Visualizing Every Los Angeles Building Since 1880: 3D Map Released

A 3D visualization site that brings up every surviving building in Los Angeles in the year it was built has been released. According to coverage by rustywasm on Hacker News (Best), the interactive map covers the period from 1880 to 2026, offering a bird’s-eye view of how the city took shape. On the published site, “lax-skyline.parcelscope.net,” each building is represented as a box that appears sequentially in the year it was constructed.

Every building in Los Angeles One box per building standing today, appearing in the year it was built. What was torn down along the way isn’t here, so this is the surviving city, not the city as it was.

As this statement indicates, only buildings that have survived to the present are shown. Demolished buildings are not included, and it does not fully recreate the city as it looked in the past. It is positioned as an attempt to visualize how the existing city has accumulated over time.

Time-Series Visualization of All Buildings in

Los Angeles

The released visualization represents buildings across Los Angeles as one box per building, making them appear along a time axis. In the initial state, only a handful of buildings that existed in 1880 are present, and as the years advance the boxes multiply, allowing viewers to grasp in three dimensions how the present-day skyline was formed.

At the top of the screen are filters such as “Decade built” and “Height.” Users can filter by decade of construction, floor count, and height, isolating only buildings erected in a particular era or comparing the distribution of low-rise and high-rise structures. A timeline in the center allows slider operation from 1880 to 2026 to check the state at any point in time.

Discussion on Hacker News drew 189 points and 92 comments, showing interest in urban-data visualization methods and historical context. It attracted attention not simply as a map, but for adding time as an axis.

Data Sources: LARIAC and the Assessment Roll

The visualization is built on two public datasets. According to the Hacker News (Best) introduction, building footprints rely on “LARIAC 2020 building outlines,” while construction years and attributes rely on the “LA County Assessor roll.” LARIAC, short for the Los Angeles Region Imagery Acquisition Consortium, is known as a joint project that provides high-accuracy geospatial data based on aerial surveys of the Los Angeles region.

The LA County Assessor roll is the tax assessment roll maintained by the Los Angeles County Assessor. It records the owner, use, year built, floor area, and other information for each parcel, and is widely used in urban research and real estate analysis. This visualization identifies the year of appearance for each building by matching polygon data showing building shapes with year-built information in the roll.

Building heights are estimated from floor counts and assessment data. The median is stated as 16 feet (about one story), and the tallest buildings can also be identified. When zoomed out, heights are exaggerated so that high-rise buildings remain easy to see in an overview. The explicit citation of data sources is important for ensuring reproducibility and verifiability.

Pan, Tilt Controls and Exaggerated-Height Design

The control scheme follows conventions common to map-based web applications. Drag to pan, and right-drag or drag while holding Ctrl to change tilt. By lowering the viewpoint, users can grasp differences in height three-dimensionally, not just planar distribution.

Exaggerating height when zoomed out is a device to improve legibility at urban scale. If shown at true scale, even high-rise buildings would look flat in a wide-area view. Exaggeration clarifies the contrast between the downtown high-rise cluster and low-rise suburban residential areas. When zoomed in, the exaggeration is reduced, allowing inspection at ratios closer to reality.

The screen also appears to offer display adjustments such as “Dim” and “Bright,” and toggles for attributes such as building “Footprint” and “Type.” If color-coding by use type and lot shape is possible, the formation of residential, commercial, and industrial areas can be visually separated. Such interactions encourage exploratory analysis not available from static maps.

In the context of similar geospatial visualizations, rendering based on WebGL and the use of vector tiles are common. Trends in GPU utilization, such as those covered in Vulkan Video Encoding Revived on Intel Alchemist GPUs, illustrate advances in the underlying technologies supporting large-scale 3D rendering in the browser. Smoothly rendering hundreds of thousands of buildings requires optimization of rendering load.

Bias and Limits of Showing Only Surviving

Buildings

A limit the site makes explicit is that it covers only buildings that still exist. Buildings that once existed but were demolished, and structures lost to land readjustment or redevelopment, are not shown. It therefore does not faithfully reproduce the city as it looked in a given year, but visualizes only the layers that remain when looking back from the present.

This characteristic creates interpretive bias. For example, if many buildings constructed in the 1920s were later lost to redevelopment, the display would understate building activity in that decade. Conversely, decades concentrated with durable reinforced-concrete high-rises could be overstated because of high survival rates. The site itself states “surviving city, not the city as it was,” warning of survivorship bias.

At the same time, this constraint offers clues for reading urban durability. How many buildings from each decade survive reflects construction quality, changes in use, and the impact of preservation policies. The fact that buildings from the 1880s are extremely scarce illustrates the renewal cycle of wooden construction and the history of urban expansion. Combined with decade filters, the city’s durability can be assessed from the perspective of survival.

Open-Data Visualization Trends and Context

In recent years, urban visualizations using government open data have been increasing. Works that combine building registries, aerial survey data, population statistics, and other sources for interactive exploration in the browser have been released by researchers and developer communities. This Los Angeles case stands out for its breadth of coverage and length of timeline.

Reactions on Hacker News include指摘 of data accuracy and gaps, alongside calls to extend the approach to other cities. With similar registries and building-footprint data being developed for New York, Chicago, Tokyo, and elsewhere, some see横 expansion of the method as technically feasible. However, because data formats and the accuracy of year-built records differ by municipality, simple transplantation is difficult.

Visualization techniques themselves are also evolving. As shown in Google Photos UI Refresh and Gemini Video Editing Features, intuitive timeline operation combined with AI assistance is advancing in the chronological organization of photos and videos. In urban-data visualization as well, intuitive controls such as year sliders and attribute filters help users without specialized knowledge understand the data.

Interest is also growing in agent-based infrastructure as a processing foundation for such large-scale data. As shown in Infinigence’s Strategy for Agentic Infra Selected by Four Companies, the importance of infrastructure that automates everything from preprocessing massive geospatial datasets to delivery is increasing. Even when individual developers handle city-scale data, efficient data conversion and tile generation in the cloud are key.

Reproducibility and Technical Takeaways for

Developers

Details of the specific implementation have not been disclosed in public information. However, the technique of extruding building-footprint polygons into 3D boxes and controlling their display based on year attributes can be realized with libraries such as Mapbox GL JS, deck.gl, and Three.js. Handling hundreds of thousands of polygons requires splitting into vector tiles, LOD (level of detail) control, asynchronous data loading with Web Workers, and similar measures.

On the data-processing side, the core step is spatially joining LARIAC building footprints with parcel data from the Assessor roll. Matching on addresses or parcel numbers requires handling notation inconsistencies and missing values. How to treat buildings with unknown construction years, and ambiguity in the definition of year built due to extensions and renovations, are also data-cleaning challenges. The fact that the published site displays attributes such as “Building Built Height Footprint Type Copied” can be seen as a consideration to allow individual verification of these matching results.

The takeaway for developers is that compelling visualizations can be created by combining public data to tell the story of a city. Without relying on expensive proprietary surveys, integrating existing assets such as LARIAC’s joint survey results and tax rolls can produce a persuasive visualization. In Japan as well, there is room to attempt similar time-series visualizations by combining data sources such as registry information from the Ministry of Justice, fundamental urban planning surveys from the Ministry of Land, Infrastructure, Transport and Tourism, and PLATEAU 3D city models.

In the future, layering not only year built but also information such as use changes, extension histories, and vacancy rates could develop into visualizations that capture urban metabolism more multifacetedly. At present it shows the emergence of surviving buildings, but integrating demolition histories and development-plan data could extend it to forecasting future skylines.

Editorial Opinion

In the short term, we expect designs that incorporate a time axis to be referenced in the developer community as an urban-data visualization method. The combination of a decade slider, height exaggeration, and pan-and-tilt operation is an effective pattern for intuitive exploration of massive building datasets, and could be reused in other cities’ cases and internal dashboards. We view the attention on Hacker News as indicating interest in such reuse.

In the long term, we believe digital-twin development of cities based on open data will contribute to participatory urban planning and disaster-preparedness studies. Visualization limited to surviving buildings makes the balance between preservation and renewal visible, providing material for debating redevelopment. If integration of 3D city models like PLATEAU with registry information over time progresses, a common foundation that anyone can verify urban history against will take shape.

The remaining question is how to correct for survivorship bias and incorporate the memory of the lost city into visualization. Records of demolished buildings are easily lost from registries, and restoration from aerial photographs and old maps takes enormous effort. Even if complete reproduction of the past is difficult, designs that explicitly show gaps as gaps would enhance data integrity.

References

• “Watch Los Angeles get built, one building at a time (1880–2026)”, by rustywasm — Hacker News (Best), 2026-09-07T18:52:06.000Z (ARR)
• Source URL: https://lax-skyline.parcelscope.net/