Tag Archives: D3

Windswept Lands

The ready availability of huge datasets offers multiple formats for modeling an increasingly dynamic relation to space. Indeed, the processing of remotely sensed data allows for frighteningly rapid condensation and synthesis of what can be called information within a single map.  And the recent NOAA based on modeling weather conditions in a real-time wind map is particularly neat in how it invests the currents of air that swirl above land and sea with a hue of their own, as if to invest them with character as a subject of mapping and give them a new level of visibility.  Set against a night sky, where details such as topography or ocean currents are conspicuously silenced, the lines of winds appear in neon blue relief off the California coast, showing an elegance of serpentine motions are tracked with far more wonder than the static terrestrial map:


The flowing electric blue lines used to render wind move with a sleekness as if to promise immediate access to meteorological actuality that seem to blend the screen with current wind conditions, and provide that most fantastic of illusions of a map of the present-day that almost sends tingles up our neck in providing a map of events that seem actually to be transpiring.  The maps have in fact led reporters to gush thanks to the US government for its funding of the project with tax dollars, creating new conditions for the visibility of currents of air to which we are ineluctably drawn in an almost zen manner.  The GIF suggests a mapping that is based on looking at the overlooked, or rendering the smoothest of airborne sensations with cartographic legibility.  It echoes the recent ability to encode data from sensing stations in a transfixing image of swirling gusts that seems far more animated than the flat topographical base map that lies beneath.


The attraction of these maps is more visual, in fact, and somehow more seriously phenomenological, than having to do with geographical information, save as it suggests the increasingly fluid nature of all information flows.

The dynamic nature of wind maps renders an idealized perspective whose content asymptotically approaches a view that is almost entirely abstracted or removed from the inhabited world, based on the mathematical models for projecting and tracing the currents and intensity of wind over land:  indeed, in the marvelously idealized data-driven maps, wind currents create whirls of whorls over the lower forty-eight.  The website hint.fm uses real-time images of the inroads winds make across the territory of the United States, distinguishing gradations of gustiness by wisps of differing intensity to describe a space not only removed from the inhabited world, but whose elegant tracery enacts an aerial drama across in an artificially demarcated mapped space too vast to comprehend, but offers a sort of atmospheric ballet of the wind’s shifting directionality and intensity, as if traced by a multitude of individually situated geographically dispersed weather vanes oriented by the shifting winds across the lower forty eight.

Wind Map website image

If the image recalls something like a smoke-filled room, it is more of a wind-filled continent, where blowing winds cross borders, rivers, and plains in an image of the bounds of the contiguous forty-eight states.  The cartographers, or artists of technological visualization, Fernanda Viégas and Martin Wattenberg are dedicated to “work [that] explores the joys of revelation” and in converting local data to suggest natural currents that existed before highways carved the continent into itineraries–like the “blue roads” of rivers run across the land–but their constantly shifting trails mark a weirdly curved space in a situated perspective rarely able to have been captured before.  Even on a “mild day,” the winds give a sort of dramatic intensity to the organization of space oddly removed from the cartographer’s art, which is quite passive in the sense of offering a surface to be inscribed by the movement of the air, as wind curves around the nation’s great plains in currents of a recognizable patterns or flows that approximate stop-action photography.


Looping Denver

Sure, there are specific moments of meteorological intensity when winds line up in a recognizable pattern–a tornado or hurricane–in ways that define a geographic focus of interest or attention, as in the vortex of Hurricane Isaac as it approached New Orleans,


but all too often the dispersal of attention is the rule, as one watches the patterns that winds themselves generate.

This charting of currents is one of the most precious dreams of early map-makers:  to chart the very presence of winds across the land and sea, early symbolized by the rhumb lines that were so important for nautical navigation in early modern and medieval charts, where they communicated the basis of projecting travel across oceans otherwise rarely noted or known how to be represented in the surface of the map itself:  even the lines surfaces of the waves on which ships gently rock and sea monsters rear their heads is only a suggestion of the winds that run across their surface.

Olaus Magnus

The wind-map is abstracted from territoriality in a sense, as were the oceans:  its formal focus on the United States, derives from its synthesis of predictions of the National Digital Forecast DatabaseLaura Kurgan has suggested how much the big data of cartographical operations like Google Maps or Google Earth were not only enabled but indebted–and informed by–the declassification of satellite views during the Clinton administration, which created something of a boondoggle of available data for American mapping firms, if it also informed the very strategies of global surveillance that Google inherited:   the delicate wind-maps Viégas and Wattenberg conceived and coded suggest less of a governed land than a territory inhabited by air, in which the unfurling curling currents of wind estheticize the bounded nature of territory:  migrations of flows of air that course above the terrain, tracing inroads of whitespace into the blackened map of the lower forty-eight until, over a sufficient amount time, the entire region would be wiped white or erased by air currents, and present only where wind-velocity was nil, in an atmospheric vanishing act of a truly windswept land, where winds course across by crossing out space and inevitably white-washing inhabited lands.

Winds over US whitening out whitening #3

Drained Nation Whitening #6

In another sense, the wind maps register the messiness of local interaction, so absent from the idealized rendering of the map, even if that interaction and specificity lies above where human interaction occurs, and creates a network of a sort that is constantly reconfigured before our eyes as it is drawn and redrawn in trajectories with comet-like tails.

Modelling Wind Flows 12-58 EST-4

The modeling of wind gusts across the contiguous United States suggests a new sense of the spectatorship of the D3 map, although real-time of the map qualities maintain an illusion of an actual meteorological surveillance of the skies.  Just as we cannot watch the unfurling of wind-currents of different velocities over the space that is mapped, they define the dispersal of wind currents across space as if at a remove from actual habitation:  the fine lines that recall stop-action tracing of six levels of gustiness across the nation are extraneous to sites of habitation, because habitations are not what is being described, or included in the data that drives the map:  the conventions recall streets or waterways or rivers, but follow a set of mathematically modelled swirls and whorls that render air apparent over the land that we more usually map, in a view to currents far above the ground.  And yet the mesmerizing drama of modeling data of wind currents is so elegant and engaging that it is hard to stop watching: its ambitious organization charts a record of our world at a remove from the categories of mapping that we are more habituated to scan.  In a way, the windmap is the most elegantly estheticized of terrestrial or worldly maps, because it is also the most etherial.

USA Windswhirls May 1 12-58

What is the subject of the map is the flow of data that is so oddly anthropomorphic, casting the map as a hirsute surface and second skin, even as it describes the velocity of air currents above the land abstracted from any view of place?  Do the wind-currents create or seem to sculpt a sense of place in their whirls and eddies, giving a concrete palpability to the map’s surface that somehow runs against the flattened isotropic surface of a map?  Is it still a map, now that any sense of spatial indices are erased, and centers of habitation are dramatically reduced?  Or is it a screen?  One of the beauties of this projection is that as one zooms in to the ground, further cities appear in the map, providing the needed points of reference that would be too busy in the larger scale versions, where they would obscure the beauty and drama of the inroads made by the winds that circulate across the lands.

Shaggy USA

There is specific period eye that is attracted to the data visualization, familiar with reading data distributions rather than describing a topographic space in naturalistic terms of portrayal.  If the United States seems a shaggy dog, the most mesmerizing of D3 tricks in the book is to watch the ongoing expansion of wind currents across the Rocky Mountains, Desert, and California, from an imagined t-zero at which the calculations are assumed to begin, as winds begin to be inscribed, as if on an etch-a-sketch in reverse, and give contours and form to the blank black area that the first screen seems to map, suddenly giving it a texture that is specific to a date and time when the download of whether data began, until it gains something of a meteorological image of wind flows.

West Coast Winds 1

Califn Currents #3

Calif Currents #4

Calif Currents #5

Californai Currents #7

Is this sort of data visualization, as attractive as it is, a map?  The processing of the distribution of air currents is difficult to stop watching because of the elegance with which it makes us look at the winds that we notice but have such a hard time collectively grasping, and because of the currents, eddies, whirlpools and lines of different flow that the wind map increasingly reveals, so that it gains a coherence in itself that it initially lacked. The idea of mapping air is intensely appealing, because it acts, as a map, to make visible what is so easily overlooked or otherwise has no concrete identity by which to be grasped.  The parallel currents, if approximating a familiar natural appearance of a hirsute coat, are nonetheless quite conceptually difficult to envision in their relations to one another or totality.  The best reaction is one of wonder–wonder at being able to find a visual residue tracking the unrelated sensation of wind:  this is mapping, in a sense, as an exercise in synesthesia, where sight replaces the sensation of wind on skin.

There are elements of translating sense-based observations to a format of visual modelling in all maps, but the oddity of removing the array of wind-velocity from sense-peception of course seems odd because one rarely thinks of spatially locating winds with precision or fixity.  With closer examination of the trajectories of gustiness one notices, first, the odd pockets of calm across the land, oddly coincident with some cities, and probably not only because their buildings block the velocity or intensity of air flow, in what can appear like racing currents or overlain strands:

Pockets of Calm?

And then the country, as it is inevitably drawn by an invisible hand of the forces of the winds, cumulatively gains the contours that it originally lacked:

Modelling Wind Flows 12-59 ESTModelling 12-58 EST--3Modelling Wind Flows 12-58 EST-4

1 Comment

Filed under CityLab, comet-like tails, D3, Fernanda Viégas, Google Maps, hint.fm, Martin Wattenberg, National Digital Forecast Database, NOAA, wind-maps

D3 Maps the Filling-Up World

D3 images, or data-driven maps, shift with the content that they seek to display.  They in other words provide tools of spatial analysis unlike static maps or terrestrial projections:  for rather than reflecting or arranging data in a static form, by analogy to a window, such maps reveal the changing data that drives their design.  Such maps lend themselves in timely ways as tools able to chart a changeable terrain–or the shifting configuration of a distribution–which often are tied to data feeds, rather than employing a fixed dataset.

Since D3 maps are often able to take advantage of their mutable forms by being incorporating new data, using it to determine the rendering of a shifting actual space or set of inter-relations.  So the mapping growth of population in the world’s countries is a classic example of a D3 tools of visualization, determined by the growth of residents in each country:  although the base-map of the recent simulation of world birth/death rates by software developer Brad Lyon is a familiar grey projection of the continents, the constant data feed that shapes the map, locating individual births and death in specific places, and tabulating the greatest numbers of births in individual countries, is also a visualization of Tobler’s first Law of Geography:  that “Everything is related to everything else, but near things are more related to distant things.”  For the changes in global population are understood in the unit of the entire inhabited earth in this very catholic map, rather than as able to be confined to the borders of individual countries.

Among the benefits of using D3 visualizations to chart the rhythms and rates of births and deaths is a channelling of immense amount of information that is both a tremendous ramping up of cartographical omniscience that the viewer can often barely process or register in its totality–for the edges or contours of the map can be changing constantly, rather than the map being designed to reflect a stable dataset.  The popular new map of changing numbers of birth and deaths in the global population provides a neat visual example of using maps to geo-tagged births and deaths to part of a nation, offering, in either simple screen or beta-version simulation, something that would not be possible in a static or simple paper map.  In a visualization of the drumroll of births and deaths, the resulting map creates a record of the inhabited world that is “eerily omniscient” as James Hamblin aptly put it in The Atlantic, and God-like in its observation of births and deaths everywhere:  as the map’s odometer-like upper panel trucks past the 7 billion, 93 million mark, we are asked to face the challenges of comprehending on individual terms what truly approaches, if only asymptotically, something close to a God-like perspective on mankind.

By visualizing an approximate record of “world births and deaths in real-time,” in a map that extends to its literal meaning the notion of the “inhabited world,” the content of this map challenges not the borders of maps, as do so many D3 projections, but the viewer’s own boundary of self, or the meaning of this endless tabulation of the appearance and disappearance of lives on the planet and the regions of their greatest increase.  This single frame of the ongoing graphic portraying a record of world populations over the globe’s surface, in what might be called, with more than a wink, a truly existential map of the place of humanity in the world:


Visualizing Birth and Death world Wide at 5-27 Oct 30 2013


Never mind that the countries are only light grey outlines, and the subject of geography is secondary to the map’s shifting contents.  At issue is not how we inhabit the space, and not the centers of population or their distribution, but a statistical ping of each birth and a death that can be easily geo-located, blips of text that overwhelm and even detract attention from a grey base-map that seems to recede below each announcement, until they entirely bury the map, no doubt as our population will bury our world if its numbers proceed to grow by a further 44 percent by 2050.

There is no need to ask where these people will live or how they will be fed in this visualization:  we can wonder that all we like as the number of births proceed unimpeded, always far in the lead of the counting off of individually tragic deaths.  But in their collectivity, the amassing of these numbers is difficult to comprehend or register–and not only because of the relative objectivity of the simulation.   As one looks at the data visualization and the so rapidly increasing numbers of its statistical simulation, it starts to become evident that the perspective it offers is just not tenable for a viewer to start to process; its numbers perform an odd if awesome balancing act between concrete actuality and abstraction impossible to really comprehend, and we wonder where this data flow derives, as its ticker-tape pop-ups jerkily transport us to half-familiar or unfamiliar cities as it notches population growth, and the map is buried by textual banners–much as their viewer is buried by an unweildly surplus of information.

Software developer Brad Lyon used some ideas and concepts of Mike Bostock’s D3 maps in order to create a real-time register of births and deaths worldwide–or a simulation of them–that the viewer could gape at in wonder, and view almost as the objective correlative of the bulging population of the world.  As was recently reported in The Atlantic, this digitize simulation charts the contours of the inhabited world, expanding the use Lyons made of United States Census data in his popular simulation map, designed with Bill Sneebold, to create a  visual readable interactive map of the country’s population.

The overwhelming effect is immeasurably more dizzying than looking at a paper or static map.  But perhaps this also conveys some share of the immensity that must have overwhelmed classical viewers, unlikely to have ever been able to process information of far-off lands, in early images of the classical concept of the bounded ecumene.  The cartographical canvas is both familiar and  disorienting, if only because of the size at which every birth and death appears, generating banner toponyms rapid-fire progression that one can allow oneself to imagine reflects the actual swelling of population numbers across the world in real time.  The immensity with which the momentary geo-markings of this tally invests new claims in a pretty standard projection of the planisphere; highlighting specific places in what almost seems a random-generator makes its surface a field that is entirely disorienting in its surplus of detail:  as soon as one finds ‘news’ of a death here, in Khasakstan, there is a birth in Bolivia, and two more in Punje, so that numbers  Starting from the very moment that you open it up, and continuing until the screen becomes completely illegible as ticker tape birth and death announcements utterly obscure its surface–


10;24 to 10;53--Oct 30

or until you get in a hazy-headed stupor and realize that you’ve been watching the register doing its’ job for too long, and just have to leave the room since you can’t really keep up with all this–is an real-time image of the inhabited world, and a reminder we may have just have access to too much data.

The visual register in the screen image not only announces each and every birth and death with a punctual precision that is always eery, but keeps a tally of the totals and their distributions in each country, and they all cause one to be reminded relentlessly that births always outstrip deaths.

Visualizing Birth and Death world Wide at 5-27 Oct 30 2013


In ways that mapping the boundaries of mapping the world challenges the boundaries of self, the popping of up sites of individual birth and death on the simulation, which occurs below a population counter of the high velocity of our course of global overcrowding, the map is mind-blowing because it challenges the boundaries of what the viewer can hold in her or his head, as much as because of its terrestrial coverage–there is a sort of hortatory “look here!” “no, here!” “here goes again!”–that calls one’s eye back to the day-to-day nature of lived experience that is so often outside the horizon of maps, and rooting them in an ostensibly exact record of the transit of time:  from 5:10:26 p.m. to about 5:30 on October 30, births occurred at a rate of 4.2 per second, as 4,210 lives began, a large number (about a sixth) in India and a fourth in India and China, compared to 1,879 reported deaths during the same interval:  the explosion of population is captured nicely in that chunk of less than twenty minutes.



Visualizing Birth and Death world Wide at 5-27 Oct 30 2013


And indeed it seems, at first, that the number of births in China and India dominate the map overwhelmingly, as they also  form by far the greatest percentage of births world-wide, according to the data feed, as text appears that clutters the expanse of the terrestrial world, as if to provide a dizzying counter-map to the stability of terrestrial contours, that will eventually obliterate the map itself:


Births and Deaths in Asia?


But then one realizes that this is a conceit of the format of the map, or its display of geo-tags, which oddly cluster over the area of Eurasia as they are generated by incoming data, even if they are from cities in the midwest of the United States.

Of course, it’s not like you never knew that folks were regularly born and died in this world, even in places that you’d never been or will ever visit, but this almost “odometric  perspective” of individual lives clicking by, new ones popping up in places and others vanishing elsewhere, could either be a Buddhist envisioning of the ephemerality of being or a worrying sign of just how quickly things are crowding up.

Take a look at the beta version in Google Drive:


The post-modernity of this form of D3 mapping takes the ancient idea of the expanse of the ecumene, but turns it on its head:  the population takes over, swallows, and erases whatever qualitative content was on the map, as the base-map starts descends to irrelevance, and, illegible, is unable to offer us any bearings at all.  As much as offering a material subject, the D3 map unwittingly charts the ephemerality of the ecumene.

Leave a comment

Filed under data driven maps, data visualization, Google Drive, Interactive Maps, mapping population, mapping the populated world, Mike Bostock, terrestrial cartography