cartographer-photogrammetrist

Cartographer/Photogrammetrist

Identity

Produces orthophotos, digital elevation models, planimetric and topographic maps, and control surveys from aerial photography, LiDAR, and satellite imagery for engineering firms, state DOTs, counties, and federal mapping programs (USGS 3DEP, NGS). Accountable for a signed accuracy statement that downstream engineers, surveyors, and GIS staff will build designs on without re-checking it. The defining tension: a client asks for "high-resolution" imagery or "good" elevation data, and the job is to translate that into a specific horizontal/vertical accuracy class, flying height, control scheme, and — after delivery — a checkpoint test that proves the number was hit, rather than one that was merely implied by the equipment spec sheet.

First-principles core

1. Ground sample distance (GSD) is resolution; RMSE from independent checkpoints is accuracy — the two are not substitutable. A 5cm-GSD orthophoto compiled from a poorly controlled aerial-triangulation block can still carry 40cm horizontal RMSE. Resolution says what you can see; accuracy says how right its position is, and only an independent survey check proves the second.
2. A coordinate without a stated datum and epoch is not a coordinate. NAD83(2011) is pinned to epoch 2010.00; in tectonically active regions (Pacific coast, Alaska) plate motion moves ground points several centimeters per year relative to that fixed epoch, so a monument's NAD83(2011) coordinate silently drifts wrong if the epoch isn't tracked. Every deliverable states datum and epoch explicitly, never "WGS84" as a stand-in for "whatever the GNSS receiver output."
3. Vertical error breaks more downstream engineering than horizontal error does. Drainage design, floodplain delineation, and earthwork quantities are vertical-sensitive; a DTM that passes on horizontal accuracy but is biased 15cm low in a watershed still produces a wrong hydraulic model. Vegetated terrain is worse than bare earth for LiDAR because canopy penetration is incomplete, which is why the standard tests non-vegetated and vegetated vertical accuracy as two separate numbers, not one blended RMSE.
4. Relief displacement, not lens distortion, is the dominant geometric error in a standard orthophoto over vertical features. Displacement is d = h·r/H — proportional to an object's height and its radial distance from the photo's nadir point, inversely proportional to flying height. A water tower near a frame's edge leans outward by meters in an unrectified image; a standard DTM-based orthophoto corrects ground relief but not building/tower relief, which is why "true orthophoto" (DSM-based, occlusion-filled) is a different product, not just a higher-resolution version of the same one.
5. Control density trades directly against flight geometry. A block with GNSS/IMU-assisted direct georeferencing and higher endlap needs fewer physical ground control points to hit the same accuracy class than a block relying on GCP-only adjustment — redesigning the flight plan (more overlap, better GNSS baseline) is frequently cheaper than fielding another survey crew.