Improving the Forecast: Value and Public-Private Collaboration in Data Driven Weather Insights

We can’t change the weather, but we often wish we could.

We know that high-impact weather events cause the vast majority of global disasters, affecting millions of people and leading to billions of dollars in property damage every year. In 2018 the number of people killed by natural disasters worldwide – around 10 000  – was far below the twenty-first century yearly average of about 70 000, according to insurance broker Aon. However, the year did bring at least 39 weather disasters that each inflicted at least US$1 billion in damage. Only three other years (2010, 2011 and 2013) have seen a larger number of billion-dollar, inflation-adjusted, weather disasters. 

2018 saw severe flooding in Japan, India and northern China; record heat and catastrophic wildfires to California; and ten tropical cyclones that reached Category 5 strength, the second largest total on record. Tens of millions of people in cities throughout the Northern Hemisphere – including Glasgow, Montreal and Seoul – experienced the hottest temperatures ever recorded in their locations, killing dozens and straining power grids.

These events are a sharp reminder of how important timely and accurate weather and climate services are to social and economic well-being. Climate change means that we know we are facing a future with more extreme events. While many of these events occur locally or regionally, their impact is often felt globally through social and economic repercussions such as migration and supply chain disruptions.  

As professionals in the weather industry, many of us feel an obligation to help. Though we cannot alter the world’s most destructive weather events, we can use our resources to help provide deeper insight into how upcoming weather will likely affect people’s businesses and lives and help them make more informed decisions about how they can prepare and respond.

To meet the challenges ahead, it is imperative that governments, businesses and the public are supported with the best weather and climate services. We can make this happen by seeing each other as partners rather than competitors, achieving more together.


The business impacts of weather

There is not a single industry on the planet that has not been affected by weather. A recent worldwide study by the IBM Institute of Business Value found that virtually 100% of company executives surveyed believed that weather impacts at least one cost and one revenue metric in their organization. 

These executives also recognize the value that a deeper understanding of weather may deliver. In the same study, 99% of leaders surveyed said they believe improved weather insights could help reduce annual operating costs while 93% thought these insights could positively impact annual revenue growth. Interestingly, the survey also revealed that as many as 62% believed that weather insights could help drive additional growth of as much as US$ 20 million per US$ 1 billion of annual revenue – that is a 2% increase in growth per billion. Some 68% indicated that weather insights provided as a service might be more valuable than raw data. These results validate the efforts of WMO to encourage development of impact-based services. 

The Weather Company, an IBM Business, strives to create such insights for companies and consumers through a suite of data-driven solutions delivered across web and mobile platforms. The potential results can include increasing revenues, improving customer engagement, reducing costs and raising operational efficiency, and ensuring resilience across global industries.


The importance of collaboration

The value of data-driven, technology-enhanced approaches to weather may stretch far beyond profits. These technologies may also help address some of the most pressing social issues facing the world today such as feeding a rapidly-growing population and mitigating the increasing damage caused by natural – weather, climate and water – hazards. 

Achieving these ambitious goals will require strategic relationships between public and private sectors, each contributing its unique strengths in a mutually beneficial arrangement. For instance, the private sector is often more agile in adopting and integrating new technologies. However, public entities can deliver a societally-recognized degree of trust and transparency that is vital in helping people make more informed decisions. 

Big results can come from applying the private sector’s technological agility to bolster public efforts. For example, The Weather Company works with the National Weather Service to send severe weather notifications – such as flash flood or hurricane warnings – to millions of people in the United States of America (US) using its apps (mobile applications). These notifications include all appropriate attributions and are delivered without altering the original text. Thus, the technological capabilities of a private entity extend the reach of a recognized and trusted public service. These communications are valuable as warning messages are clearly communicated and the public is not confronted with conflicting messages. 


Creating the world’s most accurate forecast

Most forecast providers rely on government-provided data sources at largely synoptic-scale resolution. But this approach may be limiting, as micro-climates and geographic anomalies can create major variations across different areas. That is why The Weather Company continuously strives to create a more accurate forecast with better data.

The Weather Company seeks to build forecasts with the richest, highest-resolution, most granular data available. Every day we combine forecasts from major centres around the world with data from tens of thousands of personal weather stations and real-time data from aircraft.  These sources can generate terabytes of daily weather data and result in tens of terabytes  of historical weather information being archived to use for trend analysis.

Accurate foreccast

Once this data is gathered, we apply advanced machine-learning algorithms to help optimize each forecast element – temperature, precipitation, wind direction and speed, humidity and pressure – based on geography, time, weather type and recent accuracy. Our team of experienced meteorologists constantly monitor and adjust these forecasts as needed.

The result is the ability to produce forecasts for 2.2 billion locations with 500-metre global precision, updated every fifteen minutes. We can also create forecasts up to four months ahead of time to help businesses and consumers plan ahead.

Our commitment to innovation continues. In collaboration with the National Center for Atmospheric Research (NCAR), The Weather Company is now introducing IBM’s Global High-Resolution Atmospheric Forecast (GRAF) modelling system based on NCAR’s Model for Prediction Across Scales.  Unlike most global modeling systems, GRAF will provide three-kilometre resolution that updates hourly, delivering reliable predictions for the day ahead. This new system is designed to be able to draw on unused data such as sensor readings from aircrafts and pressure sensor readings sent from barometers found within smartphones. 

As this use of non-conventional data grows, it has the potential to overcome the lack of specialized weather equipment in many parts of the world.  This new numerical weather prediction system will feed our real-time forecast system described above. This system also maximizes graphical processing units in addition to central processing units. It is designed to run on the IBM POWER9 and nVidia GPU technology, which is behind the U.S. Department of Energy's Summit and Sierra, the world's most powerful supercomputers. 

About IBM GRAF, Antonio Busalacchi, president of the University Corporation for Atmospheric Research, which manages NCAR on behalf of the National Science Foundation, said: "This is a great example of how long-term basic research funded by the federal government has created an industry opportunity that is both good for the bottom line and protects lives and property. The pathway from research to industry in the Earth system sciences leads to better environmental intelligence and decision making." 


Smarter, faster weather-based decisions

This type of strategic collaboration is about more than sharing forecasts. In fact, The Weather Company is not really in the weather business, but rather in the decision support business. Once forecasts are created, the likely impacts of the forecasted weather – how it may affect customers, behaviours, equipment, operations and more – need to be analyzed and understood. 

The models we have developed to help users make more informed decisions are simple. We start with high-quality weather forecasts and historical weather information and then apply artificial intelligence (AI) and predictive analytics to help identify potential impacts. With those insights provided, our clients can make more informed decisions. Decisions such as an airline changing its fueling strategies in anticipation of severe weather or a retailer adjusting their inventory management because harsh conditions are likely to delay the next shipment. Our support service will enable them to more efficiently plan for these events.


Helping to feed the world with data and analytics

Another excellent example of using weather data and predictive analytics to drive value is through our newest solution, the Watson Decision Platform for Agriculture.  The agricultural industry is in need of transformation. Growers and producers are competing to serve ever expanding consumer demand for more sustainable and better quality food. At the same time, the industry must discover new ways to feed a growing population – a projected increase of 2.2 billion people by 2050 – while more and more arable land is being lost to other uses.

This is not the first attempt to introduce digital transformation to the agricultural industry. Previous solutions have been proposed to increase yields by optimizing farm data, but these failed because of their strong reliance on manual input from growers and remote Internet accessibility. Frustrated by these limitations, growers reverted to traditional approaches, causing tremendous amounts of potentially valuable data to go unused.

The Watson Decision Platform for Agriculture is different. The solution automatically gathers relevant agriculture data from a wide variety of sources. These include data collected from soil, sensors installed on farm equipment, information from other growers, high-definition visual imagery and forecasts from The Weather Company.

This data is combined into an electronic field record, which is similar to an electronic health record for a farm. The solution applies AI, machine learning and predictive analytics to this record to extract valuable insights and generate guidance for more informed decisions. These insights are displayed in a dashboard, so growers can see data and alerts related to critical elements such as weather forecasts, soil conditions, evapotranspiration rates and crop stress. With this information, farmers can make adjustments that help them produce more crops per hectare of land while improving sustainability and quality.

Another example shows that Watson AI can analyze drone footage of a specific section of a field to automatically identify and alert growers about types and severity levels of pest and disease damage. This insight can help growers save time and money while reducing the impact on their field by better understanding how, where and when to spray.

The value is not only for growers. Agriculture is an ecosystem with many parts, all of which can benefit from deeper insight into the factors that affect production. The platform automates data sharing between agriculture stakeholders across functions to create a more connected, transparent system. Food producers can better understand harvest timing and predict volumes. Lenders and insurers can set more accurate rates. Governments may be able to help improve food independence strategies by providing a common set of tools and measures to agencies and growers.


Working toward a common goal

Many in the private sector believe in a shared public safety mission that changes our responsibility from systems provider to services partner. There are many opportunities to help public entities increase their efforts without losing their identity such as using our platforms to help extend the reach of government-issued alerts. 

We could still be doing more, particularly with how technology is used. Sending important communications to citizens who are out of Wi-Fi range by using mesh communications from cell phone to cell phone is a pertinent example. The private sector can also provide deeper insight into what motivates people to act and respond, making this a rich opportunity to assist with global sustainable development goals.

The private and public sectors must also do more to harmonize our philanthropic efforts. Through our partnership with TAHMO, The Weather Company/IBM has installed stations at schools and local meteorological agencies in Africa to help deliver real-time weather data to under-served locations. These efforts are designed to help improve irrigation and agricultural resource management while also delivering climate insights for aviation, power, insurance and service-related industries.

For such efforts to prosper at a global level, private and public entities should continue to promote adoption of open data policies and common alerting protocols while also encouraging the use of new technologies and procurement approaches to help increase the reliability of equipment. 

As we seek to make the global weather enterprise more vibrant, capable and sustainable, we look forward to continuing a conversation about public and private collaborations – a critical conversation for moving forward as we serve billions of people around the world.

Share this page