At 4:15 PM, he exported the map as a PDF and a GeoPackage, just in case. He hit and gave it a final name: "Final_Flood_Risk_2026.qgz."
First, he dragged in the base layers: a messy shapefile of the river basin and a satellite image from the QuickMapServices plugin. The satellite view was crisp, a patchwork of green fields and serpentine streams. He set the project CRS to EPSG:3857, the standard for web mapping, then quickly corrected it to a local projected system—EPSG:27700 for the UK’s Ordnance Survey. Accuracy was everything. qgis 3.22
Then came the trouble. The LiDAR .LAS file loaded, but the point cloud looked like angry confetti. He opened the —a vast library of algorithms that had saved his skin more times than he could count. He searched for "Noise filter." Nothing worked. The council wanted a clean Digital Elevation Model (DEM) by noon. At 4:15 PM, he exported the map as
Frustration set in. He reached for his third coffee. As he did, his elbow nudged the mouse, opening the . He stared at the noise, then an idea sparked. He navigated to Raster > Analysis > DEM (Terrain Models) . QGIS 3.22 whirred, its progress bar inching forward like a glacier. For ten minutes, Alistair paced the room, eyeing the clock. He set the project CRS to EPSG:3857, the
