Skip to Main Content

Combating drought with AI and the cloud

When an area of land becomes a desert, it loses its water, wildlife and vegetation – a process known as desertification. While deserts are naturally occurring in some parts of the world, unnaturally increasing areas of land are continuing to dry out. At present, more than a quarter of our planet is in danger of desertification due to the effects of climate change.   

In Spain for example, the Ministry for Ecological Transition has estimated that desertification will continue to expand throughout the Iberian Peninsula over this century, extending from the Costa del Sol to the foothills of the Cantabrian mountains – 80 percent of the entire country’s surface.

Climate change has accelerated the loss of fertile soil, while water shortages and prolonged periods of drought have caused further aggravation. This is particularly apparent in the southern half of Spain, where water resources are limited. Currently, the reservoirs in the peninsula are at just under 60 percent capacity. Since last year, the volume of water accumulated has been reduced by around 12.5 percent.

According to data from the United Nations Food and Agriculture Organization, we will need to increase crop production by 50 percent, to meet the world population’s food demand by 2050. If countries like Spain lack suitable soil to help increase or even maintain current crop production, we will find ourselves in a devastating situation.

Artificial Intelligence and human ingenuity
In recent years, possible solutions have been proposed to counteract this loss of viable land. The first step, which we have already overcome, is to measure the current situation for clear context, which has been made possible through technologies such as cloud computing and the Internet of Things (IoT).

The data collected and analysed by artificial intelligence (AI) algorithms and machine learning is allowing us to formulate solutions in a faster and more agile way than ever before. Thanks to AI solutions, areas can increase their efficiency, thanks to the detection of consumption patterns and more accurate forecasts.

“These new technologies allow us to make decisions to increase production and anticipate the consequences of climate change. The future of the field involves investment in technology that is accessible to all, and Artificial Intelligence can be a great accelerator of sustainable and responsible progress,” says Carlos de la Iglesia, Director of Communications and Philanthropies at Microsoft Ibérica.

An example of Microsoft’s commitment to helping combat problems like desertification is its AI for Earth initiative , which plans to invest 50 million dollars to promote projects in which AI can accelerate the preservation of the environment.

Artificial Intelligence can be a great accelerator of sustainable and responsible progress

Murcia fights against desertification .
In the southeast of Spain, Karim Claudio, data analyst at Cetaqua , participates in the AI for Earth program with the purpose of integrating Microsoft Azure machine learning tools with geospatial data analytics and machine vision techniques. Using these tools, Claudio and his team can develop predictive models of the contributions and demands of water in the agricultural sector of the Murcia region.

“The growing digitization in all processes of the water cycle means that more and more data – such as meteorological information, reservoir and groundwater level, river flow and agricultural production – is available. This allows us to predict the demand for water, whether it’s for the irrigation of crops, or working towards Agriculture 4.0,” Claudio states.

The ultimate goal of this system is to provide decision support tools, in addition to creating predictive models that contribute to the sustainability of water resource management – something that will become more critical in Spain, as each year passes.

Supporting farmers
Another example of how technology is helping to increase the efficiency of water resources can be seen in the collaboration between Microsoft and Ibercaja, to help support Aragonese farmers to optimise their farms through the development of smart agriculture.

By installing IoT devices in the field and analysing data in the Microsoft Azure cloud, farmers can make more efficient use of natural resources, particularly water. The AI system ​​analyses the evolution of the established indicators such as the environmental temperature and humidity, before providing suggestions to help farmers make more informed decisions in each stage of their production cycle. As a result, farmers can also now track detailed information about their crops, in real-time.

A global revolution
Below are some examples of other programs that already take advantage of AI around the world via the AI for Earth grant to help combat the effects of drought:

Creating a drought index
The Argentinian company S4 Agtech is a pioneer in the development of the first index in the world to cover the effects of the drought. The index itself has been created using measurements from select NASA satellites, which in turn have been analysed with Microsoft Azure machine learning tools, providing risk assessments for a number of eventualities that could occur in crop-growing regions.

“In Azure, we found a complete and friendly platform able to cover our current needs and accompany us in our growth. It offers us support and management times allowing us to occupy our main tasks,” says Sebastián Priolo, CTO of S4.

The company uses Microsoft’s cloud to create valuable information that encourages farmers to invest in the right areas. This allows them to reduce predicted risks, while increasing the use of sustainable resources.

Cooperation for the common good
Lester Mackey, a Microsoft statistical machine learning researcher in New England, and Judah Cohen, a climatologist at Atmospheric and Environmental Research, are pioneering preliminary models of sub-seasonal temperature and precipitation forecasts.

The support of Microsoft’s AI for Earth initiative has allowed them to improve and refine their techniques in order to better contribute to the protection of the environment. This initiative began with the conviction that machine learning could help improve its method to generate sub-seasonal forecasts by collecting valuable information from large amounts of historical climate data.

“The techniques explored by the project have an enormous capacity for application in climate forecasting and, in addition, they have an enormous capacity for application in broader social and economic domains,” states Lucas Joppa, who is responsible for the AI ​​for Earth program.

Microsoft’s commitment
Aware of the importance of preserving natural resources, Microsoft has established a series of internal policies aimed at reducing the impact that the company has on the environment and, by extension, on its water resources.

Since 2012, the company has had an internal carbon tax, which dictates that 15 dollars per metric ton is paid on all carbon emissions. The objective of this initiative is to make business divisions financially responsible for reducing their carbon emissions. The funds from this tariff maintain its carbon-neutrality. Following this commitment, Microsoft also joined the Climate Leadership Council (CLC), an institution founded by business and environmental leaders to put a price on carbon and spread awareness about necessary measures.

At its headquarters in Redmond, Washington, Microsoft has also begun work on seventeen new buildings that will eliminate all fossil fuels and run on electricity with a 100 percent carbon-free origin. In turn, the amount of carbon associated with construction materials will also be reduced by at least 15 percent, increasing to 30 percent through a new online tool. With this technology, Microsoft will have the first large corporate campus to achieve the goals of zero-carbon emissions and zero waste.


A sustainable cloud
Microsoft is also firmly committed to the research and development of promoting greater efficiency and use of renewable energy in its datacentres. Through cloud-based technology, IoT, blockchain and AI, the company will be able to monitor performance, and streamline the reuse, resale and recycling of its datacentre assets. The commitment to sustainability also extends to water resources, establishing a new water replenishment system that will reduce the consumption of datacentres in more arid regions before 2030.

Last month, Microsoft announced new datacentres in Sweden, designed firmly with sustainability in mind. These datacentres will be one of the most advanced and sustainable in the world, using energy from renewable sources and executing plans of operations without waste. 

“We want our datacentres in Sweden to be among the most sustainable in the world and we have the ambition to achieve operations with zero CO2 emissions. The design of the datacentre we are developing will reinforce Microsoft’s ongoing commitment to the transition to a sustainable, low-carbon future,” says Noelle Walsh, Corporate Vice President of Cloud Operations and Innovation at Microsoft.

“Humankind faces great challenges ahead of which the future of the following generations depends on,” Carlos de la Iglesia adds. “Today, where changes take place in a more accelerated way than in any other period of history, we are convinced that many answers will be reached with the help of artificial intelligence, which already allows scientists from all over the world to increase their capabilities and accelerate their investigations.”