Lucie Kupková

Mapping vegetation communities of the Karkonosze National Park using APEX hyperspectral data and Support Vector Machines

Abstract This research aims to discover the potential of hyperspectral remote sensing data for mapping mountain vegetation ecosystems. First, the importance of mountain ecosystems to the global system should be stressed due to mountainous ecosystems forming a very sensitive indicator of global climate change. Furthermore, a variety of biotic and abiotic factors influence the spatial distribution of vegetation in the mountains, producing a diverse mosaic leading to high biodiversity. The research area covers the Szrenica Mount region on the border between Poland and the Czech Republic - the most important part of the Western Karkonosze and one of the main areas in the Karkonosze National Park (M&B Reserve of the UNESCO). The APEX hyperspectral data that was classified in this study was acquired on 10th September 2012 by the German Aerospace Center (DLR) in the framework of the EUFAR HyMountEcos project. This airborne scanner is a 288-channel imaging spectrometer operating in the wavelength range 0.4-2.5 μm. For reference patterns of forest and non-forest vegetation, maps (provided by the Polish Karkonosze National Park) were chosen. Terrain recognition was based on field walks with a Trimble GeoXT GPS receiver. It allowed test and validation dominant polygons of 15 classes of vegetation communities to be selected, which were used in the Support Vector Machines (SVM) classification. The SVM classifier is a type of machine used for pattern recognition. The result is a post classification map with statistics (total, user, producer accuracies, kappa coefficient and error matrix). Assessment of the statistics shows that almost all the classes were properly recognised, excluding the fern community. The overall classification accuracy is 79.13% and the kappa coefficient is 0.77. This shows that hyperspectral images and remote sensing methods can be support tools for the identification of the dominant plant communities of mountain areas.

Classification of Tundra Vegetation in the Krkonoše Mts. National Park Using APEX, AISA Dual and Sentinel-2A Data

ABSTRACT The aim of this study was to evaluate and compare suitability of aerial hyperspectral data (AISA Dual and APEX sensors) and Sentinel-2A data for classification of tundra vegetation cover in the Krkonoše Mts. National Park. We compared classification results (accuracy, maps) of pixel-based (Maximum Likelihood, Suport Vector Machine and Neural Net) and object-based approaches. The best classification results (overall accuracy 84.3%, Kappa coefficient = 0.81) were achieved for AISA Dual data using per-pixel SVM classifier for 40 PCA bands. The best classification results of APEX though were only 1.7 percentage points lower. To get comparable results for Sentinel-2A classification legend had to be simplified. With the simplified legend the accuracy using MLC classifier reached 77.7%.

Landscape transition after the collapse of communism in Czechia

ABSTRACT The paper deals with mapping of landscape transition after the collapse of Communism in Czechia on national and local levels. Three maps demonstrate the main trends of landscape transition on the national level (by cadastral units) in the period 1990–2010. The Main Map shows the four most important Processes of landscape change (afforestation, grassing over, intensification, and urbanization). The second map demonstrates the proportion of area where any kind of land use change occurred (Index of change) and the third map (Extensification) indicates the shift to less intensive use of land (increase of forests and grasslands). Two main processes were mapped on the local level, that is, by parcels. The case of Jirny showed strong sububanization: fertile agricultural land has been turned into residential and commercial areas, roads; soil sealing was taking place. On the contrary, grassing over and afforestation was detected in Hošťka where arable land almost disappeared – it was either abandoned or replaced mainly by pastures between 1990 and 2010.

Comparison of Reflectance Measurements Acquired with a Contact Probe and an Integration Sphere: Implications for the Spectral Properties of Vegetation at a Leaf Level

Laboratory spectroscopy in visible and infrared regions is an important tool for studies dealing with plant ecophysiology and early recognition of plant stress due to changing environmental conditions. Leaf optical properties are typically acquired with a spectroradiometer coupled with an integration sphere (IS) in a laboratory or with a contact probe (CP), which has the advantage of operating flexibility and the provision of repetitive in-situ reflectance measurements. Experiments comparing reflectance spectra measured with different devices and device settings are rarely reported in literature. Thus, in our study we focused on a comparison of spectra collected with two ISs on identical samples ranging from a Spectralon and coloured papers as reference standards to vegetation samples with broadleaved (Nicotiana Rustica L.) and coniferous (Picea abies L. Karst.) leaf types. First, statistical measures such as mean absolute difference, median of differences, standard deviation and paired-sample t-test were applied in order to evaluate differences between collected reflectance values. The possibility of linear transformation between spectra was also tested. Moreover, correlation between normalised differential indexes (NDI) derived for each device and all combinations of wavelengths between 450 nm and 1800 nm were assessed. Finally, relationships between laboratory measured leaf compounds (total chlorophyll, carotenoids and water content), NDI and selected spectral indices often used in remote sensing were studied. The results showed differences between spectra acquired with different devices. While differences were negligible in the case of the Spectralon and they were possible to be modelled with a linear transformation in the case of coloured papers, the spectra collected with the CP and the ISs differed significantly in the case of vegetation samples. Regarding the spectral indices calculated from the reflectance data collected with the three devices, their mean values were in the range of the corresponding standard deviations in the case of broadleaved leaf type. Larger differences in optical leaf properties of spruce needles collected with the CP and ISs are implicated from the different measurement procedure due to needle-like leaf where shoots with spatially oriented needles were measured with the CP and individual needles with the IS. The study shows that a direct comparison between the spectra collected with two devices is not advisable as spectrally dependent offsets may likely exist. We propose that the future studies shall focus on standardisation of measurement procedures so that open access spectral libraries could serve as a reliable input for modelling of optical properties on a leaf level.

Analysis and expert assessment of the semantic similarity between land cover classes

Products of CORINE Land Cover (CLC), the National Land Cover Dataset (NLCD), the FAO/UNEP Land Cover Classification System (LCCS), etc. currently provide an important source of information used for the assessment of issues such as landscape change, landscape fragmentation and the planning of urbanization. Assuming that the data from these various databases are often used in searching for solutions to environmental problems, it is necessary to know which classes of different databases exist and to what extent they are similar, i.e. their possible compatibility and interchangeability. An expert assessment of the similarity between the CLC and NLCD 1992 nomenclatures is presented. Such a similarity assessment in comparison with the ‘geometric model’, the ‘feature model’ and the ‘network model’ is not frequently used. The results obtained show the similarity of assessments completed by four experts who marked the degree of similarity between the compared land cover classes by 1 (almost similar classes), 0.5 (partially similar classes) and 0 (not similar classes). Four experts agreed on assigning 1 in only three cases; 0.5 was given 33 times. A single expert assigned 0.5 a total of 17 times. Results confirmed that the CLC and NLCD nomenclatures are not very similar.

Laboratory and image spectroscopy for evaluating the biophysical state of meadow vegetation in the Krkonoše National Park

Abstract The paper deals with the evaluation of mountain meadow vegetation condition using in-situ measurements of the fraction of Accumulated Photosynthetically Active Radiation (fAPAR) and Leaf Area Index (LAI). The study analyses the relationship between these parameters and spectral properties of meadow vegetation and selected invasive species with the goal of finding out vegetation indices for the detection of fAPAR and LAI. The developed vegetation indices were applied on hyperspectral data from an APEX (Airborne Prism Experiment) sensor in the area of interest in the Krkonoše National Park. The results of index development on the level of the field data were quite good. The maximal sensitivity expressed by the coefficient of determination for LAI was R2 = 0.56 and R2 = 0.79 for fAPAR. However, the sensitivity of all the indices developed at the image level was quite low. The output values of in-situ measurements confirmed the condition of invasive species as better than that of the valuable original meadow vegetation, which is a serious problem for national park management.

CLASSIFICATION OF VEGETATION ABOVE THE TREE LINE IN THE KRKONOŠE MTS. NATIONAL PARK USING REMOTE SENSING MULTISPECTRAL DATA

This paper compares suitability of multispectral data with different spatial and spectral resolutions for classifications of vegetation above the tree line in the Krkonose Mts. National Park. Two legends were proposed: the detailed one with twelve classes, and simplified legend with eight classes. Aerial orthorectified images (orthoimages) with very high spatial resolution (12.5 cm) and four spectral bands have been examined using the object based classification. Satellite data WorldView-2 (WV-2) with high spatial resolution (2 metres) and eight spectral bands have been examined using object based classification and per-pixel classification. Per-pixel classification has been applied also to the freely available Landsat 8 data (spatial resolution 30 metres, seven spectral bands). Of the algorithms for per-pixel classification, the following classifiers were compared: maximum likelihood classification (MLC), support vector machine (SVM), and neural net (NN). The object based classification utilized the example-based approach and SVM algorithm (all available in ENVI 5.2). Both legends (simplified and detailed ones) show best results in the case of orthoimages (overall accuracy 83.56% and 71.96% respectively, Kappa coefficient 0.8 and 0.65 respectively). The WV-2 classification brought best results using the object based approach and simplified legend (68.4%); in the case of per-pixel classification it was the SVM method (RBF) and detailed legend (60.82%). Landsat data were best classified using the MLC (78.31%). Our research confirmed that Landsat data are sufficient to get a general overview of basic land cover classes above the tree line in the Krkonose Mts. National Park. Based on the comparison of the data with different spectral and spatial resolution we can however conclude that very high spatial resolution is the decisive feature that is essential to reach high overall classification accuracy in the detailed level.

Land Use Changes in Czechia 1845–2010

The history of land use changes on the Czech territory since the very beginning is outlined; each subchapter deals with one important historical period. The emergence of organized agriculture (Neolithic revolution) is seen as the first period when humans began to influence nature on a certain scale. For thousands of years, however, land use changes were largely limited to inhabited lowlands. The transition from wilderness towards largely agricultural landscape accelerated only during the German plantation (eleventh–fourteenth centuries) when many forests were cleared in the frontier. As a whole, however, changes were rather modest until the eighteenth century. Really important economic and social changes that fundamentally influenced land use patterns have been taking place since the eve of Industrial Revolution. In that time, agricultural society was being gradually transformed into the industrial one at the beginning of the 20th century. The second half of the nineteenth century brought general modernization; agricultural land and arable land expanded to maximum. Since the turn of nineteenth and twentieth centuries, however, reverse trends are recorded: decrease of agricultural land (due to more intensive farming) and gradual expansion of forests. Land use patterns during the twentieth century were much influenced by turbulent political events like Czechoslovak independence (1918), World War II (1939–1945), Communist coup d’etat (1948), and restoration of democratic conditions (1989). The Communist legacy included outdated technology and production-oriented agriculture that could not compete on the international markets. The post-Communist period brought restitution of confiscated property (including land) and return to market-oriented conditions. In the most recent period, the accession of Czechia to European Union (2004) has also had profound effects on land use changes.

Determination of lignin content in Norway spruce foliage using NIR spectroscopy and hyperspectral data

Contents of biochemical compounds such as chlorophyll, nitrogen, cellulose or lignin in foliage can be used as indicators of the actual tree physiological status and the tree previsible damage. Imaging spectroscopy has been often applied for estimation of relations between foliage spectral and biochemical properties of chlorophyll or nitrogen contents ([4], [10], [15]). But it can also be used for the estimation of other important biochemical compounds in foliage such as lignin and cellulose ([6]; [12]), tannin [14] or polyphenolic compounds [13].

Influence of Natural Conditions on Land Use

This chapter deals with the influence of natural conditions on land use patterns. It also examines the human impacts on land use. Basic overview of natural conditions in Czechia is outlined with special regard to geology, climate and soils. Geological conditions are seen as the key factors that form landscapes and influence the diversity of soils. Climate, of course, also has profound influence on regional farming patterns; very warm (VW) and warm climatic regions are best suitable for agriculture. The biggest part of the Czech territory is covered by moderately heavy soils. Soil types are crucial for the spatial distribution of forests, arable lands, and permanent grasslands. Climatic zones and soil types are shown in maps. Regional patterns of Czech agriculture are discussed and the so-called less-favoured areas (LFA; important for allocation of EU subsidies) are explained. The history of human impacts on land use patterns over the past two centuries (covered by this research) has three phases. First, important changes in agriculture were taking place (changing balance between extensive and intensive farming). Second, forests began to shrink as more agricultural land was needed; with the advance of intensive farming, however, this process was reversed (“forest transition”). Third, new technologies and pressures exerted by the modern society brought a significant rise of built-up land and “other” areas. The ways how recent trends influenced the natural environment are explained. Changing political climate, especially the collapse of Communism and reintroduction of market conditions, has had profound effects on land use. The same applies to mining that caused large-scale devastation in some areas. Conservation programmes that accelerated after 1990 are seen as a “return to nature”.