Geodata, privacy and all that
For those who follow geo-news, Strava made the headlines for all the wrong reasons recently. Joseph Kerski’s Spatial Reserves blog covered it most comprehensively in Privacy concerns from fitness maps and apps, but to restate briefly: Strava helps fitness enthusiasts’ motivation by having them record, measure, compare and share their running, cycling etc. paths. They collected an astronomical amount of data and released them on beautiful maps last November citing their clever code as well as MapBox’s — The Global Heatmap, Now 6x Hotter — that ‘match made in heaven’ became a hell once it became clear what facilities were inadvertently revealed in global hotpots.
Unintended consequences
These are nothing new and I encountered a few in my many travels:
- Eosat the early operators of Landsat displayed in the mid-seventies the defoliation plumes in Siberia, which combined with NOAA wind data, helped pinpoint coal plants and nuclear facilities at the windward apex — this infuriated the Soviets so much that high-resolution satellite imagery as well as D-GPS were still not allowed when I visited W Siberia in 2002
- Radarsat in Canada provided an economical way to map the vast frozen expanses in the Arctic and the mobile icebergs offshore eastern Canada since the mid-80s. While it was great to track icebergs in shipping lanes outside the mouth of the Saint Lawrence River, or in oil platform moorings offshore Newfoundland, the Dept. of Defense was not amused when all buildings camouflaged in white appeared in glorious detail across the land
- TNRIS collected all the water-related facilities in Texas, which included dams for reservoirs throughout the state and the Intercoastal Waterway facilities along the Gulf of Mexico. They published on CD-ROM when I worked in Texas in the mid-90s, before the internet was what it is today… meaning that once they were shipped they were irretrievable! So it was too late then, when TRC managing the state infrastructure expressed concern about the public availability of such data, amidst increased terrorist activity like Unabomber and Al-Queida in other states around then
For more on the technical issues from that same period you can see my YouTube Tales from the Geodetic Crypt below:
Video is too fast? step thru it here!
The beauty of geodata
This should however not detract from the huge potential of freely released data. The banner picture shows 1/2M points of climate history spanning a century over a century ago, in the offshore and at a time where and when so little data exists. Esri’s Chief Scientist Dawn Wright speaks passionately about the 70% of the globe that has 5% of data coverage according to NOAA. That same agency released a phenomenal coastline database, which still exceeds the Esri standard V.2 world vector base map described in my blog & below.
There are also many, many open data sources, below is one I helped edit the map front end, and I use the open data bible by the same Joseph Kerski as well as Robin Wilson’s Free GIS Data. And you’ll surely come up with many more…
The control of geodata
So we now see that with free data does not come the freedom to do what we want. GDPR is a sort of Sarbanes-Oxley for data… and it’s happening now!
Niall Conway recently put together a comprehensive review of its importance in geo-data. He concludes with:
“ Although GDPR is a big challenge, it presents the geospatial community with an opportunity to reinvent itself and to build better relationships with, and products for, customers. By designing transparent and fl exible data management processes and by better understanding the digital footprint of data subjects, companies will be less likely to cross a ‘familiarity’ threshold which could potentially damage trust, reputation, and bottom lines. If not, just as in the aforementioned TV shows, justice will be served.”
An to help with that, I put together a quick exercise how to get started, from bog-standard mind maps to David Swann’s innovative LINQ framework:
The discipline of geodata
So let’s review what’s at hand:
- even anonymizing data cannot hide geo-locations that might be used in ways not anticipated, and this not just recently
- open data is such a wealth in the creative commons that one certainly must ensure its access and development on top
- yet utmost care must be taken in its sourcing and delivery, not only authentication but also through secure delivery
Perhaps the movement of geo data processing — especially big data and transactional or IoT — from the desktop to the internet and back to mobile devices is a natural one?
Not only can data scale better, but mil-spec security and multi-dimensional processing are far better secured in data centers accessed via web services. Remote sensing is a patent example, where an explosion of satellite and drone or LiDAR imagery have always been on-line but now are both faster and more affordable via the internet. Last but not least, the freemium model is becoming increasingly popular, making this affordable initially for small projects and proof-of-concept, then scaling cost commensurate to volume and functionality in full-blown implementations.
But as in the challenges opening this brief overview, the work is just beginning toward making smart geospatial communities, to steal AGI’s tagline from GeoComm 2017 three months ago: in closing, AGI Chair abigail page intoned “ Today, we have heard how the availability, consistency, and quality of geographic information is key. However, improvements to that data and the way it is used cannot come about easily without the influence and advocacy of people. We must come together and act now to influence that change…”
