Hills Bending the Grid: When San Francisco Tried to Flatten Fifty Hills with Straight Lines

Look at a standard map of San Francisco and you will see a grid. A clean, confident, relentlessly orthogonal grid that marches from the Embarcadero to the Pacific Ocean with barely a wobble. It looks rational. It looks planned. It looks like the work of someone who understood the terrain they were organizing.

It is a lie.

San Francisco has more than fifty named hills. Telegraph Hill rises 284 feet above the waterfront. Twin Peaks tops out at 922 feet. Filbert Street, between Hyde and Leavenworth, achieves a grade of 31.5 percent – so steep that the sidewalk surrenders to stairs. And yet the map, viewed from directly above, shows none of this. The grid pretends the hills do not exist, because the man who drew it pretended the hills did not exist.

The O’Farrell Survey and the Tyranny of Paper

In 1847, an Irish-born civil engineer named Jasper O’Farrell was commissioned to survey and lay out the streets of the young settlement of Yerba Buena, soon to be renamed San Francisco. O’Farrell did what surveyors of his era did: he drew a grid. A handsome, regular, perpendicular grid with blocks of uniform size, oriented to the cardinal directions, with one diagonal slash – Market Street – cutting across the geometry at a 34.5-degree angle to connect the waterfront to the inland districts.

The problem was that O’Farrell drew this grid on flat paper, for a landscape that was anything but flat. The survey made no accommodation for the city’s extreme topography. Streets were platted straight up and over hills that a horse could barely climb. Blocks were dimensioned as if every lot sat on level ground, when in reality some lots tilted at angles that would later require retaining walls, stepped foundations, and the most creative garage engineering in America.

Richard Walker, in The Country in the City: The Greening of the San Francisco Bay Area (2007), describes the O’Farrell grid as an act of “willful geographic denial.” The grid was an assertion of real estate logic over physical reality – every block divisible, every lot saleable, every street a potential address regardless of whether a human being could reasonably walk up it. The hills were treated not as features to be respected but as obstacles to be overridden by the authority of the surveyor’s line.

What Happens When Geometry Meets Gravity

The consequences of this denial are the source of San Francisco’s visual identity. The cable car exists because the hills were too steep for horses pulling rail cars – Andrew Hallidie’s 1873 invention was a direct engineering response to a planning failure. The stairway streets – Filbert Steps, Lyon Street Steps, the Vulcan Stairway – exist because some “streets” on the O’Farrell grid turned out to be, in practice, cliffs. The city’s famous tilted Victorian houses, leaning into hillsides like passengers bracing on a listing ship, are the architectural consequence of building rectangular structures on surfaces that refuse to be rectangular.

From above, though, none of this registers. A satellite view of San Francisco’s Russian Hill neighborhood shows a perfect grid. The streets cross at right angles. The blocks are uniform. There is no visual evidence that some of those “streets” require four-wheel drive, or that the residents of certain blocks get their daily cardiovascular exercise simply by walking to the mailbox.

This is what makes topographic grid cities so fascinating as map subjects: the two-dimensional representation is a compression of three-dimensional drama. The map is technically accurate – those streets do exist, at those locations, with those orientations – but it omits the dimension that defines the lived experience. The grid looks calm. The city is vertiginous.

Lisbon, La Paz, and the Hillside Tradition

San Francisco is not alone in draping a grid over impossible terrain. Lisbon, rebuilt after the catastrophic earthquake and tsunami of 1755, imposed a rational Enlightenment grid – the Baixa Pombalina, designed by Eugenio dos Santos and Carlos Mardel – onto the flat land between its seven hills. But the surrounding neighborhoods – Alfama, Mouraria, Graca – retained their medieval tangle, and the result is a city where geometry and chaos exist in direct adjacency, separated by a change in elevation of a few dozen meters.

La Paz, Bolivia, sits in a canyon at 3,640 meters above sea level, its streets clinging to slopes so steep that the city eventually built a aerial cable car system – the Mi Teleferico – as public transit because buses could not reliably serve the topography. Seen from above, La Paz’s street network looks like a spider web that has been stretched and distorted by the walls of the canyon it occupies.

Valparaiso, Chile, is perhaps the purest example of a city where topography has completely defeated the grid. The flat port area (El Plan) maintains a recognizable street pattern; the surrounding cerros (hills) are served by a network of winding paths, stairways, and the famous ascensores – funicular elevators, some dating to the 1880s – that connect the waterfront to the hillside neighborhoods. The urban geographer Arturo Almandoz has written about Valparaiso as a city where the “vertical dimension is the primary axis of urban life,” a place where the conventional horizontal logic of the street grid has been rotated ninety degrees by the terrain.

The Third Dimension the Map Refuses to Show

Cartographers have tools for representing elevation – contour lines, hillshading, hypsometric tinting – but standard street maps almost never use them. The convention is to present the street network as a flat diagram, a planimetric view that records horizontal position and ignores vertical displacement entirely.

This convention is so entrenched that most people do not even recognize it as a choice. We accept flat street maps the way we accept that north is “up” – as a natural fact rather than a cartographic decision. But it is a decision, and it has consequences. It means that a map of San Francisco’s Nob Hill and a map of Kansas City’s West Bottoms look structurally similar, when in reality one is draped over a summit and the other sits in a floodplain. The grid equalizes what the terrain differentiates.

Contour mapping – the representation of elevation through isolines – reveals what the street map hides. When you overlay contour data onto San Francisco’s grid, the visual effect is startling: the calm orthogonal pattern is suddenly cross-hatched with elevation lines that bunch and spread like the grain of wood, revealing the hills that the grid pretends are not there. The intersection of these two systems – the human geometry of the grid and the natural geometry of the terrain – produces some of the most visually complex map compositions possible.

The Beautiful Lie of the Flat Map

At streetgrid.art, we render cities in plan view – the bird’s-eye perspective, the god’s-eye view. Our maps are deliberately flat. They present the street network as a two-dimensional pattern, a textile of lines and intersections, stripped of the third dimension.

This is, for a city like San Francisco, a kind of beautiful lie. The map shows you the grid as it was intended – as O’Farrell drew it on his flat sheet of paper in 1847 – rather than the grid as it is experienced, pitched and tilted and occasionally vertical. And there is something deeply compelling about that lie, because it reveals the ambition of the plan: the audacious, slightly insane conviction that straight lines could be imposed on a landscape that actively resists them.

When you hang a streetgrid.art print of San Francisco on your wall, you are looking at a document of human stubbornness. The hills are there, invisible, bending every street that crosses them into a roller coaster that the map refuses to acknowledge. The grid is a fiction stretched over a landscape that never agreed to it. And that tension – between the order we draw and the terrain we inhabit – is what makes the map so much more interesting than a simple diagram of streets. It is a portrait of the argument between geometry and gravity, and the flat map is geometry’s closing statement.