Detecting Structurally Old Scots Pine: A Crown-Metric Approach Using National ALS and LIFT Enrichment

Abstract

Old Scots pine (Pinus sylvestris L.) trees provide important habitat structures in boreal forest ecosystems, yet long-term intensive forest management in Sweden has made such trees increasingly rare and difficult to locate across large forest landscapes. Although field-based inventories can identify old trees reliably, they are time-consuming, costly, and spatially limited. This creates a need for scalable approaches that can support retention forestry and conservation planning over large areas.
This thesis develops and evaluates a method for identifying Scots pine trees with old-like structural characteristics using nationally available airborne laser scanning (ALS) data. Rather than focusing on biological age, the approach targets structural ageing. Structural ageing in Scots pine is commonly expressed through reduced height growth, lateral crown expansion, and increasingly flattened or irregular crown shapes. Individual tree crowns were segmented from canopy height models, and four crown-level structural metrics were derived: tree height, crown diameter, crown flatness, and recent height growth estimated from multi-temporal ALS data.
To account for variation in site conditions that influence tree growth and structure, trees were analysed within environmentally similar strata defined by elevation, soil moisture and peat depth. Within each stratum, an enrichment-based threshold analysis using the LIFT ratio was applied. LIFT was used to identify combinations of structural traits that occurred more frequently than expected under a null assumption of independence. Such enriched trait combinations were interpreted as indicators of old-like structural patterns.
The method was developed in the Idre–Särna forest landscape in northern Dalarna and evaluated in an independent landscape in the Lunsen–Kungshamn–Morga area south of Uppsala. Clear enrichment patterns were detected in a limited number of strata, primarily on dry to fresh mineral soils. Across both study areas, trees classified as structurally old-like represented well below one percent of all segmented crowns. A limited field validation showed high precision but moderate recall (0.37). This indicates that the method successfully identifies a small subset of trees with strong structural ageing signals but not managing to capture all old individuals.
Overall, the results demonstrate that national ALS data can be used to support the identification and prioritization of structurally mature Scots pine at the individual-tree level. The proposed method functions as a decision-support tool that can guide field surveys and complement existing forestry workflows.

Popular Abstract

Walk through a managed Swedish forest and you would find that most trees look roughly the same. Tall, straight, evenly spaced. What you rarely see are the true elders of the forest, the old Scots pine with wide, uneven crowns, thick bark, and dead branches shaped by years of wind, fire and slow growth. These trees are ecological hotspots, supporting species that younger forests simply cannot.
For foresters and conservation planners, this is a challenge. Old trees are important to retain, yet they are often scattered as single individuals across large areas. Locating them usually requires time-consuming field surveys, and covering entire landscapes on foot is rarely realistic. At the same time, detailed laser scanning data are available across Sweden and routinely used in forestry planning. The question is whether these data, that are usually used for other purposes, can also help reveal where the forest’s oldest-looking trees are hiding.
This thesis explores that possibility by focusing on tree structure rather than exact age. As Scots pine grow old, their height growth slows, crowns expand sideways, and treetops become flatter and more irregular. These visible changes can be captured by airborne laser scanning data. Using national datasets, individual tree crowns were identified and compared only with others growing under similar conditions, allowing trees with unusually wide, flat-topped crowns and slow growth to stand out.
The method was developed in a pine-rich forest landscape in northern Dalarna and tested in a second area south of Uppsala. In both regions, only a very small share of trees were identified as structurally old, reflecting the rarity of such trees in managed forests. When some of these trees were visited in the field, most showed clear signs of old age. However, the method did not capture every old tree present.
Rather than replacing field surveys, the approach works as a guide. It helps narrow down where to look, making surveys more focused and efficient. The study shows that data already collected for other purposes can be used in new ways to support conservation and retention forestry. In landscapes where old trees are few but vital, this offers a practical way to help ensure that the forest’s elders are not overlooked.

• author: Westman, Simon
• supervisor: Jonas Ardö
• organization: Dept of Physical Geography and Ecosystem Science
• course: GISM01 20252
• year: 2026
• type: H2 - Master's Degree (Two Years)
• subject: Earth and Environmental Sciences
• keywords: Geography, Geographical Information Systems, GIS, Airborne Laser Scanning, ALS, Scots pine, Structural ageing, Crown segmentation, LIFT analysis
• publication/series: Master Thesis in Geographical Information Science
• report number: 203
• language: English
• id: 9221779
• date added to LUP: 2026-02-03 09:26:25
• date last changed: 2026-02-03 09:26:25