GeoNadir

09/04/2026

Here's something we've noticed: when rehabilitation monitoring starts with "can we see the weeds," it often ends in the wrong place.

Not because seeing weeds isn't useful. It is - in the right context, at the right scale. But mine rehabilitation isn't precision agriculture. It operates at different spatial and temporal scales, with different regulatory objectives.

Confidence doesn't come from knowing exactly what's present in one place at one moment. It comes from understanding how systems are behaving through time. That's a harder question to deal with. And it requires rethinking what evidence actually defends your program.

We've been exploring this for a few months. Here's how we are changing our mindset:

In almost every conversation about remote sensing for mine rehabilitation, “seeing weeds” quickly becomes the focal point. "Can I see the weeds?" "Can I count the weeds?" "Which species are there?" These sound like reasonable questions. But rather than take them at face value, I see them as shor...

08/04/2026

Creating a slope layer from drone mapping data is not difficult. And a lot of workflows stop there. But the layer is only part of the story. Someone still has to interpret what it means, where the real risks are, and what should happen next. That interpretation step carries a lot of cognitive load, especially when the decisions need to stand up in a compliance context.

Recently, we've been spending a lot of time moving beyond that. Treating the layer as evidence, not the end point, and focusing instead on identifying where slope exceeds thresholds and which areas actually need attention.

The aim is to move from data to something you can act on, without needing to understand every step behind the scenes.

This is one of several workflows we're building behind the scenes at GeoNadir. They run on enterprise client data now. They'll ship to the broader user base later.

03/04/2026

Friday : These data are largely unusable.

In terrestrial drone mapping, flying close to midday is often recommended because the high sun angle reduces shadows and makes vegetation easier to interpret. It is a piece of advice that works well for many land-based surveys.

But marine environments behave very differently. When the sun is high, specular reflection from the water surface can produce intense sun glint (or glitter). Instead of seeing the seabed or benthic habitats, the sensor captures reflected sunlight. In nadir mapping workflows this effect is largely controlled by the sun–water–sensor geometry, which means flight direction or camera settings do very little to solve the problem.

Unless you have diffuse lighting from thick cloud cover, a high sun angle can render large parts of marine imagery unusable. It is a good reminder that drone mapping practices developed for land do not automatically transfer to water. The upshot - don't map near midday :)

02/04/2026

Tiny tip to enhance your quality of life :) One of the quiet challenges in working with spatial data is simply seeing enough of it at once. When your table of contents is constrained, you end up scrolling, collapsing layers, or constantly adjusting your view just to understand what's in your project.

Being able to resize the table of contents sounds like a minor interface change, but it gives you much more control over how you work. You can expand it when you need to explore your data structure, and minimise it when you want to focus on the map.

We think that good workflows are not just about powerful tools, but also about reducing the small interruptions that break your flow.

01/04/2026

Most coastal monitoring still starts with looking at images and asking “does this beach look different?” It is a useful starting point, but it only tells part of the story.

With drone mapping, the real value comes from moving beyond visual comparison and actually measuring change. That might be tracking shifts in elevation, extracting beach profiles, calculating sediment volumes, or quantifying how vegetation is responding over time. In many cases, the most important changes are not obvious in imagery alone.

We recently pulled together ten different ways to assess coastal erosion using drone data. What they all have in common is that they move from observation to evidence. Instead of saying a beach has changed, you can show how much, where, and in what way.

That shift is what makes the data useful for decision making, whether that is managing erosion risk, planning interventions, or understanding how a system is evolving.

Read on here:

Coastal environments are in constant flux, influenced by tides, storms, and human activity. Understanding these changes is critical for managing coastal erosion, assessing storm impacts, and informing conservation efforts. Drone mapping provides a powerful way to track these changes with high spatia...

31/03/2026

A lot of time in geospatial workflows is not just spent analysing data. Getting the data into the system in the first place can be a massive hurdle.

Uploading vector data in different projections, reprojecting it beforehand, clicking through import menus. None of it is difficult, but it adds friction to every project. Especially if you don't know the projection.

Being able to drag and drop vector files directly onto the map, regardless of their projection, removes a surprising amount of that friction. It turns what used to be a series of steps into something almost immediate.

Small improvements like this do not change what you can do. They change how quickly you can get to the point where the work actually starts.

Feel free to drag your GeoJSON, KML, CSV, GPX, and SHP files - in any projection - directly into your GeoNadir project!

30/03/2026

If you’ve ever worked on a project with multiple datasets, you’ll know how quickly the table of contents can become super messy and out of control.

Not because the data are complex, but because everything ends up sitting in one long list.

Being able to group layers, and then group those groups again, sounds like a small change. But it makes a big difference to how you organise and navigate your data. Instead of scrolling and searching, you can structure your project in a way that actually reflects how you think about it.

When you are working across multiple sites, time periods, or data types, that clarity becomes essential. It is not just about keeping things tidy. It is about being able to find what you need quickly and stay focused on the analysis rather than the interface.

27/03/2026

Friday . This image looks fine at first glance, but the shadows across the beach mean the data are not very useful for assessing erosion.

I chose to fly in the afternoon to catch the low tide, which is usually what you want for coastal mapping. But because the imagery was captured later in the day on an east-facing coastline, the trees cast long shadows across the shoreline. Those shadows completely obscure the areas where erosion features are most likely to be.

The result is that you cannot clearly see the beach profile, scarping, or subtle changes along the dune edge. In practice, that means the flight effort was largely wasted for the purpose I intended. It's also really hard to compare against flights on other days that I did in the morning when those shadows were not there.

Coastal mapping often involves trade-offs. On the east coast, morning light will usually give you better visibility along the shoreline, while on the west coast the opposite applies. At the same time, flying in the middle of the day is not always ideal over water because of sun glint.

So the “best time to fly” is not a fixed rule. It depends on what you are trying to observe, and getting that wrong can mean collecting data that looks good but doesn't answer your question. Or maybe it doesn't even look good :)

A is a failure turned into a learning

26/03/2026

Not all useful field data comes from aerial drone mapping. In many projects, ground photos provide critical context. They show species, conditions, or features that are not visible from above. But those images often end up sitting in folders, disconnected from the rest of the spatial data.

Adding photos as points changes that. Instead of treating ground photos as separate records, they become part of the map. You can see exactly where each image was taken, click into it, and use it to help interpret the surrounding data.

The value is not in the photo itself. It is in linking that photo to location, time, and the broader dataset so it can actually inform analysis and decision making.

25/03/2026

Choosing a drone for mapping isn't necessarily about picking the newest or most expensive model. Last year we upgraded from a Phantom 4 Pro to the Matrice 4 Enterprise and spent some time testing it in the field. It is a very capable platform, with improved flight time, integrated RTK, and features that reduce the cognitive load on pilots when running mapping missions.

We tested a range of capture modes, including oblique and “smart” multi-angle options, and found that a standard nadir-only flight produced the best orthomosaic. It was faster to fly, generated fewer images, and produced no meaningful difference in output quality compared to the more complex capture modes.

This is something we see quite often. More data does not always mean better outcomes. In many cases it just increases processing time, storage requirements, and complexity without improving the result.

The same applies when choosing a drone. The key question is not which model has the most features, but whether it is designed for mapping in the first place. That includes things like mechanical shutters, mission planning capability, and positional accuracy.

If you are already using a Mavic 3 Enterprise, there is probably no strong reason to upgrade purely for orthomosaic outputs. If you are still using older platforms, the jump in capability is significant.

Getting this decision right upfront saves a lot of time later. The cost of the wrong drone is not just the purchase price, it is the cost of unusable or inefficient data collection.

Full review here:

When it comes to drone mapping and surveying, two of the top contenders from DJI’s lineup are the Mavic 3 Enterprise (M3E) and the Matrice 4 Enterprise (M4E).We decided to upgrade our drone and purchased a M4E and got into the field to test its mapping capabilities. We can honestly…

24/03/2026

Finding features in drone data is often more time consuming than collecting the data in the first place. If you're trying to map something like trees, weeds, or infrastructure, the traditional approach is to manually scan the imagery and digitise what you see. It works, but it is slow and difficult to do consistently across large areas.

Tools that allow you to search for features directly in the data change that workflow. Instead of manually hunting through imagery, you can identify targets of interest and extract them much more quickly. In this case, trees can be detected and mapped in a fraction of the time it would take to digitise them by hand.

Here we can benefit from speed and consistency. When applying the same method across sites and over time, the outputs become much more reliable for monitoring and comparison.