Browsing by Subject "vaakanäkyvyys"

Horizontal visibility determines how far a person can see without any obstacles blocking their line of sight in a horizontal direction. Visibility is in great importance in numerous fields of study and lines of work, such as urban planning, wildlife surveillance and conservation and in military. Manmade obstacles, such as buildings, bridges or other built structures, and environmental elements, like topographical variations or vegetation, can cause noticeable obstruction of visibility in varying magnitude. Airborne laser scanning (ALS) based on light detection and ranging (LiDAR) technology has been used successfully to model built structures and environmental elements, such as forest inventory attributes. This study aims to find out how well horizontal visibility in Finnish boreal forests can be modelled and predicted using a predictive model based on 5 points/m2 ALS point cloud dataset provided by National Land Survey of Finland. The research was done in two separate locations, Karilasanden and Sipoonkorpi National Park, located in Uusimaa region in southern Finland. The ALS point clouds were scanned in summertime 2019 and 2020 for the two research areas. In addition, a total of 60 field measurements of horizontal visibility were collected in summertime 2023. Modelling of horizontal visibility was done with an area-based-approach on the ALS point cloud. The most suitable forest structure variables were selected using a linear regression model based on the measured data at the research areas. The variables were used to predict horizontal visibility with a wall-to-wall method for the whole study sites. Finally, map visualizations were created by generalizing vector polygons of horizontal visibility predictions, and by combining viewshed and horizontal visibility prediction results. The results indicate that number of forest structure metrics based on ALS data can be used for horizontal visibility modelling, including canopy density, median and lower quantile percentages of height distributions and relative density of points from 0.5 to 2.2-meter height. In addition, the 5 points/m2 ALS dataset from National Land Survey of Finland is sufficient dataset for estimation of horizontal visibility in Finnish boreal forests when the accuracy of the output is around tens of meters. Understandable and fit for use map visualization can be produced from the prediction results in various formats. Modelling of Finnish forest structure and using structure metrics as variables for modelling horizontal visibility can be useful for example in city planning for planning the structure of urban greenery or in planning phases of military operations. The available dataset is sufficient for rougher scale modelling, but to achieve finer scale modelling results, increasing the density of the point cloud could be tried or the ALS dataset could be accompanied with other datasets. Also, the availability of the data is somewhat limited since the data is not open. The combining of ALS datasets with TLS datasets for higher point cloud density, usage of TLS dataset for verification purposes of this research's methods or for gathering additional field measurements rose as possible future research topics, as well as the possibility of modifying the point cloud dataset in order to examine the most important forest structure variables more accurately.