The diploma thesis is focused on the application of the Geographically Weighted Regression (GWR) in forestry models. This is a prospective method for coping with spatially heterogeneous data. In forestry, this method has been used previously in small areas with good results, but in this diploma thesis it is applied to a bigger area in the Region of Murcia, Spain. Main goal of the thesis is to evaluate GWR for developing of large scale height-diameter model based on data of National Forest Inventory of Spain. Final model is compared with local height-diameter model on validation plots. The obtained results are very different according to the level of input data and GWR calibration type. The best result is obtained with individual tree data and with fixed kernel calibration. In order to improve quality of GWR calibration several suggestion were made, such as change weighted functions. GWR method is highly promising because in the case of need a particular model for some large area, there is possible to make sufficiently precise model at any point of the area of interest without need of any additional measurements in particular forest stand.
| Identifer | oai:union.ndltd.org:nusl.cz/oai:invenio.nusl.cz:190857 |
| Date | January 2015 |
| Creators | Quirós-Segovia, María |
| Source Sets | Czech ETDs |
| Language | English |
| Detected Language | English |
| Type | info:eu-repo/semantics/masterThesis |
| Rights | info:eu-repo/semantics/restrictedAccess |
Page generated in 0.011 seconds