---
  student: Matthew Heaton
  image: cover.gif
  tags:
  - students
  ---
  
iso 3D
matthew heaton | columbia gsapp m.s. cdp | colloquium ii: explore, explain, propose

REMOTE SENSING

*

to a geographer, the internet is a challenging object of inquiry . . . the web flattens the intuitive experience of space while its users and infrastructure still have a measurable position in the real world

this project attempts to parse this dissonance by asking:

how can we measure digital connectivity as it relates to a physical world?
where are the furthest corners of the internet?
where are the furthest corners from the internet?
how can we listen to the mechanisms that determine access?

I could conceive of connectivity like Barrett Lyon - his map of the internet presents a network topology of nodes and edges (or portals and linkages) that define its virtual quote-unquote “space.” But this type of spatiality is defined more through metaphor than felt experience. As a 'spatial scientist', I prefer to think about this much more literally:

BARRETT LYON: MAPPING THE INTERNET (2022)

FM RADIO TOWER LOCATIONS AND THEIR SIGNAL PROPAGATION CURVES

this is from a project I worked on over the summer in colloquium i -

Using US Federal Communications data, I extracted imagery from FM radio transmitters, as well as the estimated propagation distance of each transmitter signal. One can clearly see the relationship between the landscape and shape of each signal contour

I find that phenomenon of electro-physical interference compelling, so I’m digging in

propagation area extent
propagation contour
transmitter location: zoom level 15

point of entry

pictured : an old hunting camp built by my grandpa in the adirondack mountains

until recently, there has never been an internet connection anywhere nearby

cell service is still largely nonexistent within a couple miles’ drive, but we put a landline broadband hookup there a couple years ago

i have a complicated relationship with this newfound connection: i lost my ability (and excuse) to definitively sign off for a while, but it’s enabled more visits with flexible work arrangements . . . etc . . .

i still leave my phone in the car once i arrive

proof of concept

As an exploratory exercise, and to wireframe potential interface design for my final output, I’ve been developing a three dimensional mapping browser-based application to place built cell and radio topology in the context of the physical landscape.

New York State is my primary study area

Nearest-neighbor connections give a sense of cardinality for these line of sight dependent propagation signals

Lines of sight are fundamental to this modeling methodology

I’ve estimated potential cell phone dead zones to highlight areas of interest for later investigation

This was part proof of concept, part data explorer tool, mostly a learning exercise: figuring out how to reproject 2D geospatial data, which I’m used to working with, into a three dimensional space where signals propagate in all directions

click the 'fullscreen map' button, bottom right, to explore

MINIMUM SPANNING TREES: FM & CELL TOWER NETWORK, NYS

form factor

i make maps for a living, but they’re are about as conventional as they come, out of necessity: GIS input, output to PDF, print on a plotter, send to the field

for my capstone i’m eager to envision new methods of describing geographic information - specifically one that de-emphasizes the visual in favor of other sensory perspectives. ideally, too, my map-as-instrument would encourage discovery and play

Monome Norns Shield
monome norns shield, an open-source diy sound computer
Teensy Microcontroller
a teensy microcontroller driving a liquid crystal monochrome 'memory display'

I spend quite a bit of my free time tinkering with electronic instruments in my studio, often soldering my own, so incorporating that ethic into my cartographic practice is an exciting opportunity. This is also informed by my experience writing music first with virtual instruments and eventually being able to afford tactile interfaces. There is definitely a difference in the experience and output.

Likely the platform for this map will have a similar form factor to a GPS receiver (or etch-a-sketch) and will be raspberry pi or arduino based: I feel strongly about having a sort of “artifact” to hold and interact with

monome norns shield, an open-source diy sound computer
my disposable income

accessibility modelling

Here is a map of global travel time to the nearest high-density urban center, which is how this particular study operationalized remoteness Isochrone maps like these are commonplace: they estimate a person’s traversal time through a given geographic unit (cell) based on its characteristics

This typically includes ground cover, slope, presence and conditions of roads, vegetation density, and so on

I’d like to apply this same logic of gridded “friction surfacing” to the electronic domain.

By assigning weights to these different dimensions, I can run a sort of composite site “unsuitability” analysis within my study area, where each geographic study area (grid cells, townships, census tracts, or some other unit) will have a relative isolation score

central and remote

The US department of agriculture bins zip codes into four increasing intensities of “FAR” (frontier and remote, ha!) depending on their distance from high-order specialized (level 1) to low-order necessary (level 4) goods and services
this includes

    emergency services
    places of commerce
    transport hubs

usda 'frontier and remote' aka FAR places

signal erosion

I am applying the same logic that defines physical accessibility to explore electromagnetic signal as an electronic shadow. (Ironically I’m calling these literally analog waveforms “digital” in this case)

These signals use the earth as transmission substrate, either directly via lines of sight (their viewshed) or passively through the soil or ionosphere. The same physical phenomena that affect FM radio also affect cell signal.

Radio signal - and signal loss - is thus a useful proxy to operationalize and investigate electronic isolation

signal: sightlines

Drawing inspiration from my earlier terrestrial radio experiments, I’m most interested in signals that propagate through the air and how their signals decay or interfere with one another

Plus, there isn’t a ton of complete reporting done on cell dead zones, since the FCC allows telecomms to self-report and then uses those reports to make their coverage maps. Networks will model their network coverage based on perfect physical conditions on the ground

for reference, relative propagation distances per cell frequency band:

    2G (~50km)
    3G (~30km)
    4G (5-15km)
    5G (1-3km)

cellular antennae: manhattan, nyc
cellular antennae: blue mountain lake, adk

signal: landlines

Broadband access is of course fundamental to digital accessibility

Since this signal travels through hard wires, its transmission properties are much more easily monitored. Therefore, spatial coverage is well documented, like in this data explorer tool created by the new york state Public Service Commission

geocoded broadband data explorer
nys Public Service Commission 2023

signal: extraterrestrial lines

another dimension worth noting is connectivity via satellite

In addition to GPS and television constellations, EUTELSAT ONEWEB AND SPACEX STARLINK operate in the broadband domain

Their orbital positions are available via NORAD

SPACEX STARLINK
EUTELSAT ONEWEB

lines of sight

cell signals use the earth as transmission substrate, either directly via lines of sight (viewshed) or passively through the soil or ionosphere. The same physical phenomena that affect FM radio also affect cell signal.

Radio signal - and signal loss - is thus a useful proxy to operationalize and investigate electronic isolation

one method of roughly modelling megahertz level signal propagation is through a viewshed analysis. by placing a cell network in 3d space with a digital elevation model (DEM), one can estimate a receiver's signal on the ground

cell signals technically can penetrate solids, but there is immediate signal absorption and refraction

bounding network nodes in three dimensions using convex hulls
this technique can be used to visualize cellular network catchment areas

ground truth

my assumptions re: lines of sight as shorthand for cellular reception are corroborated by my own experience referenced earlier

triangles on this map are cell locations

darker areas are places out of view from any cell towers within 45 miles, the maximum potential propagation distance for 2G cellular signals under ideal conditions

true dead zones are likely more expansive

click the 'fullscreen map' button, bottom right, to explore this map in three dimensions

the cyan circle is the location of my family's adirondack camp
white areas are raster cells out of view from the nearest cell tower

the map as instrument

and measurement as performance technique

What comes next? When exploring ways to measure, simulate, and visualize communications infrastructure so far this year, my methods have been exclusively graphical

Next, I hope to develop a data sonification engine that maps audio parameters onto spatial ones

Here is a capture from a project where I built a “tunable” map that crossfades between recordings scraped from radio stations based at each node

sonification

since my phenomena of interest - sound + signal - are physically waves there is a surprising amount of semantic overlap

for example:

    Sonic and electromagnetic waves reflect and refract similarly, so sound propagation models may be helpful
    Amplitude modulation (AM, like the radio) and frequency modulation (FM) are critical to radio signal propagation as well as common methods to modify a sonic palette. In both contexts, there are fundamental “carrier” waves that are affected by an external “modulator” waveform
    Signal strength measured in decibels (just like loudness level on a stereo)
    Wave interference between competing bands produces eerie and unique audio fingerprints
    noise is decoded by signal receivers in similar ways - think about white noise static

AN ‘ÆSTHETIC PERSPECTIVE SPACE’ FOR CLASSIFYING AUDITORY DISPLAYS IN THE ARS MUSICA DOMAIN
(VICKERS & HOGG, 2006)
mapping potential sonic parameters onto physical attributes

sound, signal, terrain

“wavetables” are quasi-periodic sound-shapes that can be navigated through at high frequencies to produce distinctive sonic timbres - mountains of signal

a sonic map could reverse engineer the concept to decode physical landscapes through sound

precedents

plenty others have explored the similarities between musical composition, performance, and the science of transmission. There is a rich precedent in the “transmission arts,” but it typically refers to terrestrial radio transmission

i'd like to take it a step further and examine other modalities of transmission: cellular, wifi, bluetooth

alternative semiotics

the map as instrument is not a spatial chopping block

I hope to try and subvert the ingrained behavior of the map as a tool for creating and perpetuating arbitrary divisions in space

For instance, analog radio propagation signal contours don’t represent a harsh dropoff of signal, but rather the range without signal interference. I’m trying to think about sound, signal values, and space in terms of interfering wave patterns.

Compared to the continuity of signal waves, Digital signal receivers have a cardinality: the device is linked to one and only one signal.

Treating the map user as a signal receiver, my map will have a method of sonifying this switching mechanism which would be difficult to understand visually.



some things to think about:
    a place’s audio fingerprint vs arbitrary lines
    interfering signals vs. binary dropoffs aka continuous gradients
    + rhythmic oscillations vs discrete edges