Day 16/100: Higher bars mean more people 📊
Early #ThreeJS experiment visualizing world cities as vertical bars where height represents population. Tokyo towers, Mumbai rises high, smaller cities barely register.
Technique: Extrusion - taking 2D map points and adding third dimension driven by population data. Density becomes physical landscape.
Challenge: Occlusion hides smaller cities behind larger ones. Added rotation controls for multi-angle exploration.
Day 15/100: Mapping education access in Sierra Leone 🎓
Client project (2021) for Fab Inc: How far are children from schools?
Approach: #MapboxGL + Isochrone API to calculate realistic walking times (not straight-line distance). Generated 15/30/60-minute zones from each school location across actual terrain.
When mapping serves education equity, every pixel represents a child's opportunity to learn. ✨
Project details: https://maptheclouds.com/work/sl-isochrones/
Week 2: From GIS foundations to modern tools 🛠️
Days 8-14 traced 7 years evolution: satellite settlement mapping → real-time #WebGL terrain. Through #Copernicus data, #H3 indexing, electoral cartograms, 8 map projections (same data, different truths), Aiudului 3D.
Core questions stayed consistent: How represent density? Make terrain understandable? Show political complexity readably?
Tools evolve. Cartographic problems endure 📐
Week 3: COVID-19 meets D3 simulations.
Day 14/100: Ten years later, same place, different tools 🔄
Aiudului Gorges (Trascău Mountains): my reference mapping location since 2011.
2011: #OpenLayers + MapTiler. 2D tiles over OSM.
2021: #ProceduralGL. Smooth 3D terrain with #WebGL + real elevation. Routes draped over high-res topography.
Core question unchanged: how help people understand terrain before arrival?Problems consistent. Solutions transform.
Day 13/100: The year I became MapTheClouds 🌍
2021 = transformative. #30DayMapChallenge + #30DayChartChallenge
Same year: launched https://maptheclouds.com, created brand/logo, refined COVID-19 network viz for Romania.
Lesson: discipline beats waiting for perfect conditions. Some days → proud work. Others → functional experiments. Both mattered. Showing up builds creative muscle memory.
End 2021: not just projects, but practice + platform + direction.
Day 12 of #100DayMapChallenge: Projected Time Zones 🌐
Same dataset - countries grouped by 6-hour time zones - rendered through 8 projections: Van der Grinten, Eckert III, Eckert V, Sinusoidal, Mollweide, Cassini, Equidistant Conic, Polyconic.
Each preserves something, distorts something else. Area, shape, distance, direction. You can never keep all four.
A projection is always a decision. Most maps hide it. This one makes it visible.
https://maptheclouds.com/playground/30-day-map-challenge/earthnotflat/
Day 11/100: When geography lies about votes 🗳️
Romania 2019 Presidential Elections dashboard. Obsessed with cartograms - maps where size = data, not geography.
Data: #QGIS + Spatialite joining ANCPI geometries + electoral data. 3,186 voting stations → municipality. Fixed București Siruta codes. #Svelte + #D3js dashboard.
Why: București tiny on map, massive in votes. Distortion becomes information.
https://blog.maptheclouds.com/events/for-whom-shall-i-vote-2
Day 10/100: Why hexagons for population density 🔶
Romania at 400m resolution via Uber #H3 hierarchical hexagonal index. Admin boundaries rarely reflect how people occupy space.
Hexagons > squares/triangles: 6 equal neighbors, minimal directional bias, reduced edge effects, circular distance approximation, perfect tessellation.
#QGIS for H3 processing + #Qgis2threejs for 3D export = browser scene with semitransparent hexagonal columns. Height + color encode density.
Day 9/100: A ball that rolls uphill… or does it? 🎱👀
Gravity Hill, Pittsburgh—terrain optical illusion. Downhill slope appears uphill. Built web map with #MapLibreGL + #MapTiler showing location + video demo of phenomenon.
Experimented with animation: marker rolls onto screen on page load. Small detail, reinforces playful nature.
Reminder: spatial perception isn't just data/measurement—it's context, expectation, sometimes wonder.
Your eyes vs. physics. 👁
️
#100DayMapChallenge
Day 8/100: Mapping where people live from space 🛰️
Cluj-Napoca settlement analysis via World Settlement Footprint 2019. #Copernicus Sentinel-1/2 data at 10m resolution, processed by ESA, DLR, GEE.
The power: petabytes processed + openly available. Anyone can study settlement patterns anywhere on Earth through Copernicus program.
Open satellite analysis transforms what's possible for independent researchers.
Interactive demo: https://maptheclouds.com/playground/30-day-map-challenge/urbanrural/
Week 1: The long game 🗺️
7 projects, 2011-2022. OpenLayers → #LiDAR → Landsat → #ThreeJS Antarctica.
Tools evolved dramatically. #WebGL existed (Mapbox) but wasn't customizable. #PDAL hadn't opened LiDAR beyond enterprise. Shaders weren't on radar. But question stayed: how make terrain feel real online?
Answers accumulated slowly, project by project.
Week 2: GIS foundations → interactive data viz. Acceleration begins ⚡
#100DayMapChallenge Days 1-7
Day 7/100: Mapping a restless volcano 🌋
Taal Volcano, Philippines - one of the world's most active volcanic systems. Caldera inside a lake formed by earlier eruptions. #PDAL #Entwine #Potree workflow on 172M points from PHIL-LiDAR Program.
Volcanic LiDAR as temporal document: spatial data is never static. Terrain evolves.
Context isn't recreation - it's about understanding landscape that may change without warning.
Week 1 complete! GIS foundations → D3.js phase next.
Day 6/100: Antarctica doesn't look like this (but what if it did?) ❄️☁️
200m resolution elevation data from RAMP rendered in #ThreeJS with a twist: fluffy clouds instead of ice and rock.
Pure visual experimentation.
Workflow: #QGIS → export as image → Three.js fly navigation + transparent texture
Sometimes maps don't need realism. Sometimes looking at data in an unfamiliar way reveals patterns you wouldn't notice otherwise.
Day 5: Point clouds cross borders 🌍
In 2015, NYU captured one of the densest public #LiDAR datasets: Dublin at sub-meter resolution. Years later, I processed 500M+ points using #PDAL and #Potree, applying the same pipeline from Yosemite.
NYC captures. Romania processes. Anyone explores.
Geography is local. Tools are global.
https://blog.maptheclouds.com/learning/lidar-pdal-experiments-dublin
Day 3/100: 25M points in one file. Where to start?
FOSS4G 2019: Connor Manning & Adam Steer demo'd PDAL/Entwine. Hooked instantly.
Sequence matters: classify, filter, mesh, interpolate.
Filter too early → lose detail. Too late → noise corrupts.
570M points. 19h processing. Open-source JSON pipelines. All documented. Visualization: seconds. Processing: 19 hours.
https://blog.maptheclouds.com/learning/lidar-pdal-experiments-yosemite-valley
Day 4/100: Maps for decisions 🔥
Aug 2019: 1M acres burned across Bolivia-Paraguay-Brazil.
Landsat 8 dNBR: pre/mid/post-fire scenes, TOA correction, NBR differencing, classification.
988,069 acres total. Max continuous: 638,673 acres (~491K football fields).
ArcGIS ModelBuilder + Python automation. Four severity classes for recovery planning.
Not every map needs beauty. Some need accuracy + speed.
#GIS #Landsat8 #RemoteSensing #FireMapping #BurnSeverity #100DayMapChallenge
Day 2/100: Mapping walls you can't see from above 🧗♂️
Yosemite. LiDAR 0.5m DTM. Extracted walls >75°. Flow algorithms estimated heights.
Classified: yellow to red (45°/100m → 75°/900m).
Challenge: LiDAR can't see overhangs. What climbers care about most. 🏔️
Map shows possibility, not certainty.
Some questions need both data and dirt.
https://blog.maptheclouds.com/learning/yosemite-climbing-walls
Day 1/100: The origin 🗺️
Aiudului Gorges, Romania. 2011. Karst terrain I still hike and climb.
Hand-digitized contours from topographic map. Same one from my thesis. Tiles with GDAL2Tiles + MapTiler, rendered in OpenLayers.
Before Three.js in cartography. Before WebGL was mainstream.
Same question: how do we make terrain feel real online? ⛰️
Still asking, 14 years later.
#WebMapping #GIS #OpenLayers #GDAL #Cartography #100DayMapChallenge
End of an era. Start of a new one.
10+ years building maps: LiDAR → GIS → WebGL → network graphs → spatial storytelling.
Something’s shifting. Drawing me toward immersive work that blurs data, art, and experience.
Starting Feb 9: #100DayMapChallenge
100 days. 100 projects. Tracing my transition from traditional GIS to interactive 3D viz.
From static maps to immersive environments.
From QGIS to WebGL. 🗺️
