The Heat is On for Better Climate Data

Jan 18, 2018   |   Lucas Joppa, Chief Environmental Scientist

About a month ago, in Nature, I wrote that time is too short for companies to ignore one of society’s greatest challenges – managing and adapting to our changing climate. In the past week, several different reports have been published that indicate that time is shorter and the challenge greater than we previously thought.

Today, NASA and NOAA released their annual data on global temperatures and climate-related trends of 2017. While there are slight statistical differences between the two reports, they show that last year was among the third hottest ever recorded and global warming continued its upward climb. These results come on the heels of a draft U.N. report that suggests warming is likely to exceed the 1.5C ceiling set by the Paris Climate Agreement by the middle of this century, although just how troubling these findings are is a debate unto itself, as researchers also point out in a recently published report in Nature.

These reports provide valuable information about the scope of the challenge ahead of us and the stubbornly slow progress being made and play an invaluable role in helping society understand the true risks associated with our changing climate. While the facts and figures will attract most of the attention, I want to focus on what else these reports tell us about the state of climate and environmental science and what we’ll need to improve upon to address climate change.

We need more data. While we have global data on temperature, we need more data on a more granular level. Scientists still struggle to predict the effects of climate change at the resolution of cities or regions, or over time-frames of months or weeks — largely because they don’t have the kinds of data needed to constrain their models of earth’s systems appropriately. We need to close this data gap to provide better focus and prioritization, and that will take a coordinated effort between academia and the public and private sectors to fund this research and make the findings publicly available.

We must rapidly convert data into information. In this digital age we often find ourselves data rich, but information poor. For instance, high resolution imaging platforms like satellites are producing a proliferation of imagery of the earth. Yet our ability to turn that data into information on where our forest, fields, water, and infrastructure lags well behind. In the United States, the best available data sets on land cover, at a resolution of 30 meters, are nearly seven years out of date. Globally, the picture is much worse. This lack of information makes it difficult for housing developers, foresters or other land use planners to make evidence-based decisions about how to efficiently use our natural resources. And large delays in the face of rapidly changing systems leave us flying blind to climate-based risks.

Changing this reality is part of why I was so proud to serve as a member of the previous Advisory Committee on the Sustained National Climate Assessment. This group had the responsibility for advising the U.S. government on how technology can help move the monumental National Climate Assessment, produced every four years, from a static, bulky process into a more dynamic one, constantly updated with the latest information about our climate systems. The recent announcement that this critical, cross-sector committee that has been reconstituted under the strong leadership of New York Governor Andrew Cuomo will help ensure that this work will continue, providing more information to those who can use it.

AI can help. AI can help humans collect more data, in a more efficient manner and ensure it is actionable. Already, AI systems can be trained to classify raw data from sensors on the ground, in the sky or in space, using categories that both humans and computers understand, and at appropriate spatial and temporal resolution. With enough data on which to train, and with human feedback, these systems can learn to tag photos, acoustic recordings and genetic information with species names; or to convert satellite imagery into information on water availability at a landscape scale. This in turn helps tell us how species are adapting to changing habitats and temperatures, or when and where and how severe water scarcity issues may be before it becomes an issue.  AI can even help improve the results of existing climate models – recently researchers showed that ‘super resolution’ AI techniques can be used to improve Earth-system models by statistically ‘downscaling’ low-resolution projections of around 100 square kilometers to high-resolution ones of around 12 square kilometers that are more relevant to local land-use planners.

Microsoft is working on these issues through our AI for Earth program. There is much to be done. Resources are scarce. The needs are great. At times like these, in the face of these reports, it can seem like an insurmountable challenge. But we think good data helps galvanize effective action, and we believe that AI can help by providing more information, more quickly. Decisions about what actions to take will be easier to make — and less vulnerable to politicization — if we know what is happening on Earth, when and where. It’s time to put AI to work for the planet.

Learn more at:

Tags: , , ,