|
|
Analysis of important grass species distribution in the Krkonoše Mts. tundra using remote sensing
Ježek, Vít ; Kupková, Lucie (advisor) ; Červená, Lucie (referee)
Analysis of important grass species distribution in the Krkonoše Mts. tundra using remote sensing Abstract The aim of this thesis was to test the application of maximum likelihood classification, Random forest, Support vector machine and object-oriented classification with the Support vector machine classifier on selected areas in the Krkonoše Mts. relict arctic-alpine tundra for the purpose of mapping the distribution of vegetation with a focus on conservation-important grass species. The research used pre-processed multitemporal hyperspectral data and multispectral data from UAS with a spatial resolution of 0.03 m and 0.06 m and hyperspectral aerial data with a spatial resolution of 0.6 m together with training and validation data collected by botanists directly from the fields using GPS (all data are from 2019-2021). The classifications achieved excellent results. The best overall accuracies were achieved by the object-oriented classification, with accuracies ranging between 80-95 %. Similarly, good results were also achieved by pixel methods - Random forest and Support vector machine (highest overall accuracy 94 %). Of the important grass species, Calamagrostis villosa (producer's accuracy 99.73 %, user's accuracy 99.95 %) and Deschampsia cespitosa (producer's accuracy 99.98 %, user's accuracy 99.33 %)...
|
|
|
Analysis of tundra vegetation developement using a time series of ortoimages in the Krkonoše Mountains
Pajmová, Petra ; Kupková, Lucie (advisor) ; Potůčková, Markéta (referee)
Analysis of tundra vegetation developement using a time series of ortoimages in the Krkonoše Mountains Abstract The aim of this study is to analyse changes in arctic-alpine tundra vegetation in the Krkonoše Mountains using archival and current aerial imagery with red, green and blue bands and spatial resolution of 0.5 m. Three small areas of interest (cca 100 100 m) with different types of vegetation and a one larger area of the eastern tundra were studied. Several classification methods (Maximum likelihood classification, Random forest and object-based classification) were tested to obtain the best classification results. For more detailed analysis of grass species development, unsupervised classification and extended time series (5 orthoimages) were used for the area of Bílá louka. Classification were executed in softwares ENVI 5.5 and R 4.2.1. The highest overall accuracy of the 2020 image classifications were over 70% in all study areas, in some cases over 80%. With the exception of the Luční hora area (58%), the best overall accuracies for 2004 image were above 65%. After comparing classification results between years 2004 and 2020, a possible development trend was revealed. But due to low accuracy of the 2004 data classifications, this cannot be reliably demonstrated. Key words: classification,...
|
|
|
Determination of chlorophyll content in selected grass communities of Krkonoše tundra based on field spectroscopy and aerial hyperspectral data
Pinlová, Gabriela ; Červená, Lucie (advisor) ; Lhotáková, Zuzana (referee)
The thesis focuses on the determination of chlorophyll content from hyperspectral data in grass communities in the arctic-alpine tundra in the Krkonoše Mountains, namely Nardus stricta, Molinia caerulea, Calamagrostis villosa, and Deschampsia cespitosa. Leaf chlorophyll content (LCC) was measured using two methods - spectrophotometric destructive determination in the laboratory, and the LCC assessed non-destructively by fluorescence portable chlorophyll meter CCM-300. Leaf area index (LAI) values for canopy chlorophyll content (CCC) retrieval were also acquired by destructive biomass sampling and indirectly using LAI-2200C. Relationships were established between the LCCs, LAI, CCCs, and vegetation indices (VI) calculated from respective spectra, i.e. leaf level spectra acquired with contact probe coupled with an ASD FieldSpec4 Wide-Res spectroradiometer, canopy level spectra measured by the spectroradiometer and extracted from hyperspectral images (HSI) acquired by Headwall Nano- Hyperspec® mounted on the DJI Matrice 600 Pro drone. Chlorophyll content maps were created based on the results of multiple stepwise linear regression applied to HSI. For the model, derived from the non-destructive data sampling and used for the LCC map, a RMSE of 66.55 mg/m2 was achieved. Keywords: leaf chlorophyll...
|
|
|
The influence of spectral resolution on land cover classification in Krkonoše Mts. tundra
Palúchová, Miroslava ; Červená, Lucie (advisor) ; Kupková, Lucie (referee)
The influence of spectral resolution on land cover classification in Krkonoše Mts. tundra Abstract The aim of this diploma thesis was to specify the spectral resolution requirements for classification and to identify the most important spectral bands to discriminate classes of the predefined legend. Aerial hyperspectral data acquired by AisaDUAL sensor were used. The method applied for the selection of the important bands was discriminant analysis performed in IBM SPSS Statistics. The most discriminative bands were found in intervals 1500-1750 nm (beginning of SWIR), 1100- 1300 nm (longer wavelengths of NIR), 670-760 (red-edge) and 500-600 nm (green light). The classification of the selected bands was realized in ENVI 5.4 using the Support Vector Machine classifier, achieving overall accuracy of 80,54 %, Kappa coefficient 0,7755. The suitability of available satellite data for the classification of tundra vegetation in Krkonoše mountains based on spectral resolution was evaluated as well. Keywords: tundra, Krkonoše, classification, spectral resolution, class separability, discriminant analysis, hyperspectral data
|
|
|
Laboratory spectroscopy for selected Krkonoše Mts. tundra vegetation species
Tomcová, Jana ; Červená, Lucie (advisor) ; Lhotáková, Zuzana (referee)
Laboratory spectroscopy for selected Krkonoše Mts. tundra vegetation species The diploma thesis is focused on testing the methodologies of measuring the reflectance of grasses from the tundra of Krkonoše Mountains (Nardus stricta, Molinia caerulea, Calamagrostis villosa). The spectoradiometer ASD FieldSpec 4 Wide-Res with added contact probe ASD Plant Probe is used for measurements. Since it is not common to measure such narrow leaves that do not cover the whole FOV, the thesis is looking for methodologies that are the most repeatable and influenced by a minimum of errors. Factors influencing the measurement results are also monitored. Furthermore, the differentiation of the studied species is observed based on their spectral properties. Based on the measured data the medians and standard deviations are calculated and compared among each other. An analysis of variance (ANOVA) is used to determine the bands where the influence of individual factors is more apparent and where the individual grasses are distinguishable. As the most suitable methodologies for measuring grasses depend on the grasses structures and properties, the best methodology is different for each of selected species. The two layer leaf measurement is most suitable for the Nardus stricta, the measurement of the abaxial sides of leaves fits...
|
|
|
Laboratory spectroscopy for selected Krkonoše Mts. tundra vegetation species
Tomcová, Jana ; Červená, Lucie (advisor) ; Lhotáková, Zuzana (referee)
Laboratory spectroscopy for selected Krkonoše Mts. tundra vegetation species The diploma thesis is focused on testing the methodologies of measuring the reflectance of grasses from the tundra of Krkonoše Mountains (Nardus stricta, Molinia caerulea, Calamagrostis villosa). The spectoradiometer ASD FieldSpec 4 Wide-Res with added contact probe ASD Plant Probe is used for measurements. Since it is not common to measure such narrow leaves that do not cover the whole FOV, the thesis is looking for methodologies that are the most repeatable and influenced by a minimum of errors. Factors influencing the measurement results are also monitored. Furthermore, the differentiation of the studied species is observed based on their spectral properties. Based on the measured data the medians and standard deviations are calculated and compared among each other. An analysis of variance (ANOVA) is used to determine the bands where the influence of individual factors is more apparent and where the individual grasses are distinguishable. As the most suitable methodologies for measuring grasses depend on the grasses structures and properties, the best methodology is different for each of selected species. The two layer leaf measurement is most suitable for the Nardus stricta, the measurement of the abaxial sides of leaves fits...
|
|
| |
|
|
The influence of spectral resolution on land cover classification in Krkonoše Mts. tundra
Palúchová, Miroslava ; Červená, Lucie (advisor) ; Kupková, Lucie (referee)
The influence of spectral resolution on land cover classification in Krkonoše Mts. tundra Abstract The aim of this diploma thesis was to specify the spectral resolution requirements for classification and to identify the most important spectral bands to discriminate classes of the predefined legend. Aerial hyperspectral data acquired by AisaDUAL sensor were used. The method applied for the selection of the important bands was discriminant analysis performed in IBM SPSS Statistics. The most discriminative bands were found in intervals 1500-1750 nm (beginning of SWIR), 1100- 1300 nm (longer wavelengths of NIR), 670-760 (red-edge) and 500-600 nm (green light). The classification of the selected bands was realized in ENVI 5.4 using the Support Vector Machine classifier, achieving overall accuracy of 80,54 %, Kappa coefficient 0,7755. The suitability of available satellite data for the classification of tundra vegetation in Krkonoše mountains based on spectral resolution was evaluated as well. Keywords: tundra, Krkonoše, classification, spectral resolution, class separability, discriminant analysis, hyperspectral data
|
|
|
Changes of land cover above the tree line in Krkonoše National Park based on Landsat data
Vyvialová, Linda ; Červená, Lucie (advisor) ; Suchá, Renáta (referee)
CHANGES OF LAND COVER ABOVE THE TREE LINE IN KRKONOŠE NATIONAL PARK BASED ON LANDSAT DATA Abstract This study evaluates land cover changes in the area above the tree line in Krkonoše Mts. National Park. The stress is put on the changes of Pinus mugo shrub. Two methods of change detection based on Landsat data in four time horizons from the eighties up to now were tested (years 1984, 1992, 2002 and 2013). The first method was classification of scenes with the Maximum Likelihood Classifier individually and evaluation of changes based on their overlay. Overall accuracies from the oldest scene were 86,04 %, 88,44 %, 86,91 % and 86,43 %. The second method evaluates detection of change above one dataset that consists of scenes for all the time horizons. Overall accuracies were from the oldest 86.63 %, 88.64 % and 86.11 %. The second method was more appropriate for this study of land cover changes. On the map results there can be seen thinning of Pinus mugo scrub (between the years 2002 and 2013, 1984 and 1992) as well as its natural thickening and spreading. Key words KRNAP, tundra, land cover, change detection, Landsat
|
|
|
Classification of selected vegetation land cover categories in the Krkonoše Mts. tundra from Sentinel-2A imagery using multi-temporal data
Roubalová, Markéta ; Kupková, Lucie (advisor) ; Suchá, Renáta (referee)
Classification of selected vegetation land cover categories in the Krkonoše Mts. Tundra from Sentinel-2A imagery using multitemporal data Abstract The aim of this thesis was to evaluate the possibilities of multi-temporal approach to improve classification accuracy of vegetation cover in eastern tundra in the Krkonoše Mts. National Park. Sentinel-2A imagery - 10 spectral bands with spatial resolution 10 and 20 m - was used. The classification legend was created by a botanist of the national park. Maximum likelihood classification for 11 categories of vegetation land cover was executed in software ENVI 5.3. The overall accuracy of the best classification result was 53,4 % which is similar result as in the case of single image classification (overall accuracy was 51,2 %). Key words: multi-temporal classification, vegetation, spectral features, Sentinel-2A, tundra, The Krkonoše Mts. National Park
|
guest ::
