Hello everybody,
today I would like to present two articles on the use of artificial intelligence in predicting the weather. After reading the articles, please answer the following questions:
1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
4. Where can weather prediction be used?
Articles:
1. https://interestingengineering.com/ai-might-be-the-future-for-weather-forecasting2. https://medium.com/syncedreview/ai-local-rainfall-forecasting-using-weather-radar-maps-d4aa041343a
Hi Artur.
ReplyDeleteinteresting issue, in fact.
Obviously, what AI can be used in, if not in forecasting? So, yes, in my opinion the scale of potential AI possibilities for this purpose will be increased in the future, I guess not so far away from today.
I suppose that forecasting using AI methods will be more accurate than the other ones, because of their intelligence sewed into algorithms. They will online learn current data and apply for further predictions, limiting errors and providing more and more accuracy. While traditional methods just correspond to actual data and must relay also on their specificity.
In my opinion, we will rather need more efficient computers rather than faster calculations methods. Efficient, I mean prepared for storing a lot of particular data, gathered all the time, online.
Weather prediction has many applications. All of us check it before we prepare for our planned activities and voyages. Very precise weather forecasting helps to prevent from catastrophes and even avoid disasters. So this way we protect our life and property.
Many people relay on weather forecasts to decide what and when do their work, for example farmers. Electricity and gas companies rely on weather forecasts to anticipate demand, which can be strongly affected by the weather.
Weather forecasts can be used to invest in the commodity market. Because the aviation industry is especially sensitive to the weather, accurate weather forecasting is essential.
Br, marta
Thank you for your reply. I agree with you about the need for more accurate forecasting, because industry is largely dependent on weather conditions (aviation, agriculture). As for the processing of data, whether the efficient machines themselves are enough to significantly accelerate the artificial intelligence in forecasting?
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteYou know what? Forecasting anything is a fortune-telling of tea leaves. I'm more inclined to weather predictions based on the probability calculus (meteo.pl) than the artificial intelligence algorithms. For me personally forecasting anything, choices, weather, even my own expenses is a great challenge.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
I think not. Any prediction whether AI or other is burdened with a large error or error. For me personally ;-) I always treat the weather forecast as a joke. I never believe in weather people.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
I don't think that very powerful computers have to be used to predict the weather. This, in my opinion, is a simple analysis of facts and data. Maybe I'm wrong, but ...
4. Where can weather prediction be used?
For sure in sailing, other water sports. Mountain climbing, construction, agriculture. Everyone, however, is planning their activities, trips, something, taking into account the weather forecast :-D
Thank you for your reply. From your statements, it seems that you do not believe in weather forecasting, because as you have described it is a fortune-telling. However, at the moment we are able to predict the weather fairly well for 2-3 days, while the long-term forecast leaves much to be desired. You also pay attention to the large errors that are burdened with artificial intelligence, perhaps because the algorithms used are not fully adapted to the data obtained or it is just too early to accurately predict weather conditions with the use of AI.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteI think it will be used on a large scale but its accuracy won't be so large as if predicting wather for some local and short term weather predictions. As it is going larger it is less predictable.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
I think that after good hyperparameter tunnning and lot of historical data it is going to be more efficient and accurate. However it may be not so good in some special cases and lng term climate changesif we do not own corresponding historical data.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
We need suuper effcicient computer to predict the weather in a fast way from large amounts of data. That means forecasting the weather is usefull if we can used and have time to share the data with others. The most important data come from a very recent period so it is importatn to include it in a predictions and make calculations fast.
4. Where can weather prediction be used?
I think everyone sometimes use weather predictions. So it may be used on a daily basis or to plan some events on an open air by farmers etc.
Thank you for your reply. I agree with you regarding the amount of data and their timeliness. Ideally, the data should be as current and comprehensive as possible. So we pay special attention to the quality of our data, because uncertain or incomplete data affect the quality of calculations and the algorithm used.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteOf course it's possible. What the weather will be like depends on thousands of factors, and in addition, these factors are interlinked over time. The main way to predict is observation. Dangerous phenomena coming in the near future can be seen by observing the surroundings. Recently, I have read that Artificial Intelligence from Google has learned to predict dangerous weather. The weather forecast is a fortune teller from the bushes. It does not make sense for longer periods of time. This is an undetermined chaos system.
Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
Whether it will be more accurate or not, it is enough to calculate the multidimensional distribution of Receiver Operating Characteristic. On the other hand, the current models are also numerical. Only the final model has a different source. In addition, the weather forecasting network will learn from the same data as the existing models. It is interesting that weather forecasting became possible only after the invention of the telegraph, when it was possible to collect many observations at the same time.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
Computer modelling is currently the most effective tool for the analysis of complex meteorological problems. The main advantage of this approach is the high accuracy of the obtained results and the possibility of dynamic change of physical parameters of the model or its working conditions, without the necessity of constructing a virtual representation of the analyzed phenomenon or object. But it is also the most complicated and costly in terms of calculation. Maybe it will be possible to build a system that will do it faster and cheaper.
4. Where can weather prediction be used?
It is obvious that nowadays from the quality of the weather forecast, will prohibit the functioning of the whole world. I would have to change all occupations, events and services so that I could answer that question.
Thank you for your reply. Your statement brings a lot of interesting and new things. I agree with you in everything except the amount of data needed to calculate weather parameters. On one of the fairly reliable websites (unfortunately I do not remember), I saw information about weather models that used millions of weather data, where whole clusters of computing machines were used for calculations. The algorithms were reportedly very optimized.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteYes, I think so. Artificial Intelligence may speed up big data analysis and increase the accuracy of prediction. Weather phenomena are very predictable when you have sufficiently collected data so AI can deal very well in this situation.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
Nobody knows it - only after a reliable comparison will it be able to deduce something more. For now it looks like AI will be better. It will certainly be able to improve accuracy of predictions into long-term precipitation forecasts.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
Efficient computers are needed to analyse big sets of gathered data. Weather prediction must be based on a lot of data gathered from different sensors to be accurate.
4. Where can weather prediction be used?
Predicting weather is helpful primarily in agriculture, planning flights, shipping, tourist trips and in many other areas where weather is crucial.
Thank you for your reply. I agree with you as to the amount of data needed, their accuracy and how they are processed. Learning algorithms must have the best data (it does not mean that they are selected) to be more accurate. The computing power of computers can significantly accelerate the data analysis and the quality of the predicted progression, but I would also not forget about the optimization of the algorithms used in the analysis of these data.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteThe weather is dependent on many factors. I think that artificial intelligence can be used to predict the weather. Currently, the verifiability is not too high, but if we used AI on a large scale, it could learn and perhaps we could better predict the weather at least for the next few days.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
I think the most effective way of doing this is to combine these two approaches. Although it will still be very complicated and perhaps not as effective as we would expect.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
If we want to analyze such large data sets that can be collected about the weather, it is necessary to have a large and computationally efficient infrastructure. If we could create and learn a model, we could start using less computing power, but the weather is so variable that such a model will not be effective in the long run either.
4. Where can weather prediction be used?
Weather forecasting is useful at almost every part of our lives. First and foremost in maritime, land and air transport, as well as in predicting weather anomalies that are dangerous for people. The weather is also important for military operations, but also for the more prosaic organisation of outdoor events.
Thank you for your reply. I believe that you are right about teaching artificial intelligence on weather data. However, in my opinion, models created using weather data are very flexible, because the data is constantly changing.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteWell, I don't think that predicting anything that is connected with nature and its power on the basis of algorithms can be treated as something reliable. I will stick to normal weather forecasts.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
As I already stated, I don't think so. Nature cannot be fitted into any algorithm functioning rules whatsoever.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
I don't think that one really need a very efficient and powerful computers to provide a decent and reliable weather forecast. It's a matter of appropriate data analysis ‒ simple and not very complex data per se.
4. Where can weather prediction be used?
It's something that everybody needs, regardless of what they do.
Thank you for your reply. As for the prediction in nature you have a lot of reasons, but the weather forecasts are based largely on models created by the whole team of analysts, programmers and other specialists in the field. We use millions of data for analysis, which in addition to optimization should be properly processed and it is best in real time.
DeleteHi Artur,
ReplyDeleteThank you for interesting article, regards your questions:
1.Do you think that artificial intelligence will be used on a large scale in predicting the weather?
Due to the fact how complex weather forecasting models are, and how big amount of data they have to process to make the prediction AI is perfect match for this job. From what is presented in the articles you found for us, AI is already used on a big scale in weather forecasting, and next implementation are waiting around the corner, just like Deep Thunder presented by IBM.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
Weather forecast obtain through AI algorithms can be more accurate but only for shorter timescales prediction. The limitation hear is how rapidly changing Earth weather is, and in how complex is the weather system we live in.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
From what I read in presented articles, computer power is not the problem in this case, we are limited by nature. Based on general algorithm computing I would say that we can increase weather forecasting accuracy by creating short term prediction for smaller regions at the same time.
4. Where can weather prediction be used?
There are plenty of application for weather prediction, starting with alarms when heavy storms, hurricanes or tornado approaching, and ending with worldwide cargo flow optimization.
Thank you for your comprehensive answer. I believe that you have many reasons for the accuracy of predicting the weather using artificial intelligence for a given area. It is easier to limit yourself to a specific area to collect data from it. In addition, such data can be processed more quickly and more efficiently to match the model that meets our criteria.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteWheater forcasting is data modeling task, so I believe soon, when we will have enough granular data for this task and enough computing power to train some deep recursive neural network we will start using this kind of predictor instead of current ones.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
Even today in some specific use cases AI prediction is more accurate than clasic ones. I read article (https://www.theverge.com/2019/2/26/18241632/google-deepmind-wind-farm-ai-machine-learning-green-energy-efficiency) that using AI one of US wind farm got much more accurate predictions of energy production.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
I think weather predicion is similiar problem to detecting particles and forces in particle accelerator. There is millions of possible factors and we need first to create some filltering algorythm to shape input to reasonable size, without it data stream will be to big to use it.
4. Where can weather prediction be used?
I believe almost everywhere. Currently our energy production is more than ever dependent on weather (sun, wind, tides). More over our electronic and infrastructure is more susceptible to extreme weather events, so we need to predict it.
Thank you for your comprehensive answer. You are right about the amount of data being processed and that to limit it to a certain size. The algorithm learning artificial intelligence may have a problem with adjusting to too large data "obscuring" the area of research on the creation of a fairly accurate model. In addition, artificial intelligence develops very dynamically and will certainly be used when creating variable weather models.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteI don't think so. AI could support people in forecasting wather. The forecasting process is a hard task. We need a huge amount of data from many sensors. In short-term AI maybe can improve forecating accuracy but in long term it is difficult to say. The weather is changing and the historical data can be wrong because in last april it was sunny and hot and in this year is cold and rainy. Researchers still working on algorithms to improve theirs accuracy. We will see what the future brings.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
I think in small scale short-term AI forecasting can be more accurately but in long-term forecasting im not sure. The forecasting is a hard task.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
I need very efficient computers because there is a huge aomount of data that must be analyzed to forecast weather accuratly. We need data from many sensors located in many places. Mayby when we used a small amount of data for one region the process could be faster for this region.
4. Where can weather prediction be used?
I think in short-term weather forecasting for small regions. Maybe in industries which work depends on weather.
Thank you for your reply. The huge amounts of data needed to produce more and more accurate weather models require high computing power and algorithm optimization. In a smaller area, it is easier to predict the weather, however, it is dependent on the weather prevailing in a much larger area.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteI think that artificial intelligence can be used for this type of task, especially since nature has different types of patterns and cycles that artificial intelligence can detect.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
If they adapt the right method, then yes. Due to the fact that artificial intelligence is able to quickly catch things that an ordinary person bypasses or does not notice. Thanks to this, the effectiveness of prediction can increase, but at the beginning the cost of preparing such an algorithm will be quite high.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
High-quality computers are needed due to the reduction of computation time and the increase in prediction accuracy, as with any such project. The weather is so complicated that various types of microclimate are created, eg by the sea, in some places a zone may appear where practically rain does not fall while all around the rainfall is intense. The same must be taken into account that some rivers attract clouds, which creates a microclimate around them that is governed by its laws. You could exchange many of the types of exceptions necessary to take into account.
4. Where can weather prediction be used?
In air systems for setting flight plan, in Maritime systems for sea navigation plans. For agriculture, for the management of hydropower plants, wind farms, solar farms in general too.
Thank you for your reply. I am glad that you believe in the possibility of artificial intelligence in this area. I think that the beginnings are actually difficult and very expensive, but something should start with something. The use of appropriate algorithms to build accurate weather models will improve the quality of forecasts. Unfortunately, as in other cases of using artificial intelligence, we need time and many attempts to make the system as accurate as possible.
DeleteHello Artur, thank you for the interesting subject.
ReplyDelete1. As Andrzej said, the weather prediction is a task for a fortune teller, because it is an undetermined chaos system. I think that it would be doable in future, however, I would say that right now it would be very hard to achieve.
2. I think that if there would be an accurate and reliable way to make predictions then those models eventually would be more accurate than the weather forecasted by current methods.
3. Wikipedia says that even with the increasing power of supercomputers, the forecast skill of numerical weather models extends to only about six days, and I think there is something in that statement thus we need further algorithms research to achieve better results.
4. I think that one of the good examples that already has been mentioned is the share market, where you need to predict something that is very variable during the time.
Thank you for your reply. In fact, weather forecasting is a bit like a fortune-telling, but more and more accurate models are used to improve the quality of forecasts. Of course, every day we are growing more and more in this field and we can say that we are able to predict the weather quite accurately for a maximum of three days. However, a lot of work and trials have yet to be done in order to be able to accurately accurately create a long-term forecast.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteIf the result of their prediction will be much more better it will happen for sure.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
That's really hard question. The weather is very unpredictable and change very fast. Todays method aren't very precisely becouse our climate change. Maybe AI could change it, but I can't answer on this question.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
All of todays models are based on historical data and actual weather. To speed up this process we could don't checking historical data, and based only on actual weather and a few logical statement like if raining then later humidity would be higher but I'm sure that results of that model will be fatal. Faster but worse.
4. Where can weather prediction be used?
I think that it is using in many aspect e.g. transport, tourism and daily routine when we planning about what time we should go running ;)
Thank you for your reply. You're right about predicting the weather. It is very difficult and more often resembles divination. Artificial intelligence must make many attempts to create the most-realistic models based on a multitude of variable data.
Delete1. I think so. I've seen several articles on this topic before. Artificial intelligence is getting better and can predict more and more complex issues. Predicting the weather is a complex issue that requires checking many factors. If we would like to prepare training data, we should identify the most important ones. You would also need a very large amount of test data from the most diverse weather conditions. I'm just wondering if it would be possible to create one system that predicts the weather for, say, a specific climate or latitude, or whether each country would have to have its artificial intelligence to predict the weather.
ReplyDelete2. It's hard to tell. I think that numerical prediction of the weather could work in short periods of time, i.e. in predicting weather for 1-3 days. Atmospheric phenomena and weather are sometimes unpredictable and can change within an hour. Although it seems to me that with rapidly changing parameters, artificial intelligence would be able to predict some kind of weather breakdown during the day.
3. As I think about the acceleration of calculations, I always have a picture of quantum computers in my head :D In contrast to efficient computers, they are needed to predict weather in short intervals, so we need to analyze the data collected by measuring devices and analyze them.
4. I think everywhere. Weather forecasting has a huge impact on all types of transport. In addition, it is important to alert you when the weather breaks, such as heavy rains, typhoons, hurricanes, etc. Also, weather forecasts can tell us a lot about the ever-changing climate.
Thank you for the comprehensive answer. I agree with you regarding the selection of training data. It's easier to predict weather for a specific area than globally. I think it is worth trying to create a system that covers countries located in the same climate zone, because a lot of data can cover it. Predicting weather is important in many areas and it is used for many purposes.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather? The process of predicting weather patterns is a very complicated science. It requires the need to analyze and decode massive data sets gathered from thousands of sensors and weather satellites every day. But like any kind of forecast, weather forecasting is something of an educated guess. Since we cannot control the weather, the best meteorologists can do, is to use past and present data and patterns to attempt to predict the future.
ReplyDelete2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
Chaotic nature of the atmosphere and incomplete understanding of the processes mean that forecasts become less accurate as the range of the forecast increases.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
The complicated equations which govern how the state of a fluid changes with time require supercomputers to solve them.
4. Where can weather prediction be used?
The prediction of the weather through application of the principles of physics, supplemented by a variety of statistical and empirical techniques. In addition to predictions of atmospheric phenomena themselves, weather forecasting includes predictions of changes on Earth’s surface caused by atmospheric conditions—e.g., snow and ice cover, storm tides, and floods.
Thank you for your reply. You have taken a fresh look at the conversion of weather data using artificial intelligence and areas of application of weather prediction. It is difficult to properly build a good weather model, but through experience, historical data, the use of computing power of computers, artificial intelligence and other similar elements, we can make more accurate and faster measurements.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteI think yes, it will be. We have access to lots of data, and looks like in every field where lots of data is available, AI solution pops up.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecast by applying weather data to weather maps?
It might be, but not by much. The computation cost, measurements precision and grid size requirements grows super-polynomial and we will hit the wall we won't have resources to scale.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
Well, there is lots of equations to be solved before prediction turns into history, for dense grid of points. There is no other way than use efficient computers, no magic shortcut.
4. Where can weather prediction be used?
Short term - agriculture, transport, construction sites, probably all human outdoor activities. Long term - agriculture again, planing of rain/water infrastructure, real estate planing?
Thank you for your reply. I agree with you regarding the use of artificial intelligence in creating weather forecasts due to the huge amount of data. Such a system will support meteorologists in their daily work, where the bane is the processing of such large data sets.
Delete1. Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteI think that in the future we will notice a very large involvement of AI in almost every aspect of life and in my opinion one of this aspects will be the weather prediction. First of all, it’s easier than leaving all the arithmetic to the people, it is more accurate and eliminates the amount of mistakes making during the proccess. So I am sure that artificial intelligence will be used on a large scale in predicting the weather.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
I am not a specialist on this topic and I must say that today was the first time when I read any article on this subject, but as far as I am concerned using AI in predicting the weather indeed will be more accurate than the traditional way of forecasting it by applying weather data to weather maps. Computers are for sure more efficient and can proccess more data at once, so they may concern greater amount of factors influencing the weather. What is more, as I mentioned answering first question, there is less room for any mistakes than in case of predicting weather by actual people.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
Unfortunately I don’t know any other methods to speed up calculations involved in the proccess of forecasting the weather, but I am sure that in case to do it porperly – using AI – you need very fast and efficient copmuters, because of the amount of data that must be taken under consideration, like wind’s direction, sea surface temperature, etc, etc.
4. Where can weather prediction be used?
I think that the most obvious implementation is in the tv news, when at the end all the viewers expect to hear about weather predictions for the next couple of days ;) but there are also some other things where we can use weather prediction. For example, it would be very important before the sport events or some concerts taking place on the open field. Moreover, people would like to know specific weather conditions before they plan their wedding. As for the natural disaster, it is crucial to know beforehand, if on some location we can expect for example flood, hurricane, earthquake or tornado, because the proper preparations would be in need.
Thank you for your reply. I also hope that artificial intelligence will speed up and eliminate at least some of the errors that may arise during the calculation. In addition, very rapid development of computer science will contribute to the development of artificial intelligence in this field.
Delete1.Do you think that artificial intelligence will be used on a large scale in predicting the weather?
ReplyDeleteI think that artificial intelligence will be applied on a large scale in predicting the weather soon. Predicting weather is very complicated process and it requires a lot of data from many sensors. The authors mentioned that AI is able to provide better the accuracy of predicting weather. Some processes in the weather are recurrent and they appear periodically. The AI can learn how the weather was in past based on deep neutral networks. The solutions already exist but I hope that they will be more practical in the near future because I think that AI will determine the future of the world.
2. Will the weather predicted using artificial intelligence (numerical prediction of weather) be more accurate than the weather forecasted by applying weather data to weather maps?
I do not know but I suppose that the weather predicted based on the AI may be more accurate. I am not a specialist of predicting weather but when I read the articles about the comparisons between human translator and the machine translator sometimes the authors say that the machine translator is more accurate so maybe the situation with weather is similar. I think that it requires a lot of data to compare which approach is better. There is one more option – combining the two mentioned options.
3. Why do you need very efficient computers to predict the weather, or maybe you can speed up calculations in a different way?
Predicting weather is based on very big data sets so if you want to do some analysis, I mean, predict weather, you need very efficient computers or server where you can perform calculations. There are some algorithms, for example, to decrease the dimensions of analysis but there will be still lot of dimensions. If you applied the less efficient computer, the calculations would be done very long and you could not announce the predicted weather because you would not do on time.
4. Where can weather prediction be used?
I think that the prediction of weather may be applied in many fields of industry in the world. The air industry depends on the weather. If the wind does not blow in the good direction, the airplane is not able to take on and take off. If you are able to predict whether you can for example change the place of take off. The following industry are also strongly dependent on prediction of weather: agriculture, marine industry and tourism.