Spatial Ox

Spatial Ox GIS, WebGIS, RemoteSensing, GIS Dashboard, Environmental Monitoring, Geospatial Intelligence, Satellite Imagery, Drone Mapping

Your water network loses 35% of its supply. Where is it going?A loss percentage tells you how big the problem is, not wh...
02/11/2026

Your water network loses 35% of its supply. Where is it going?

A loss percentage tells you how big the problem is, not where to dig. GIS can narrow it to streets using district metered areas, 3 a.m. night flow and pressure data, but only when the valve records are right.

In one UK zone survey (2007, an illustration rather than a benchmark), 487 of 3,433 surveyed valves were missing from the GIS.

A simple first step: pick one DMA, find its boundary valves and watch how the numbers move.

Full breakdown: https://spatialox.com/blogs/industry-specific-use-cases/your-water-network-loses-35-of-its-supply-gis-can-find-where-if-the-valve-records-are-right?utm_source=facebook&utm_medium=social&utm_campaign=water-loss-valve-records&utm_content=carousel

Need your network data cleaned up? Our GIS team can help: https://spatialox.com/pages/gis?utm_source=facebook&utm_medium=social&utm_campaign=water-loss-valve-records&utm_content=carousel

Deed: 10 acres. GIS: 9.63 acres. Why does GIS parcel acreage never match the deed?Because one parcel carries five differ...
31/10/2026

Deed: 10 acres. GIS: 9.63 acres. Why does GIS parcel acreage never match the deed?

Because one parcel carries five different area figures: the record area, the survey-computed area, the GIS planar area, the GIS geodesic area, and the stored attribute. Each is a valid answer to a different question.

In our own test, a 10-acre square measured in a local projection stayed within about 0.01 acre of the geodesic area, while the same square in Web Mercator read about 16.9 acres at Denver's latitude. A 3.7% gap on a 10-acre square works out to roughly a 1.9 m shift on each side. That is the number to compare against the boundary's positional accuracy.

Our parcel-area QA workflow shows how to reconcile the figures before they reach a tax roll, a valuation report, or a development application.

Full breakdown: https://spatialox.com/blogs/industry-specific-use-cases/same-parcel-five-different-areas-why-gis-acreage-never-matches-the-deed?utm_source=facebook&utm_medium=social&utm_campaign=parcel-area-mismatch&utm_content=carousel

Your LiDAR clearance report says the power line is clear. For one day.It records the conductor at the temperature and lo...
30/10/2026

Your LiDAR clearance report says the power line is clear. For one day.

It records the conductor at the temperature and load it had when the aircraft flew over, and the trees at that season's height. It can't tell you how the line sits at the end of a hot, heavily loaded afternoon, or what a fast-growing canopy does by next spring.

One more catch: a Texas co-op reported that over 95% of its vegetation-related outages came from trees outside the right-of-way. If the analysis only covers the strip beside the wires, it can miss the tree that actually brings the line down.

Our new GIS Beats post covers conductor sag, buffer width, point cloud classification errors, and a validation workflow you can run before a crew is sent out.

Full breakdown: https://spatialox.com/blogs/industry-specific-use-cases/your-lidar-map-says-the-power-line-is-clear-what-happens-when-the-trees-grow?utm_source=facebook&utm_medium=social&utm_campaign=lidar-clearance&utm_content=image

A satellite damage map can be sharp and still arrive too late.Resolution tells you what a sensor can see, not when the i...
29/10/2026

A satellite damage map can be sharp and still arrive too late.

Resolution tells you what a sensor can see, not when the image arrived. In the Tonga eruption, the first image came roughly a day and a half after the event. Clouds, revisit gaps, and mismatched before/after pairs add more delay and error, and radar only flags that something changed. It still needs checking.

Ask when each image was taken, when the map was published, and what was not observed.

Full breakdown: https://spatialox.com/blogs/industry-specific-use-cases/the-disaster-happened-today-why-is-your-satellite-damage-map-already-out-of-date-latency-in-emergency-mapping?utm_source=facebook&utm_medium=social&utm_campaign=emergency-mapping-latency&utm_content=carousel

Contains modified Copernicus Sentinel data.

An InSAR map showing 30 mm of ground movement is not proof that the ground moved. One phase unwrapping slip is about 28 ...
28/10/2026

An InSAR map showing 30 mm of ground movement is not proof that the ground moved. One phase unwrapping slip is about 28 mm, the size of your headline number.

Atmosphere, reference points, and viewing geometry can all fool the numbers, and line of sight is not the same as vertical settlement. The article covers how to check each one before anyone acts on it.

Full breakdown: https://spatialox.com/blogs/industry-specific-use-cases/the-ground-moved-30-mm-or-did-it-how-to-tell-real-insar-subsidence-from-processing-error?utm_source=facebook&utm_medium=social&utm_campaign=insar-subsidence-check&utm_content=carousel

Want an independent check on a deformation result? Talk to our remote sensing team: https://spatialox.com/pages/remote-sensing

Flying is solved. Proving your drone mapping data is correct is the hard part.A clean-looking model is not the same as a...
27/10/2026

Flying is solved. Proving your drone mapping data is correct is the hard part.

A clean-looking model is not the same as an accurate one. Checkpoints, surveyed like control points but left out of processing, are the independent test. Warped edges, stair-stepped surfaces, and points scattered below the ground are early warning signs. And a software confidence score is not independent validation.

We cover how to build a QA/QC process that holds up when someone challenges it, including a real hydropower corridor survey of 837 images and more than 54.6 million points.

Full breakdown: https://spatialox.com/blogs/orthomosaic-and-photogrammetry-processing/from-collecting-data-to-trusting-it-why-qa-qc-is-the-real-bottleneck-in-drone-mapping?utm_source=facebook&utm_medium=social&utm_campaign=qaqc-drone-mapping&utm_content=carousel

A land cover map can be 92% accurate and still get your most important class wrong.If wetland is only 4% of a district, ...
26/10/2026

A land cover map can be 92% accurate and still get your most important class wrong.

If wetland is only 4% of a district, missing all of it costs just 4 points of overall accuracy. The JRC's own 2020 Global Forest Cover map shows the same effect: 91% overall, but 18% of the area mapped as forest wasn't forest and 8% of real forest was missed.

For land cover accuracy assessment, check user's and producer's accuracy for the class that drives your decision, using independent validation points.

Full breakdown: https://spatialox.com/blogs/industry-specific-use-cases/your-land-cover-map-is-92-accurate-what-if-it-gets-the-most-important-class-wrong?utm_source=facebook&utm_medium=social&utm_campaign=land-cover-accuracy&utm_content=carousel

Photos: jarmoluk and Efraimstochter via Pixabay.

GIS and BIM models can look perfectly aligned in a viewer and still clash on site.Both teams can be right. A GIS layer s...
24/10/2026

GIS and BIM models can look perfectly aligned in a viewer and still clash on site.

Both teams can be right. A GIS layer sits in a projected coordinate system with a defined datum. A BIM model usually sits in a local grid with its own origin, units and idea of north. What joins them is a transformation, and that is often missing, undocumented or wrong.

Nudging the model onto the basemap by eye fixes one spot, not scale or rotation. A 0.1° rotation error moves a point about 1.7 m at 1 km.

The post also covers datum and unit traps, vertical offsets, a six-point handover checklist and a table for finding where a mismatch started.

Full breakdown: https://spatialox.com/blogs/data-formats-and-deliverables/your-gis-and-bim-models-align-on-screen-why-dont-they-align-on-site?utm_source=facebook&utm_medium=social&utm_campaign=gis-bim-alignment&utm_content=image

Your AI mapped 10,000 buildings overnight. That count tells you how many things it detected, not how many are real.Two n...
23/10/2026

Your AI mapped 10,000 buildings overnight. That count tells you how many things it detected, not how many are real.

Two numbers decide whether an AI-extracted layer is usable: precision (of what it drew, how much exists) and recall (of what exists, how much it drew). One overall accuracy figure hides both, because false detections and misses can cancel out in the total.

The article walks through the errors you will actually see (phantom buildings, merged rooftops, missed sheds, roads that look right but don't connect) and a sampling workflow to run before anything is published.

Full breakdown: https://spatialox.com/blogs/industry-specific-use-cases/your-ai-mapped-10-000-buildings-how-many-are-actually-right-validating-ai-feature-extraction?utm_source=facebook&utm_medium=social&utm_campaign=ai-validation&utm_content=image

Nullah Leh reached a reported 29 ft at Kattarian Bridge on 1 September 2026. The spill level is 30 ft, so the margin was...
22/10/2026

Nullah Leh reached a reported 29 ft at Kattarian Bridge on 1 September 2026. The spill level is 30 ft, so the margin was about a foot.

The gauge cannot tell you where the water goes once it leaves the channel, or which fix gives the most relief. A flood simulation can. We ran a sample one in ArcGIS Pro for the Nullah Leh corridor in Rawalpindi: where water gathers, how to test a wider choke point, and where a quick model stops being reliable.

It is a sample run, not a prediction, and it does not replace an engineering model.

Full breakdown: https://spatialox.com/blogs/industry-specific-use-cases/nullah-leh-reached-29-feet-one-foot-short-of-spilling-what-a-flood-simulation-can-tell-rawalpindis-planners?utm_source=facebook&utm_medium=social&utm_campaign=nullah-leh-flood-simulation&utm_content=carousel

Address

Office 205, 2nd Floor Chenab Plaza, Opposite Ufone Tower Blue Area Islamabad
Islamabad
44000

Opening Hours

Monday 08:00 - 18:00
Tuesday 08:00 - 18:00
Wednesday 08:00 - 18:00
Thursday 08:00 - 18:00
Friday 08:00 - 18:00
Saturday 08:00 - 18:00

Telephone

+923167333888

Alerts

Be the first to know and let us send you an email when Spatial Ox posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Contact The Business

Send a message to Spatial Ox:

Shortcuts

Share