Identifiering av gallringsbehov med hjälp av flygburen laserskanning
Thinning is one of the most important silvicultural activities in middle aged forests, partly to minimize damages and partly because it leads to earlier harvesting revenues and increases the value of the remaining trees. There are many different indices used to describe stand density and thereby the...
| Autor principal: | |
|---|---|
| Formato: | L3 |
| Lenguaje: | sueco Inglés |
| Publicado: |
SLU/Dept. of Forest Resource Management
2008
|
| Materias: |
| _version_ | 1855572024665047040 |
|---|---|
| author | Halvarsson, Joakim |
| author_browse | Halvarsson, Joakim |
| author_facet | Halvarsson, Joakim |
| author_sort | Halvarsson, Joakim |
| collection | Epsilon Archive for Student Projects |
| description | Thinning is one of the most important silvicultural activities in middle aged forests, partly to minimize damages and partly because it leads to earlier harvesting revenues and increases the value of the remaining trees. There are many different indices used to describe stand density and thereby the need of thinning in forests today. Studies have shown that forest variables can be estimated with high accuracy using airborne laser scanning and it is likely that the method could also be used to estimate forest density indices. In this study, the possibility of using the forest density index ΣH² to determine and map priorities for thinning operations for a forest area was examined. The objective was to test how different variables derived from airborne laser scanning data can be used to estimate the index ΣH². Tests were performed with low and high density laser data (densities of 0.8 and 10 measurements/m²). In field, ΣH² was measured on 30 field plots with a size of 30x30m. The field plots were allocated in forests with a variety in structure, height, age, and tree species mixture. Different thinning indices that correlated to field measured ΣH² were examined. These indices were computed directly from raw laser data, from raw laser data converted to a raster with different cell sizes, and using single tree detection. The results show that ΣH² describes the need of thinning in a relatively good way for forests with a variety in structure, height, age and tree species mixture even though there are some exceptions. The method to convert raw laser data to a raster with 6m cell size gave a thinning index with the highest correlation to ΣH². The method worked well for both low density (0.8 measurements/m2) and high density (10 measurements/m2) laser data with adjusted R² values of 89.7% and 90.8%, respectively. The relative RMSE was 14.3% and 13.5% for low and high resolution laser data, respectively. |
| format | L3 |
| id | RepoSLU12229 |
| institution | Swedish University of Agricultural Sciences |
| language | swe Inglés |
| publishDate | 2008 |
| publishDateSort | 2008 |
| publisher | SLU/Dept. of Forest Resource Management |
| publisherStr | SLU/Dept. of Forest Resource Management |
| record_format | eprints |
| spelling | RepoSLU122292017-11-02T09:05:07Z Identifiering av gallringsbehov med hjälp av flygburen laserskanning Prioritizing thinning operations using airborne laser scanning Halvarsson, Joakim LIDAR laser gallring täthetsindex höjdkvadratsumma Thinning is one of the most important silvicultural activities in middle aged forests, partly to minimize damages and partly because it leads to earlier harvesting revenues and increases the value of the remaining trees. There are many different indices used to describe stand density and thereby the need of thinning in forests today. Studies have shown that forest variables can be estimated with high accuracy using airborne laser scanning and it is likely that the method could also be used to estimate forest density indices. In this study, the possibility of using the forest density index ΣH² to determine and map priorities for thinning operations for a forest area was examined. The objective was to test how different variables derived from airborne laser scanning data can be used to estimate the index ΣH². Tests were performed with low and high density laser data (densities of 0.8 and 10 measurements/m²). In field, ΣH² was measured on 30 field plots with a size of 30x30m. The field plots were allocated in forests with a variety in structure, height, age, and tree species mixture. Different thinning indices that correlated to field measured ΣH² were examined. These indices were computed directly from raw laser data, from raw laser data converted to a raster with different cell sizes, and using single tree detection. The results show that ΣH² describes the need of thinning in a relatively good way for forests with a variety in structure, height, age and tree species mixture even though there are some exceptions. The method to convert raw laser data to a raster with 6m cell size gave a thinning index with the highest correlation to ΣH². The method worked well for both low density (0.8 measurements/m2) and high density (10 measurements/m2) laser data with adjusted R² values of 89.7% and 90.8%, respectively. The relative RMSE was 14.3% and 13.5% for low and high resolution laser data, respectively. Gallring är en av de viktigaste skogsvårdsåtgärderna i medelålders skog. Idag finns många olika täthetsindex vilka beskriver tätheten och därmed gallringsbehovet i skog. Studier har visat att skogliga variabler kan skattas med god noggrannhet med hjälp av flygburen laserskanning och tekniken kan därmed tänkas vara användbar även för att skatta täthetsindex. I denna studie undersöktes hur väl täthetsindexet ΣH² beskriver gallringsbehov samt hur det kan skattas ur laserdata med en punkttäthet på 0,8 respektive 10 punkter/m². ΣH² mättes i fält på 30st 30x30m koordinatsatta provytor i skog med varierande struktur, höjd, ålder och trädslagsblandning. Olika sätt att beräkna variabler från laserdata som korrelerar mot fältmätt ΣH² provades. Resultatet visar att ΣH² beskriver gallringsbehovet på ett relativt bra sätt för skog med varierande struktur, höjd, ålder och trädslagsblandning även om vissa undantag finns. Metoden att omvandla rådata till ett raster med 6m rastercellstorlek gav ett gallringsindex med högst korrelation mot ΣH². Metoden fungerade bra för både glest och tätt laserdata och justerad R² för korrelationen blev 89,7% för glest laserdata samt 90,8% för tätt laserdata. Då gallringsindexen omräknades till skattad ΣH² visade sig skattningarna ge ett relativt RMSE på 14,3% samt 13,5% för glest respektive tätt laserdata. SLU/Dept. of Forest Resource Management 2008 L3 swe eng https://stud.epsilon.slu.se/12229/ |
| spellingShingle | LIDAR laser gallring täthetsindex höjdkvadratsumma Halvarsson, Joakim Identifiering av gallringsbehov med hjälp av flygburen laserskanning |
| title | Identifiering av gallringsbehov med hjälp av flygburen laserskanning |
| title_full | Identifiering av gallringsbehov med hjälp av flygburen laserskanning |
| title_fullStr | Identifiering av gallringsbehov med hjälp av flygburen laserskanning |
| title_full_unstemmed | Identifiering av gallringsbehov med hjälp av flygburen laserskanning |
| title_short | Identifiering av gallringsbehov med hjälp av flygburen laserskanning |
| title_sort | identifiering av gallringsbehov med hjälp av flygburen laserskanning |
| topic | LIDAR laser gallring täthetsindex höjdkvadratsumma |