Although there is still a great deal of skepticism about climate change, reading last summer reports is enough to be convinced of how delicate and precarious the balance of our ecosystem is.
According to experts in the sector, there is no doubt that the never-seen-before rains that have affected various regions of Germany, Belgium and China are closely related to global warming.
Even Ursula Von Der Leyen, president of the European Commission, defined these floods as «a clear sign of climate change».
According to data collected by the National Meteorological Institute of Germany, on the last 14 and 15 July some 150 mm of rain fell in Rhineland-Palatinate and North Rhine-Palatinate Westphalia, causing hundreds of casualties and economic damage that nearly brought the affected area to its knees.
In China, in the central province of Henan, about 200 people have been displaced because of what has been called the worst flood of the last 60 years and at least 69 people died. To give an idea of the severity and exceptionality of the event, just think that, during a 4 days time, almost the same amount of water that usually falls on average in a year in that area (640 mm) has fallen. In addition, it is estimated that about 20,000 hectares of crops have been lost.
What is the link between climate change and these extreme phenomena?
Global warming increases the ability of the atmosphere to retain humidity. Each additional degree centigrade corresponds to a 7% increase in the capacity to retain water. The direct consequence of this atmospheric phenomenon is that torrential rain is most likely to occur.
According to a paper published on the 30th of June in the Geophysical Research Letters, global warming could cause a 14% rise in frequency of these events by 2100.
In the light of these and other observations, it is clear how urgent the cut in CO2 emissions has become: the European countries declare they want to pursue that through the opean European Green Deal, an agreement favouring the introduction of new laws for a lower environmental impact.
Europe, in fact, through the Green Deal, aims at becoming the first “climate neutral block” by 2050.
How can artificial intelligence help in achieving these goals?
An interesting dissertation on this topic was presented in the Report entitled “Tackling Climate Change with Machine Learning“, prepared by a group of 16 researchers from various universities (including MIT, Carnegie Mellon University, ETH Zurich, Google AI division…).
In particular, the researchers identified 13 different opportunities to use artificial intelligence in support of policies to reduce greenhouse gas emissions:
1-in the electricity sector: production of electricity is currently responsible for about a quarter of man-made greenhouse gas emissions each year. AI can accelerate the transition to a “cleaner” energy by perfecting the calculations of energy demand forecasting (short-term forecasts) and improving the management, maintenance (predictive maintenance) and monitoring systems.
2-in the transport sector: like electricity, the transport sector is also responsible for about a quarter of CO2 emissions, but has not made the same progress towards a lower environmental impact. The transport sector is considered harder to decarbonize, mainly because most circulating vehicles require fuel. AI can support the planning, maintenance and operation of transport systems, as well as providing better forecasts, demand-wise. In addition, it can also be used in the design of more efficient vehicles concerning design and fuel supply (through alternative fuels).
3-in the real estate sector: the adoption of easy implementation techniques and cutting-edge strategies can reduce emissions of existing buildings by up to 90%, as well as allowing the construction of new buildings (smart buildings) with zero impact. Machine Learning can help identifying the best solutions for individual properties to renovate, as well as monitoring the efficiency of the same through smart control techniques (for example through remote sensing and satellite data). In addition, it can optimize urban planning, through the collection and control of data concerning factors such as population density, pollution rate, the presence of services, etc.
4-in the industrial field: AI can be employed in order to reduce the polluting impact of all the supply chain, forecasting the demand and supply in a more accurate way, identifying the low carbon emission goods and optimizing the navigation routes. Moreover, it can avoid the waste of food by improving delivery routes, forecasting demand in retail outlets and refrigeration systems. Machine Learning can also be used to study alternative materials and production processes requiring less carbon.
5-in agriculture and forestry: deforestation and intensive agriculture account for at least a quarter of total greenhouse gas emissions. An increase in precision agriculture can help reducing the amount of carbon released from the soil and also determining a better crop yield, curtailing the use of harmful fertilizers and chemicals, resulting in less need for deforestation. Satellite images and machine learning can also allow effective monitoring of greenhouse gas emissions, other than the health status of forests and moorlands, preventing the risk of fires and contributing to sustainable forestry.
6-in carbon dioxide removal processes: once reached the difficult goal of reducing harmful emissions to zero, we would still have a quantity of carbon dioxide that could cause negative consequences for the climate. It is therefore necessary to imagine technologies that are capable of removing greenhouse gases that have already been produced, as well as avoiding the emission of new ones. For this very purpose, Machine Learning solutions are still in a phase of experimentation.
7-in climate forecasting: Machine Learning models can provide better responses in terms of climate forecasting, exceeding the natural limits of current techniques, which do not have the same capacity in terms of data analysis. By analyzing the data of the past, it is possible to understand the trends of the future, in order to adapt resources and infrastructures to the type of events the model could predict in the long term. As part of the short-term predictions, Artificial Intelligence can help predict extreme weather events, enabling affected populations to protect themselves and try to contain damage.
8-in the social impact of climate change: droughts, floods and other extreme climatic events have consequences not only on the environment, but also on society, in terms of economic damage to infrastructures, crops, etc. Artificial Intelligence technologies, thanks to their potential in terms of predicting and preventing such events, can also become a strategic ally in the mitigation of negative effects on society.
9-in the field of solar geoengineering: also called climate engineering, i.e. that set of technologies aimed at reversing the global warming process in place, through solutions such as the marine cloud brightening or the stratospheric solar injection (techniques aiming at reflecting sun rays and creating a solar darkening). These technologies are not risk-free and also very controversial, although theoretically very effective. Machine Learning may deeply help estimating the possible effects of applying these techniques and try to make them safer and more effective.
10-in individual decisions: Machine Learning can help understanding the so called “personal carbon footprint” through the analysis of personal data and house ones. It can predict household emissions regarding transport, energy consumption, waste production and purchased food, allowing to suggest targeted actions to reduce its impact.
11-in collective decisions: in order to tackle climate change effectively, collective measures need to be taken quickly, involving society at multiple levels (governments, trade unions, businesses, NGOs). The affected interests are various and often in conflict, so it is necessary to understand how to encourage change, minimizing any negative consequences. Artificial Intelligence, through its analytical capacity, can provide the tools to enable data-driven decision-making and to monitor its effects.
12-in education: education plays a key role in building a society that aims at a sustainable development. Proper education enables us to know the causes and effects of global warming and provides the means to adapt to the change taking place and provide a remedy. Artificial Intelligence and Machine Learning can contribute in many ways to teaching, for example by providing personalized learning or tutoring pathways.
13-in finance: during the last years, investments in low-carbon resources are increasing in the financial sector. To encourage this choice, “green indexes” have been created, measuring the environmental impact of different resources and encouraging the purchase of “clean” services and technologies. In this context, Machine Learning techniques can be used to identify climate risks, which are more impactful in the field of renewable sources, to predict the evolution of prices and to guide the selection of the investment portfolio.
In conclusion, Artificial Intelligence, used in conjunction with other technologies and forms of knowledge, can be a valuable aid to address one of the most complex challenges of our time: the fight against climate change.
Author: Claudia Paniconi | Marketing Manager at DMBI
Photo by Li An-Lim on Unsplash
