Forestry in Forest Remote Sensing

Forestry requires experts in various fields, but also people with broader knowledge who are able to understand the bigger picture. Sound knowledge is essential to ensure sustainable forestry. In the Nordic countries, forests were affected by the industrialisation that began nearly two hundred years ago. Forests were a very important raw material, used to build industries and to lay the foundations of today’s modern society.

The forest is an important part of renewable resources

About a hundred years ago, the first national forest inventory was carried out in Sweden. Even then, there was an awareness of how important it was to manage this valuable resource wisely. Today, it is also very clear that the biological diversity of the world’s forests is under threat. The global population has increased, and a larger proportion of people now enjoy a higher standard of living compared to just fifty years ago. The society of the future must be based on renewable resources, and in this context, forests play a crucial role. Across several scientific disciplines, we generate knowledge about how ecosystems function, but the question remains how this knowledge is applied in forestry.

We have multiple objectives concerning timber value, biodiversity, and social benefits. To achieve these, information about the entire forest is required at different scales and over different time horizons. Strategic planning in forestry employs methods to simulate forest management over long time horizons, for example, a 100-year period. This level of planning ensures that forestry is conducted sustainably and is therefore often applied simultaneously across large forest areas. Strategic planning then guides tactical and operational planning, which determines where and how various activities will be carried out over different time periods.

Remote Sensing in Forestry

To manage forests wisely, it is very important to have reliable information over large geographic areas. Collecting detailed information across extensive regions is costly, and therefore we often need to make trade-offs between spatial and temporal resolution. For example, satellites can be used to create maps of the world’s forests, but these maps do not have the same resolution as data collected using drones.

Remote sensing involves acquiring information about objects using sensors, without being in physical contact with them. This is done by detecting electromagnetic radiation across different wavelength bands, ranging from visible and invisible light to radio waves. Both the sun and artificial radiation sources can be used. Technological developments have made it possible to access sensors that can collect data with much higher resolution for objects at greater distances than was previously possible.

Before the introduction of digital remote sensing, aerial photographs were used in forestry. These photographs were taken with overlap, making it possible to manually measure tree heights.

The Use of Sensors as Data Sources

The first digital remote sensing used satellite images and ground reference data to create models for generating comprehensive maps. About 30 years ago, airborne LiDAR (light detection and ranging) was developed, allowing the automatic creation of 3D models over large forest areas. Digital photogrammetry was later developed, enabling the calculation of 3D data from stereo images using only sunlight as the illumination source. This technique can be applied at various scales, from cameras on drones to satellites. However, optical sensors are limited to favourable weather conditions; for instance, satellite images are unusable when clouds cover the forest being observed. In such cases, radar sensors can be used, which rely on an artificial radiation source. The resolution of radar images depends on the wavelength and the antenna used. The strength of the signal returning to the sensor depends on whether there are objects similar in size to the radar wavelength. To achieve higher resolution, a synthetic radar antenna can be created, for example by exploiting the movement of the aircraft carrying the sensor.

Remote sensing involves using sensors as data sources and then converting that data into useful information. Achieving this requires knowledge in several areas: sensor technology, physics, computer science, and statistics. Moreover, it is important to understand the specific problems for which the information is needed. We are currently experiencing rapid technological development. Sensors for forest remote sensing are becoming better, cheaper, and easier to use. We have seen a shift from large, expensive sensor systems that required a team of engineers to operate, to smaller, inexpensive systems that are simple to use. As a result, remote sensing will be applied by more people and across a wider range of applications. Using airborne optical sensors, we can measure tree canopies, although we are often also interested in the properties of tree stems.

From Manual Field Inventories to Automation – The Future of New Sensors

In the past, we have relied on manual field inventories to collect reference measurements for models, which are then used, for example, to estimate biomass or timber volume across large forest areas. In the future, new sensors will enable automated field inventories that can be used to gather reference data for a wider range of variables, such as those related to timber properties and biodiversity. There will be access to more data sources, which can be combined using highly accurate positioning, allowing us to optimise where and when data are collected. To identify patterns in the data and extract valuable information, we will leverage the ongoing revolution in artificial intelligence. This will open up entirely new opportunities for developing the forestry of the future.