution one of the major avenues in terms of food security is food affordability and that's what we're modeling on the consumer side. So what the actual effect would be on rice prices for core consumers and what that means in terms of additional consumption that could be enabled. So we captured that distribution issue through the price issue and really the matter is you can have any supply and demand equilibrium that you want. It's all reflected in what price you end up with there. So it's not really a question do you have enough food for 1 billion people it's a matter of at what price do you have food that could be available because of what we substitute. So we reflect that in the price issue. If there is I'll entertain one final question specifically on the spatial data not on the other aspects of the process and use presentation. If there is one question on that I will take that. Say it's cold. First of all I congratulate you on trying to put some numbers on some variables. First of all, this is not the objective this is an activity that can help us assign research priorities and resources. That's all we're trying to do. You take a look at how we've been doing this for 20 years within the CG system we just say priorities is the problem. We make them up with something else like this gentleman was talking about Africa. We've done a monstrous work in Africa which presented very well we assume priorities is the problem. It's not priorities. We're at least trying to put a scientific method of assigning the world's resources on priorities. And you know we may not be very accurate but no one else is doing it. Thanks. We'll break for I don't think there is a response to that. The coffee break will be back at 11 o'clock sharp please.