There has been a lot of talk lately about “big data” and how the future of innovation and business success will be dominated by those best able to harness the information embedded in big data. So how does remote sensing play a role in this discussion?
We know remote sensing data is big. For example, the NASA Earth Observing System Data and Information System (EOSDIS), which includes multiple data centers distributed around the U.S., currently has more than 7.5 petabytes of archived imagery. Within the EROS data center alone there are over 3.5 million individual Landsat scenes totaling around 1 petabyte of data. And this is but a subset of all the past and currently operating remote sensing instruments. There are many more, particularly when considering the various international and commercial satellites, not to mention the array of classified military satellites and the many instruments yet to be launched. Remote sensing imagery therefore certainly satisfies the big data definition of size.
But what about information content? A significant aspect of the big data discussion is geared towards developing large-scale analytics to extract information and applying those results towards answering science questions, addressing societal needs, spurring further innovation, and enhancing business development. This is one of the key aspects – and challenges – of big data, i.e., not just improving the capacity to collect data but also developing the software, hardware and algorithms needed to store, analyze and interpret this data.
Remote sensing researchers have long been using remote sensing data to address localized science questions, such as assessing the amount of developed versus undeveloped land in a particular metropolitan area, or quantifying timber resources in a given forested area. Subsequently, as software and hardware capabilities for processing large volumes of imagery became more accessible, and image availability also increased, remote sensing correspondingly expanded to encompass regional and global scales, such as estimating vegetation biomass covering the Earth’s land surfaces, or measuring the sea surface temperatures of our oceans. With today’s processing capacity, this has been extended yet further to include investigations of large-scale dynamic processes, such as assessing global ecosystem shifts resulting from climate change, or improving the modeling of weather patterns and storm events around the world.
Additionally, consider the contribution remote sensing makes to the planning and development of transportation infrastructure in the northern hemisphere, where the opening of new trans-arctic shipping routes and changes to other existing high-latitude shipping routes are being predicted using models that depend on remote sensing data for input and/or validation. And also consider agricultural crop forecasting, which relies heavily on information and observations derived from remote sensing data, and can not only have economic impacts but also be used to indicate potential regions of economic and political instability resulting from insufficient food supplies.
Such examples, and others like them, represent a logical progression as research and applications keep pace with greater data availability and ongoing improvements in processing tools. But the field of remote sensing, and its associated data, is continuing to grow. What else can remote sensing tell us and how else can this immense volume of data be used? Are there relationships yet to be exploited that can be used to indicate consumer behavior and habits in certain markets? Are there geospatial patterns in population expansion that can be used to better predict future development and resource utilization?
There’s a world of imagery out there. What are your ideas on how to use it?