Quantification of a continuous-cover forest in Sweden using remote sensing techniques
(2015) In Student thesis series INES NGEM01 20142Dept of Physical Geography and Ecosystem Science
- Abstract
- Mapping and quantifying forest information about e.g. land cover, tree height and biomass has traditionally been a both time-consuming and labour-intensive part of forestry and forest research as field measurements typically are collected manually using handheld equipment. Remote sensing has proved to be a valuable complement to field based measurements as it enables for fast and relatively cheap collection of data from areas that would be hard to access from the ground. The aim of this thesis was to map and quantify the Romperöd forest outside Glimåkra in southern Sweden where selective thinning forestry has been practised since the 1960’s. The study was carried out using high resolution multispectral aerial images and small-footprint... (More)
- Mapping and quantifying forest information about e.g. land cover, tree height and biomass has traditionally been a both time-consuming and labour-intensive part of forestry and forest research as field measurements typically are collected manually using handheld equipment. Remote sensing has proved to be a valuable complement to field based measurements as it enables for fast and relatively cheap collection of data from areas that would be hard to access from the ground. The aim of this thesis was to map and quantify the Romperöd forest outside Glimåkra in southern Sweden where selective thinning forestry has been practised since the 1960’s. The study was carried out using high resolution multispectral aerial images and small-footprint discrete-return LiDAR data included in the Swedish national elevation model in conjunction with field measurements. The results revealed a mixed forest where Norway spruce was the most dominating tree species, accounting for 40.2 % of the total coverage of the study area, followed by Scots pine (13.8 %), broadleaved trees (8.7 %), succession (6.7 %) and bare-ground (4.1 %). The elevation of the terrain varies between 76.2 and 107.3 meters above sea level, with a ridge extending from south to north. The canopy height of the forest varies greatly throughout the study area and ranged between 1.0 and 34.6 m with an average height of 15.1 m and a standard deviation of 8 m. Above-ground biomass (AGB) was estimated by fitting a multiple regression model to LiDAR-derived vegetation metrics (independent variables) and AGB estimates based on field measurements (dependent variable). The model managed to explain 70 % of the variability in the field measured AGB estimates and was applied to the entire study area yielding an average AGB of 122 900 kg/ha and a standard deviation of 50 497 kg/ha. The inclusion of remote sensing data improved the AGB estimates compared to those based solely on field measurements. The results were compared to the AGB data included in the SLU Forest Map which showed low correlation with AGB estimates based on field measurements (adjusted R2: 0.14), proving it unsuitable for the part of the Romperöd forest characterized by selective thinning. (Less)
- Popular Abstract (Swedish)
- Att karlägga och kvantifiera skogsinformation angående exempelvis marktäcke, terräng, trädhöjder och volym är en traditionellt både tidskrävande och dyr del av skogsbruk och forskning eftersom mätningar vanligtvis samlas in i fält med handhållna instrument. Fjärranalys har visat sig vara ett värdefullt komplement till fältbaserade mätningar eftersom det möjliggör för snabb och relativt billig insamling av data från områden som skulle vara svåra att besöka i fält. Syftet med denna uppsats var att kartlägga och kvantifiera Romperödskogen utanför Glimåkra i nordöstra Skåne där blädningsskogsbruk har praktiserats sedan 1960-talet. Studien genomfördes med hjälp av fjärranalysdata i kombination med mätdata som samlats in i fält och blottlade en... (More)
- Att karlägga och kvantifiera skogsinformation angående exempelvis marktäcke, terräng, trädhöjder och volym är en traditionellt både tidskrävande och dyr del av skogsbruk och forskning eftersom mätningar vanligtvis samlas in i fält med handhållna instrument. Fjärranalys har visat sig vara ett värdefullt komplement till fältbaserade mätningar eftersom det möjliggör för snabb och relativt billig insamling av data från områden som skulle vara svåra att besöka i fält. Syftet med denna uppsats var att kartlägga och kvantifiera Romperödskogen utanför Glimåkra i nordöstra Skåne där blädningsskogsbruk har praktiserats sedan 1960-talet. Studien genomfördes med hjälp av fjärranalysdata i kombination med mätdata som samlats in i fält och blottlade en blandskog där gran utgör det dominerande trädslaget (40,2 % av studieområdet), följt av tall (13,8 %), lövträd (8,7 %), föryngringar (6,7 %) och bar mark (4,1 %). Terrängen varierar från 76,2 till 107,3 meter över havet med en ås som sträcker sig från söder till norr. Höjden på krontaket är heterogent i hela studieområdet och varierar mellan 1,0 och 34,6 m med en medelhöjd på 15,1 m och en standardavikelse på 8 m. Biomassa ovan jord uppskattades för hela studieområdet och visade ett genomsnitt på 122 900 kg/ha med en standardavikelse på 50 497 kg. Resultaten jämfördes med biomassa ovan jord enligt SLUs Skogskarta som visade låg överensstämmelse med skattningar baserade på fältmätningar vilket visar att SLU:s Skogskarta ej är applicerbar för den del av Romperödskogen som kännetecknas av blädning. Den föreslagna metodiken kan användas för att planera skogsbruk eller för att studera framtida förändringar eller störningar i skogen. Resultaten kan även vara till hjälp vid framtida forskning angående Romperödskogen och dess kolutbyte med atmosfären då biomassa ovan jord direkt kan konverteras till kolförråd, vilket är ett viktigt steg för att kunna studera effekten blädningsskogsbruk har på kolcykeln i skogen. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/8561391
- author
- Westin, Johan LU
- supervisor
- organization
- course
- NGEM01 20142
- year
- 2015
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- digital elevation model (DEM), land cover classification, selective thinning, Physical Geography and Ecosystem analysis, remote sensing, above-ground biomass, AGB
- publication/series
- Student thesis series INES
- report number
- 367
- language
- English
- id
- 8561391
- date added to LUP
- 2016-01-21 08:48:02
- date last changed
- 2016-01-21 08:48:02
@misc{8561391,
abstract = {{Mapping and quantifying forest information about e.g. land cover, tree height and biomass has traditionally been a both time-consuming and labour-intensive part of forestry and forest research as field measurements typically are collected manually using handheld equipment. Remote sensing has proved to be a valuable complement to field based measurements as it enables for fast and relatively cheap collection of data from areas that would be hard to access from the ground. The aim of this thesis was to map and quantify the Romperöd forest outside Glimåkra in southern Sweden where selective thinning forestry has been practised since the 1960’s. The study was carried out using high resolution multispectral aerial images and small-footprint discrete-return LiDAR data included in the Swedish national elevation model in conjunction with field measurements. The results revealed a mixed forest where Norway spruce was the most dominating tree species, accounting for 40.2 % of the total coverage of the study area, followed by Scots pine (13.8 %), broadleaved trees (8.7 %), succession (6.7 %) and bare-ground (4.1 %). The elevation of the terrain varies between 76.2 and 107.3 meters above sea level, with a ridge extending from south to north. The canopy height of the forest varies greatly throughout the study area and ranged between 1.0 and 34.6 m with an average height of 15.1 m and a standard deviation of 8 m. Above-ground biomass (AGB) was estimated by fitting a multiple regression model to LiDAR-derived vegetation metrics (independent variables) and AGB estimates based on field measurements (dependent variable). The model managed to explain 70 % of the variability in the field measured AGB estimates and was applied to the entire study area yielding an average AGB of 122 900 kg/ha and a standard deviation of 50 497 kg/ha. The inclusion of remote sensing data improved the AGB estimates compared to those based solely on field measurements. The results were compared to the AGB data included in the SLU Forest Map which showed low correlation with AGB estimates based on field measurements (adjusted R2: 0.14), proving it unsuitable for the part of the Romperöd forest characterized by selective thinning.}},
author = {{Westin, Johan}},
language = {{eng}},
note = {{Student Paper}},
series = {{Student thesis series INES}},
title = {{Quantification of a continuous-cover forest in Sweden using remote sensing techniques}},
year = {{2015}},
}