Role of Big Data in Fighting Climate Risk

In a world wrestling daily with the threats posed by terrorism, war and political instability, we can often forget the dangers posed by the planet itself, in the form of natural disasters and extreme weather events. Aon’s Annual Climate and Catastrophe Report from its Impact Forecasting team found that the planet suffered more than 300 natural catastrophes during 2016, which collectively caused around $210 billion of economic damage, the seventh highest annual total on record. It is becoming a truism that the world we live in is getting more unpredictable. But in one very important way, that trend is going in reverse. When it comes to weather, we are finally beginning to get our collective heads around a system of historical complexity. Better data tools are already unlocking new potential, improving our resilience to weather-related risk and at the same time, opening new opportunities. Better data means better understanding of the scale of the risk, which can help reduce the impact of climate-related disasters through improved planning. It means having a greater ability to develop products like micro-insurance to help some of the world’s poorest. It can even help improve investment returns by better predicting the energy output of renewable energy projects. Humanity has been trying to find ways to influence or predict the weather since prehistory. But it’s only the emergence of big data and predictive modeling that is now truly giving us ways of understanding and mitigating the worst impacts of the weather. We live in an age of continuing unpredictability, but we have more solutions than ever before to help deal with these unknowns. The Ascent of Data Tools In the last few years, machine learning, artificial intelligence, and more sophisticated forms of data collection have provided an array of tools to help manage unpredictable events, and these are improving all the time. For instance, tools can combine information obtained from sophisticated satellite and on-the-ground data collectors, that visualize data in a way that can be quickly translated into action. “These kinds of tool are enabling opportunities to understand the impact of weather on property and people in new ways”, says Brad Weir, Head of Aon Benfield Analytics, APAC. For example, Rick Wall, exposure analyst at Talbot – which operates in the Lloyd’s insurance market – refers to a project he worked on with Aon when he was previously studying at University College London. Wall describes how “Aon was able to help insurers and aid workers in Cambodia create risk maps to identify the potential impact of flooding on local communities”. The teams created models that took into account both weather data and a range of social data, including the age and income of populations, to determine which groups were most at risk. These models could then help relief teams allocate resources in the event of catastrophic flooding. See also: Big Data Can Solve Discrimination   Better Data Analysis Means Greater Resilience The applications of these catastrophe modeling tools need not only be applied to natural disasters. A variety of industries, from agriculture to energy, can benefit from detailed insights on everyday weather trends. For instance, there are agricultural regions from India to the U.S. where small changes in climate can have serious economic consequences. A slow wet season can cause sickness in cattle, and make them less fertile, significantly reducing farmers’ assets. Specialized microinsurance packages can help small-scale farmers cope with these losses, but providers also need to be confident in the accuracy of their information, to prevent their low margins making such offerings unsustainable. This precision can be achieved with new and innovative recording techniques. Reviewing the weather alone may not be enough to give a truly granular insurance value to a farmer’s assets. Looking at other indicators can help give a more contextual picture of how climate risk is actually impacting on farmers’ bottom lines, and how effective coverage can be provided. For example, a cow’s rib cage stands out more when it is malnourished. Integrating seemingly obscure data like this helps us get a more contextual picture of the potential economic impacts of climate risk. With emerging data and tools, insurers are now able to analyze the weather risk exposure of extremely small areas, so farmers can take out insurance policies that can help them overcome weather-related liabilities where previously they may have gone bankrupt. Optimizing Power Generation From Wind, Wave And Solar Understanding and hedging against climate risk with such detailed data points can also have more directly commercial uses. Many forms of renewable power – such as wind, wave or solar – are what is known as ‘variable’ power. This means that they are highly dependent on the environment in which they operate. Wind farms don’t operate when the air is still, wave power generators fail to deliver if the sea is calm, and solar farms are of little use if it’s cloudy. Understanding the weather in advance can help power generators anticipate and plan for these risks, which is important when establishing risk profiles at the funding stage. For example, by running historical weather data through data tools, a wind-farm operator could calculate wind speed trends on a particular patch of land (and even at a particular altitude above the ground), allowing them to determine where to site plants for maximum efficiency. This also includes making sure winds are not too strong, as too much activity can damage the mechanisms of a wind turbine – or even destroy it completely, as in the recent case of a snapped turbine in Nova Scotia. This data can then be used to demonstrate to investors their likely returns. If a prospective wind farm operator is going to get financing, they can apply these data tools to their site to map the risk of weather variability. By understanding the likely energy output based on historic data, you can also build in financial mechanisms to hedge against losses – for instance, if a wind farm doesn’t get X amount of wind hours per annum. Understanding risk is good for business. Using Data to Meet the Climate Challenges of the Future We are living in a time of unprecedented technological advancement. By leveraging the power of data tools and predictive modeling we will be able to uncover deeper insights than ever before. And in an age of persistent climate risk and increasing demographic pressure, having a clear vision of what we might face has never been more important. “It is impossible to predict the future. But we know that there will inevitably be major weather and environmental disasters next year and beyond,” says Steve Bowen, Director and Meteorologist, Aon Benfield. “As computational weather forecasting capabilities accelerate and become more precise, this will only increase the importance of better communicating imminent or future hazards to those living in vulnerable locations. This understanding will help us better prepare and ultimately improve a global awareness to the growing risks of natural catastrophes and the impact of the weather.” See also: Strategist’s Guide to Artificial Intelligence   Talking Points “We don’t need an army of actuaries to tell us that the catastrophic impacts of climate change will be felt beyond the traditional horizons of most actors – imposing a cost on future generations that the current generation has no direct incentive to fix.” – Mark Carney, Governor, Bank Of England “The distribution of climatic events is not fair. In our 20-year analysis of weather extremes nine out of the ten most affected countries are developing countries in the ‘low’ or ‘lower-middle’ income category. The results of the Global Climate Risk Index remind us of the importance to support resilience policy, to mitigate the negative effects of climatic events on people and countries.” – Sönke Kreft, Executive Director, Munich Climate Insurance Initiative