paper coding culture - max van der pluijm
DESCRIPTION
Numerical weather predictions nowadays rely on the representations and numerical modelling of the atmosphere by a computer model. Since mathematical equations that are being used in order to govern the atmospheric outlines are not precise, they also tend not to represent the real atmosphere entirely and accurately. Seen from a critical point of view this is a big shortcoming, since numerical weather predictions are positioned at the basis of modern meteorology. Nevertheless we rely much upon these computer simulations, which are being used in order to prevent certain catastrophes or simply stating that it is going to be nice weather. This paper investigates some specific problems within the field of numerical weather predictions.Computer simulations can be seen as powerful tools if understood and used properly. The introduction starts off with a definition and historical perspective surrounding numerical weather predictions. Then goes through a centralised case about various weather forecasts that have a great impact upon economical, political and cultural structures. This will also be the main problem addressed within this paper, in what way the outcomes of computer simulations that are being used to produce numerical weather predictions affect our society. Throughout the paper, there is an on- going discussion on issues concerning the use of computer simulation within everyday life. Connected to this the issues will be put onto a globally put level.TRANSCRIPT
Numerical Weather Predictions
Do they affect our society?
Coding Culture Paper
Max van der Pluijm – 3727726
Trimester 4, 2011/2012
July 2, 2012
Teacher: Johannes Paßmann
ABSTRACT Numerical weather predictions nowadays rely on the representations and numerical
modelling of the atmosphere by a computer model. Since mathematical equations that
are being used in order to govern the atmospheric outlines are not precise, they also
tend not to represent the real atmosphere entirely and accurately. Seen from a critical
point of view this is a big shortcoming, since numerical weather predictions are
positioned at the basis of modern meteorology. Nevertheless we rely much upon these
computer simulations, which are being used in order to prevent certain catastrophes or
simply stating that it is going to be nice weather. This paper investigates some specific
problems within the field of numerical weather predictions.
Computer simulations can be seen as powerful tools if understood and used
properly. The introduction starts off with a definition and historical perspective
surrounding numerical weather predictions. Then goes through a centralised case about
various weather forecasts that have a great impact upon economical, political and
cultural structures. This will also be the main problem addressed within this paper, in
what way the outcomes of computer simulations that are being used to produce
numerical weather predictions affect our society. Throughout the paper, there is an on-‐
going discussion on issues concerning the use of computer simulation within everyday
life. Connected to this the issues will be put onto a globally put level.
2
INDEX
Abstract ................................................................................................................................. 1
1.0 Introduction ................................................................................................................. 3
2.0 The case: the 21st of june, a warning was given ................................................ 5
2.1 Issues regarding numerical weather predictions – part one ............................................. 6
2.2 The eruption of the Eyjafjallajökull Icelandic volcano .......................................................... 8
2.3 Issues regarding numerical weather predictions – part two ............................................. 9
3.0 Conclusion .................................................................................................................. 11
Bibliography ..................................................................................................................... 13
3
1.0 INTRODUCTION Weather forecasting is an important controlling factor in many activities within our
society, such as air, road and water traffic, the planning of gas and electricity production,
drainage purposes, air quality, tourism and agriculture. Numerical methods play a major
role in the realization of the forecast. A century ago weather forecasting was a random
process, it was very imprecise and most of all, very unreliable. The observations that
were made back in the early days were very irregular and the methods used or
practiced held a kind of crude notice. Forecasters made use of rough techniques,
knowledge of the local weather conditions and foremost it was an art of guessing.
Nowadays it is unthinkable that modern meteorology would be successful without the
help of powerful computing facilities. Since the early days people made use of the
mathematical models of the atmosphere and oceans present at hand based on the
current weather conditions. In other words we are now talking about numerical
weather predictions.
The idea of weather predictions based on numerical conditions is not something
that is new. The central idea is that you could predict the weather by solving physical
mathematical equations. For in fact the basic equations were formulated in 1904 by
Vilhelm Bjerknes, a Norwegian physicist and meteorologist who did much to the
founding of modern practice of weather forecasting (Gramelsberger 2006, 78; Lynch
2008, 3432; Tisler 2006, 7). Although Bjerknes was ahead for his time by developing a
qualitative, graphical method, it was not until the British mathematician Lewis Fry
Richardson that this was pushed on to the next level (Lynch 2008, 3433). In his
published book Weather Prediction by Numerical Process, Richardson addresses in the
preface a pure dream, which, in light of this written paper, stands out quite nicely
(Richardson 1922, xii). The dream Richardson spoke of was focused on the possibility to
advance computations faster than the weather advances, which became possible (Lynch
2008, 3433). As we know now, weather forecasts are produced on regular basis while
running algorithms on computers, sometimes even supercomputers, which are very
similar to the ideas thought of by Bjerknes and Richardson.
This was due to the important work of John von Neumann, who was one of the
leading mathematicians of the previous century. Von Neumann experimented a lot with
complex numerical equations and for him it was more than clear that ‘very fast
automatic computing machinery was required’ (Ibid., 3435). Though the complete
history of the computer shall not be discussed here, it is still clear that with the rise of
the computer the process of numerical weather predictions accelerated effectively. This
4
is because researchers started the implementation of models within computer
simulations and thus in recent years we can now simulate all sorts of meteorological
conditions. From the predicaments of typhoons, snowy weather forecasts up to the
global warming and melting of the icecaps.
The aim of this written paper is to look into some specific problems that present
themselves with the use of computer simulations benefitting numerical weather
predictions. Instead of only looking onto the positive side of these uses, we should
consider the possible downside(s) that are connected to it. Recent events have shown
that with the use of computer simulations we tend to overreact at some levels and
decisions have been made based on computer simulations for some time now. In this
paper we address a central question, which will be discussed throughout the theoretical
standpoint and foremost the case that will be used. The central question of this paper is;
in what way do the outcomes of computer simulations that are being used to produce
numerical weather predictions affect our society?
I will make clear that with the use of computer simulations regarding numerical
weather predictions, we rely too much on the information given by these computer
simulations. Although history has shown us that they could be of great use, the opposite
also has occurred.
This research consists of two main sections divided in several sub layers. The
first section is mainly covered by the case, which will be discussed extensively. The
centralised case, which is about the extreme weather forecasts within the Netherlands
of the past several years, will serve as a good example of the possible factors that could
go wrong with the interpretation of computer simulations. Furthermore there will be an
extra element within the case that exemplifies the thoughts on a more global
perspective. This will be met by taking the eruption of the Icelandic volcano
Eyjafjallajökull as a fine example. Through the use of theoretical backgrounds the
arguments will be made clear. Where the second section will consist of the conclusion,
supplemented by the discussion, regarding the centralised question.
It is important to note that within the elaboration of the case, issues regarding
political systems, economic structures and capitalism will be addressed. Since these are
all connected to major events and decisions regarding extreme weather forecasts, they
are implemented within this paper. Knowing what consequences numerical weather
predictions has on the political, economical and cultural structures within a society
would help one person. By stating the underlying aspects of the use and trust
5
researchers have in computer simulations we can address even the higher educational
segments of the discourse we are positioned in.
Thus we can start off with the centralised case, which holds various examples of
numerical weather predictions that led to some big decisions with vigorous
consequences towards the different structures within societies.
2.0 THE CASE: THE 21ST OF JUNE, A WARNING WAS GIVEN On the 21st of June 2012 the Royal Dutch Meteorological Institute foresaw a big
irregularity within the climate. This irregularity led to the warning of code orange. This
means that the weather is becoming dangerous and unusual meteorological conditions
are expected. In the worst-‐case scenario’s you could speak of damages to houses,
infrastructure and many more whereas accidents are not excluded ("KNMI Kleurcodes…”
2012). The warning was submitted for five provinces of the Netherlands and the
expectation was that within a short period a massive amount of rain was coming down
from the sky, combined with heavy lighting storms and possibly hail (Novum/ANP
2012).
This is one of the tasks that the Royal Dutch Meteorological Institute has to do,
warn the population when there are signs of possible heavy weather. Making these calls
asks for a lot of grounded data and conclusions. When the call was made, immediately
everybody was in an uproar. This led to a couple of decisions made by various
corporations. The first example that could be given here is that the Dutch Automobile
Association gave notice to the call that had been given. They expected that the heavy
weather would cause a lot of problems within the traffic, and especially during peak
hours. The Traffic Intelligence Service said that travellers should take into account that
the heavy weather would cause extensive delays and of course the possibility of damage
and accidents ("Avondspits blijft..." 2012). The Dutch Railway Corporation on their part
had to make a lot of preparations in light of the upcoming weather, mostly because of
the fact that a few days before the warning submitted by the Royal Dutch Meteorological
Institute; lightning struck the net of the Dutch Railway Corporation. This led to excessive
long waiting periods, because the Dutch Railway Corporation was not prepared for this
kind of problem ("Blikseminslag ontregelt..." 2012). So they had to prepare a lot of
possible scenarios. One of those scenarios meant the use of ninety busses, which would
shuttle the stranded passengers between the various railway stations if needed. When
the heavy weather would shift its position the busses would follow its route.
6
In the end the weather was not that bad, certainly not an orange tinted warning.
Thus all the preparations and the uproar were for nothing. Nonetheless, all the
discussed organisations had made preparations and thus spent a lot of money, or even
perhaps lost a lot of money in the case of the Dutch Railway Corporation. Mainly
because they advised people to avoid the use of their services and even advised people
to stay indoors. So in retrospect, the weather forecast, the weather simulation that had
been run that day by the Royal Dutch Meteorological Institute had caused a pretty
intense uproar within the Netherlands.
2.1 ISSUES REGARDING NUMERICAL WEATHER PREDICTIONS – PART ONE This particular aspect as stated before, the fact that a single computer simulation can
alter an entire infrastructure by its core fundaments, is something that I would like to
discuss here. Despite the fact that it still had rained and lightning was spotted that
evening, it was not as bad as what the computer simulation had shown when looking
back into the history of the simulations run by Buienradar, associated to Meteox, the
European rainfall radar ("Buienradar.nl…" 2012).
A lot of people and organisations nowadays rely on the outcomes of computer
simulations. For in fact, what we might have learned from the particular case discussed
above, computer simulations are rendered by a number of variables and these variables
change. As Orrin H. Pilkey, professor Emeritus of Earth and Ocean Sciences, and Linda
Pilkey-‐Jarvis, geologist in the State of Washington’s Department of Ecology, argue in
their written book Useless Arithmetic that the outcome of natural processes in general
cannot be accurately predicted by mathematical models (Pilkey and Pilkey-‐Jarvis 2007).
Although they make this clear by addressing mathematical models, which are
descriptions of processes or predictions about the end results of certain processes
expressed as equations, they still relate to the central placed topic (Ibid., 24). Especially
quantitative mathematical models are in order here since they are predictive models
that seek the answer to the questions where, when, and how much (Ibid., 24). Within
quantitative mathematical models it is crucial that everything is done precisely in order
to understand the process and the variables connected to it (Ibid., 26).
Although at some levels the Royal Dutch Meteorological Institute was right
about the weather that day 21st June 2012, the outcome of the process was not correct.
The mathematical model used for their computer simulation is one with quantitative
roots. Except what stands out most in this case is that relation between the computer
simulations run by the Royal Dutch Meteorological Institute and the interpretation
between the institute and the rest of the country, which is one of a unique nature.
7
As Gabriele Gramelsberger, a philosopher who is interested in the influence of
computation on science and society, points out that with the rise of computational
science there is a twofold epistemological function to recognise (Gramelsberger 2006,
84). On the one hand you have the computer-‐based simulations, which describe the
matter at hand, and on the other hand they determine their object of research. For
example, the numerical weather predictions that were produced by the Royal Dutch
Meteorological Institute. Gramelsberger reflects upon unexpected insights, where for
instance she talks about the world as a computer simulation, which is enclosed in the
memory of her USB stick (Ibid., 87). The interesting aspect of this example is that a
simulation of the world, in terms of global climate changes, can in fact be stored on a
USB stick. But without the help of supercomputers and fast numerical equations, and
don’t forget the necessary batch files, she had nothing to put on her USB stick. The
reason why I point this out is because of the fact that the scientific world has become
dependent on the fast growing industry that produces the needed supercomputers. It is
perhaps an inevitable path that will be walked upon, except, spoken on a more
philosophical level, who says the supercomputers are always right. I do see the benefits
of the use of supercomputers, I also use computers every day, but the question could be
asked whether or not computers define our perception of the world. Gramelsberger also
addresses this particular issue when she talks about her third and final surprise. When a
super computer like the Japanese Earth Simulator unveils a perforated world—missing
pieces as big as entire countries and time gaps of hours and days—one’s perception of
the world could be altered (Ibid., 78). It also can be related to the choices that were
made when starting the computer simulation when people decide which variables are
applied (Quiggin 2008, 204).
Nonetheless, this issue can be solved up to a certain degree when applying narrative
concepts, explaining computer simulations via the use of a story’s. The most simplified
example we can think of is the weather forecast during the daily news broadcast. People
explain the computer simulations through a story and most of the time with the help of
semiotic concepts. This is no different from the scientific stories that are held. Scientific
stories are a special type of narration since they are successful and believable. The
scientist is perceived as being a romantic and his narration is the simplified version.
Simulation models (the mixture of computer simulations and the models as
components), as Gramelsberger points it out, are based on such narrations but they add
realistic details to the abstract core (Gramelsberger 2006, 80). This resembles the
characteristics of story telling a lot that we all know contains “drama, actors, interesting
8
locations and an enormous catastrophic potential” (Ibid., 80). Except, what happens is
that these stories tell us a lie. The moment when a computer simulation has been run all
sorts of tools are used in order to diagnose its validity, accuracy and try to uncover
mistakes. In the end these simulations give a false pretence, a false story about the
world.
This is no different from the case we are dealing with. Since the story of our case
was leaning towards a dreadful afternoon and evening, full with rain, lightning and
possible hail, in the end it was a ‘lie’. The reason eventually that it was a lie, had to do
with the many variables that are related to the ever changing weather. The words ‘ever
changing’ already stipulates it a bit.
As Pilkey and Pilkey-‐Jarvis made clear, a computer simulation of a natural
process over time and space may involve hundreds of lines of equations (Pilkey and
Pilkey-‐Jarvis 2007, 26). They pose a very interesting question whether or not the
software or computer code actually model what the authors say it models and in the end
if it comes true (Ibid., 26). In our case it did not. The warning, or the story, that had been
released right after the simulations had been run was not correct. Though it has to be
said that a lot of variables are needed to be taken into account. Pilkey and Pilkey-‐Jarvis
explained it in a reasonable sense. Though perhaps the most important reason that
quantitative predictive mathematical models of natural processes on earth are doomed
to fail, had to do with ordering complexity (Ibid., 32). When one variable changes within
the complexity that is numerical weather prediction, another one may change
exponentially. Several variables may even change simultaneously, which makes it even
harder. So one could ask to what length a system can cope with this, these never-‐ending
changes that take place.
2.2 THE ERUPTION OF THE EYJAFJALLAJO ̈KULL ICELANDIC VOLCANO Although Numerical weather predictions tend to be hard to transfer towards the public
opinion, they often succeed in a manner one cannot measure. When linking back to the
centralised case, we saw that a vast majority of the institutes or organisations
responsible for the better part of the infrastructure met their end of the deal. Except,
this meant, especially for the Dutch Railway Organisation, a lot of planning and the loss
of a certain income. They advised people to stay at home when it was not needed to go
anywhere. The following logical step in the chain was that people really stayed at home
and thus the Dutch Railway Organisation missed some income when talking about train
tickets. Besides that they had to reserve ninety busses that would shuttle between
different stations. Although this is a hypothesis, it still seems valid and thus proclaiming
9
that numerical weather predictions have a significant impact on the politics and
economics of a company and in the end the cultural field. This is not something that is so
different from the past. With a slight form of hindsight we can state that this is a
returning aspect.
The example given about the 21st of June 2012 is still our centralised case,
however, to point out an example of the impact that numerical weather predictions
could have, there is another fine example that took place on a much larger scale. The
eruption of the Eyjafjallajökull Icelandic volcano during the period 14 April to 21 May
2010 caused an unmatched disruption to the European airspace. This was caused due to
the massive amounts of volcanic ash particles that were spread across Europe. It is said
that when volcanic ash particles appear in the engines of airplanes they become
damaged, damage flight control systems and cause jet engines to fail (Neal and Guffanti
2010, 1). Of course this is a very dangerous aspect and has to be prevented at all times.
Although the ash cloud was not observable by the naked eye, nevertheless it was
there and had a massive impact on the European aviation. Aircrafts were grounded in
most parts of Europe for more than five days. The air traffic bans for the different parts
of Europe were mostly based on the forecasted ash cloud dispersion run by computer
simulations. Thus, the forecasting, or numerical weather predictions, of the dispersion
of the ash clouds had become a major public issue, which affected a much broader
perspective than only the aviation industry (Emeis et al. 2010, 2690).
2.3 ISSUES REGARDING NUMERICAL WEATHER PREDICTIONS – PART TWO To start off with the local effects, which were quite dramatic at a certain point. People
living in the rural areas ‘down wind’ of the volcano had to wear goggles and facemasks
in order to prevent ash from reaching in to the longs and of course eyes. This was
because the ash was really thick. Further more there was a serious risk that the local
farmers and businesses were suffering enormous from the eruption.
When the volcano erupted the scientific world started running computer
simulations in order to predict the course of the massive ash cloud that was formed by
the volcano. These computer simulations, which were run by various institutes,
immediately stated that the aviation of various countries should be suspended. Although
the predictions of another eruption around that time were rather unpredictable, the ash
cloud itself was already airborne. Ground-‐base observations and numerical predictions
of the dispersion of the cloud were nonetheless possible, and there was a fundamental
need for reliable predictions. These predictions were of course not only for the aviation,
but also for the quality of the air that is. Of course, when there is a large amount of ash
10
particles within the sky, it is bound to fall down eventually, so this aspect was rather
important as well.
Results emitted by several institutes and unions showed us that for air traffic
security, the combination of a good dispersion model with special profiling
measurements means they could predict the dispersion of the volcanic ash cloud
(Bartnicki et al. 2010, 7; Emeis et al. 2010, 2699). This lead to the closure of several big
airports, thus leading to a lot of missed incomes. It has to be said that it was a good thing
that these predictions were made, for the better part of the safety of course, but it also
meant a tremendous amount of impact on economical, political and cultural levels. The
economic aspect was already stipulated in the previous paragraphs, but it is also
important to note that the ash cloud on a global perspective affected the world. The
International Air Transport Association made an estimate regarding the losses of a
roughly 160 million euros a day, which meant 1,3 billion in total during an eight day
period ("Iceland volcano..." 2012). This was not the only economical impact that the ash
cloud had. When airports are closed, logically transport of goods are also suspended, air
transport that is. Thus, resulting that companies had to import or export their goods via
alternative channels. A good alternative, that is, when there is no economical pressure to
it. So the different aspects of an economical backstab were present.
Political speaking, tensions between different parts of the world, which is an on
going phenomenon for centuries nowadays, meant that meetings had to be cancelled or
rearranged. Although people can have the decency to wait a couple of days when the ash
clouds would lay down, nonetheless a lot of summits and visits of ministers and
royalty’s had to be cancelled. And of course, people have little patience when they are
under pressure, so one could fill the blanks in for themselves when someone would
state that the tensions between different countries would only become higher.
And last but not least we can speak of a cultural impact. The ash clouds
disrupted television broadcasts because planned guests could not appear on television
due the cancelling of flights ("Iron Man 2 premiere…” 2010). The music industry had to
deal with a massive blow because a lot of artist could not realise their appearances on
the stages due to the cancelling of flights ("Whitney Houston, John Cleese…” 2010).
Besides the music and the television industry, the ash clouds also affected the sports
industry, again because of the cancelling of all the flights.
Once more it has to be stipulated that without the numerical weather
predictions, the risk of casualties would be way to high. The fact that a computer
simulation can render a very precise schematic of the route the ash clouds would take is
amazing enough. But, the predictions were also a forecast, which meant that they were
11
run before the actual trajectory of the ash clouds were set in motion. This is something
that reoccurs almost every time. To put it bluntly, data is being gathered, put into a
numerical model and run by a computer simulation. This variables with which we have
to deal with are enormous and very unpredictable up to a certain degree, so when we
link this particular example to the centralised case we can see a very clear resemblance,
the fact that numerical weather predictions are made of equations and tell us a story
which can alter any possible minute. Nothing new to that, but they could have a massive
impact on economical, political and cultural levels. Thus, in the next chapter we will
discuss the centralised case, in relation to the main question that has been asked in the
beginning of this paper.
3.0 CONCLUSION The numerical weather predictions made by the Royal Dutch Meteorological Institute,
the eruption of the Eyjafjallajökull Icelandic volcano and all the consequences that were
connected to these events showed us a perfect example of how computer simulations
have an impact upon society.
It became clear that with the important responsibilities that the various
institutes have, a certain degree of shift within a country, or even globally seen, becomes
affected. By taking the centralised case we saw that with the computer simulations that
were run, in order to apprehend numerical weather predictions, a lot of consequences
are connected to the outcomes. This is of course not a bad thing, but when the
predictions, in the case of 21st June of 2012, are not proceeding as foreseen, a lot of
levels within society are being tested. The fact that the Dutch Railway Corporation had
made a lot of last-‐minute plans in order to prevent a certain infrastructural catastrophe
was paired with a lot of losses in terms of money. Which in the end was not necessary at
all.
Although it is hard to predict the weather, or to put it in a better sense of words,
the climate, computer simulations play a very big role in this particular event. We
‘translate’ these computer simulations into a story, which makes it a lot easier for the
better part of the people to understand. These stories, however, can be deceitful when
looking at a strict level. It is not so much the fact that they often do give us a reliable feed
of information; it is more the part that they have a rigorous impact upon various fields
we all move in. As explained in throughout this paper, the computer simulations run
daily affect economical, political and social structures. By discussing it on a simpler, a
more concrete level, it is now possible to lift it to a broader context of speaking.
12
Though numerical weather predictions are a perfect subject of interest for me as
an author to stipulate my point, it can also be transferred upon other scientific fields
such as economics, political, medicine and a many more. This is because of the fact that
within computer simulations we have to take into account the various variables. As we
have seen with the case we started with and the discussed consequences of the volcanic
eruption, variables are always playing with the outcomes of a computer simulation. It is
very hard to program it into the software, but the underlying fact is, we still translate the
outcomes of the computer simulations and follow or act upon them. This phenomenon
then takes us to a new level of awareness.
All the examples that were put forth within this paper help us in understanding a
bit of the consequences, which computer simulations can have. In the case of the
eruption of the Icelandic volcano, which meant a lot of ash clouds, the computer
simulations were put to the test in order to predict the trajectories. But again, variables
such as winds and pressure areas can alter the trajectory that easy. These variables then
cannot be foreseen at the moment of simulating the Numerical weather predictions. To
put it quite bluntly, in light of the described events within this paper, we can ask
ourselves in which way computer simulations are valid enough to rely upon. The
economical, political and social consequences can be quite dramatic at times. Then again,
without the computer simulations that can be run and used, we cannot foresee the
trajectory of the ash clouds for instance. Connected to this particular example, airplanes
for instance, would be damaged and in some cases even crash, which would lead to a lot
of casualties.
Numerical weather predictions in the end are a good thing to make us aware of
the possible events that might happen. Although the events sometimes neglect to appear
in real life, they do have an impact. An impact, which can be measured upon economical,
political and social structures. Someone might say that we do rely most of the time on
the outcomes of these computer simulations and to put in a more philosophical light one
might ask; in what way do these simulations control us? Of course this is a very broad
and difficult field to do research on, but in a way, computer simulations do control our
lives, decisions and major predictions subdued to big consequences. In light of this
written paper and the more philosophical question put forward in the end we can
conclude that simulations, whether we like it or not, are important for the wellbeing of
our civilisation. Perhaps these new insights will show that the problem at hand with
computer simulations is closer than one might think. Either way, uncertainty about the
future does not justify inaction in the present.
13
BIBLIOGRAPHY "Avondspits blijft slecht weer bespaard." De VerkeersInformatieDienst.
http://vid.nl/Nieuws/article/VID.2012.172.10 (accessed June 24, 2012).
Bartnicki, Jerzy, Øystein Hov, Alvaro Valdebenito, and Michael Gauss. "Eyjafjallajökull
eruption. Prediction of atmospheric dispersion of volcanic ash from the
Eyjafjallajökull eruption -‐ An example of operational post-‐processing of
numerical weather prediction data." Meta 2 (2010): 4-‐10.
"Blikseminslag ontregelt treinverkeer Utrecht." De VerkeersInformatieDienst.
http://vid.nl/Nieuws/article/VID.2012.170.02 (accessed June 24, 2012).
"Buienradar.nl -‐ Europa -‐ Historie opvragen." Buienradar.nl -‐ Europa.
http://europa.buienradar.nl/hist.aspx (accessed June 25, 2012).
Emeis, S., R. Forkel, L. Ries, F. Meinhardt, W. Birmili, C. Münkel, F. Obleitner, P. Suppan,
W. Junkermann, K. Schäfer, H. Flentje, S. Gilge, W. Fricke, M. Wiegner, V.
Freudenthaler, and S. Groß. "Measurement and simulation of the 16/17 April
2010 Eyjafjallajökull volcanic ash layer dispersion in the northern Alpine
region." Atmospheric Chemistry and Physics 11, no. 6 (2011): 2399-‐2421.
http://www.atmos-‐chem-‐phys.net/11/issue6.html (accessed June 29, 2012).
Gramelsberger, Gabriele. "Storytelling with Code. Archeology of Climate Modelling."
TeamEthno-‐Online 2 (2006): 77-‐84. www.teamethno-‐
online.org.uk/Issue2/Gramelsberger.pdf (accessed June 26, 2012).
"Iceland volcano: Airlines 'to lose $200m a day'." BBC News -‐ Iceland volcano: Airlines
'to lose $200m a day'. news.bbc.co.uk/2/hi/uk_news/8624663.stm (accessed
June 28, 2012).
"Iron Man 2 premiere switches from UK to US." BBC News -‐ Iron Man 2 premiere
switches from UK to US.
http://news.bbc.co.uk/2/hi/entertainment/8630813.stm (accessed June 28,
2012).
"KNMI Kleurcodes waarschuwingen, Weeralarm en Meteoalarm." KNMI -‐ Koninklijk
Nederlands Meteorologisch Instituut.
http://www.knmi.nl/cms/content/32692/kleurcodes_waarschuwingen_en_we
eralarm (accessed June 27, 2012).
Lynch, Peter. "The origins of computer weather prediction and climate modeling."
Journal of Computational Physics 227, no. 7 (2008): 3431-‐3444.
http://www.sciencedirect.com/science/article/pii/S0021999107000952
(accessed June 25, 2012).
14
Neal, Christina A., and Marianne C. Guffanti. "Airborne Volcanic Ash -‐ A Global Threat to
Aviation." U.S. geological survey -‐ reducing the risk from volcano hazards 3116
(2010): 1-‐6.
Novum/ANP. "KNMI waarschuwt voor extreem weer." KNMI waarschuwt voor extreem
weer | nu.nl/binnenland | Het laatste nieuws het eerst op nu.nl.
www.nu.nl/binnenland/2840858/knmi-‐waarschuwt-‐extreem-‐weer.html
(accessed June 23, 2012).
Pilkey, Orrin H., and Linda Pilkey-‐Jarvis. Useless arithmetic: why environmental scientists
can't predict the future. New York: Columbia University Press, 2007.
Quiggin, John. "Uncertainty and climate change policy." Economic Analysis & Policy 38, no.
2 (2008): 203-‐210.
Richardson, Lewis Fry. Weather prediction by numerical process. 1922. Reprint, New
York: Dover Publications, 1965.
The Huffington Post. "Whitney Houston, John Cleese, Coachella, Opera & More: How The
Volcanic Ash Has Affected Celebs & Entertainment." Breaking News and
Opinion on The Huffington Post.
http://www.huffingtonpost.com/2010/04/17/whitney-‐houston-‐john-‐
clee_n_541543.html (accessed June 28, 2012).
Tisler, Priit. Aspects of weather simulation by numerical process. Helsinki: Finnish
Meteorological Institute, 2006.