Google offers global geospatial analysis for the masses
Tuesday, February 20, 2018
A plethora of global remote-sensing technologies acquire terabytes of environmental data minute by minute. Much of it is routed to and through the cloud. This was not the case perhaps a decade ago.
But, by the 21st century's second decade, this growing fount of geospatial information prompted Google, the world's third-most valuable IT company (after Samsung and Apple), to develop a geospatial "platform" — set ways and means whereby a computer user can crunch the abundant earth data quickly and smartly, at no cost and without any other undue restrictions.
Known as the Earth Engine, the platform gives users intelligent access to the data collected by the earth's vast multitude of sensors and satellites. It's already a crucial mainstay of a new, mission-dedicated digital infrastructure, which Google nearly touts as a public good. But make no mistake, the company owns and will undoubtedly maintain its complete discretion over it — even as the world only slowly and reluctantly lurches into action to curb climate change.
It's clear that the platform — designed to make planetary-scale remote sensing analysis quick, easy and convenient — materially assists the drive to reduce greenhouse gas emissions. With the Earth Engine, Google has deftly brought to bear its awesome computational capacity on a range of urgent ecological issues such as deforestation, drought, disaster prediction and relief, public health, food security, water resource management, and overall climate monitoring and environmental protection.
But the sad truth remains: Even large-scale technical programs like Earth Engine are dwarfed by the political challenge of forging the international cooperation that would be necessary to make even nominal progress against climate change in the near future.
In a recent paper from the journal Remote Sensing of Environment, a group of six Google scientists and operatives discuss features of Earth Engine. They make it clear that Google has come up with ways for visitors of even the smallest commercial sites to conceive climate change solutions and conduct geospatial analyses from data produced by remote-sensing devices.
Earth Engine "is unique in the field as an integrated platform designed to empower not only traditional remote-sensing scientists, but also a much wider audience that lacks the technical capacity needed to utilize traditional supercomputers or large-scale commodity cloud-computing resources," say the authors.
From the beginning, Google has eyed five major factors or realities that beckon the new platform:
- Large-scale number-crunching capability is ubiquitous, but really the expertise required to use it is scarce.
- Google has every interest in devising a platform that would make global-scale remote-sensing analysis much more user-friendly.
- The company can readily locate a large catalog of data sets with its trademark massive CPUs to promote unheard-of interactive data exploration by new users and groups.
- A speedy and easy-to-use system would invariably accelerate scientific discovery.
- The evolving architecture would slowly become more amenable to programming processes currently ill-suited to it, such as recursion.
Google has seen for years that the petabytes (millions of gigabytes) of archived remote sensing data — available gratis from U.S. government agencies like NASA, NOAA and the U.S. Geological Survey, and from the European Space Agency (via its Copernicus Data Access Policy) — can feed existing platforms designed to expedite large-scale geospatial data processing. The latter have included the likes of TerraLib, Hadoop, GeoSpark and GeoMesa.
One major obstacle in taking full advantage of these resources remains many researchers' lack of basic IT management skills and resources, especially in the areas of data acquisition and storage; mastering obscure file formats; managing databases; machine allocation; networking; jobs and job queues, CPUs, GPUs, and using any of the vast array of geospatial DP frameworks now accessible.
Google and its team maintain that such an IT burden effectively puts such tool sets beyond the reach of many investigators and operational users. This limits access to the information found within many remotely-sensed data sets to remote-sensing experts who have previously arranged access to super-computing and other resources.
The answer? Earth Engine, a cloud-based and high-performance supercomputing center. Can doing this work really be as easy as creating your own website has become? That seems to be Google's MO. Once a remote-sensing programming routine resides on Earth Engine, a user needn't be an application developer, web programmer or HTML coding king to produce well-organized data products or interactive applications.
As a final righteous bonus, the Earth Engine is designed to let researchers easily disseminate results to other functionaries, policymakers, field workers, NGOs and — perhaps most angelically and magnanimously of all — the general public.
- EPEE: Cooling has an essential role to play
- Back to the future with Ford bioplastics
- US vs. Europe: Comparing different approaches to renewable energy
- Can solar energy compete with fossil fuels?
- Impressive new smartphone apps in health and medicine
- Emerging green building material technologies to watch
- The environmental benefits of LED lighting
- Window film improves building system performance
- AR-15 ‘overpenetration’ only exists if you do this wrong
- More big brands join the war on plastic
- Why I started reading storybooks to my older students
- Is there a market solution to the feral hog problem?
- A long road ahead for Heathrow expansion, despite optimism
See your work in future editions
Your content, Your Expertise,
Your Industry Needs YOUR Expert Voice & We've got the platform you needFind Out How