Satellite images, laser scans and digital maps increasingly shape how we understand forests, cities and coastlines. Yet we still lack understanding of how the data behind them are produced.
Remote sensing is everywhere.
Satellites continuously observe the Earth's surface. Aircraft map terrain with lasers. Smartphones carry sensors capable of collecting vast amounts of spatial information.
Together, these technologies produce the data that shape how landscapes are analysed, managed and designed.
For landscape architect Maximilian Schob, this development raises a fundamental question: What happens when a discipline comes to depend on sensing technologies and data infrastructures that largely operate outside its own field?
Through his PhD research, he argues that remote sensing and geospatial data have become deeply embedded in landscape architecture, while the sensing and modelling processes behind them operate as a hidden infrastructure that remains difficult to understand.
“Remote sensing already permeates many aspects of landscape architecture. At the same time, the conditions under which sensing and modelling take place are becoming increasingly opaque. We rely on these systems to understand and manage landscapes, but often have limited insight into how they actually work.”
More than data
Remote sensing is often associated with its outputs: images, maps, datasets and digital models. But Schob argues that focusing only on these products risks overlooking the much larger systems that produce them.
Behind every map lies a chain of targets, signals, sensors, data and interfaces. Together, these shape what becomes visible in a landscape and how it can be interpreted.
“We frequently encounter remote sensing through finished products such as maps, classifications or visualisations. My research suggests that we also need to pay attention to the processes that generate them.”
According to Schob, this challenge is becoming more significant as sensing technologies multiply and data infrastructures grow more complex. The result is a disconnect between those who produce remote sensing data and those who use them.
Missing framework
The starting point for the PhD is a paradox: landscape architects have become frequent users of remote sensing data. Yet landscape architecture lacks a robust conceptual framework for understanding how remote sensing operates and how its products relate to the landscapes they represent.
To address this gap, Schob develops a framework that describes remote sensing through five interconnected elements:
- Target: what is being sensed
- Signal: the radiation or other energy being measured
- Sensor: the device collecting information
- Data: the resulting digital record
- Interface: the means through which people engage with the information
The purpose is neither to simplify remote sensing nor to reduce it to technology. Instead, the framework is designed to make the process visible.
Schob describes this approach as inverse sensing. Rather than only using sensors to observe landscapes, he argues that landscape architects increasingly need to examine the sensors, models and data infrastructures themselves.
“Remote sensing is not just about data. It involves targets, signals, sensors, data and interfaces operating together. Understanding their relationships helps us understand how landscapes become visible, knowable and ultimately actionable.”
New way of thinking about models
Building on this framework, the dissertation develops what Schob calls the “model view” of remote sensing.
This is the central theoretical contribution of the research. In Schob's view, models are not merely representations of landscapes. They are active mediators between sensing infrastructures and the landscapes being sensed.
“Sensing and modelling are often treated as separate activities. My research argues that they are fundamentally intertwined. We sense through models, and we model through sensing.”
The goal is to provide landscape architects with conceptual tools for engaging more critically with increasingly complex sensing environments.
From forests to flood landscapes
The research tests these ideas through four case studies involving forests, flood landscapes and marine spatial planning. Together, they show how different combinations of sensors, data and models reveal different aspects of the same landscape.
“The case studies show that there is no single view of a landscape. Different sensing processes reveal different qualities, relationships and possibilities.”
Systems behind the data
One of Schob's broader arguments is that remote sensing has evolved from a specialist technology into a global sensing infrastructure. Sensors are now embedded in everything from smartphones and cars to drones and satellites, shaping how societies observe and manage landscapes.
The findings do not argue for less remote sensing. If anything, they argue for a deeper engagement with it.
According to Schob, landscape architects need to understand the processes of sensing and modelling rather than treating geospatial data as neutral inputs or finished products.
“The question is no longer whether landscape architecture should engage with remote sensing. It already does. The challenge is learning how to engage critically with the sensing infrastructure that increasingly determines what landscapes are, what can be known about them, and what can be done with them.”
As sensing technologies continue to expand across the planet, that challenge will gain in importance, not only for landscape architecture, but for anyone concerned with how landscapes are observed, managed and transformed.
Research highlights
- Remote sensing has become fundamental to landscape architecture.
- The sensing and modelling processes behind geospatial data are often difficult to understand.
- The dissertation develops a framework for remote sensing based on five elements: target, signal, sensor, data and interface.
- It introduces a “model view” that explains how models mediate between sensing technologies and landscapes.
- The research argues that landscape architects must engage with the sensing infrastructure itself, not only with the maps and data it produces.
- Interventions into models precede interventions into landscapes.
Maximilian Schob will defend his PhD thesis The Model View: Models for Remote Sensing in Landscape Architecture on 26th August at the Norwegian University of Life Sciences. See the event webpage for further details.