|
|
Estimation of winter wheat yield using machine learning from airborne hyperspectral data
Švik, Marian ; Pikl, Miroslav ; Janoutová, Růžena ; Veselá, Barbora ; Slezák, Lukáš ; Klem, Karel ; Homolová, Lucie
Methods based on optical remote sensing allow nowadays to assess crop conditions over larger areas. The assessment of crop conditions and potential estimation of crop yields in the early growth\nstages can help farmers to better target their management practice such as application of fertilizers. In this study we analysed airborne hyperspectral images acquired several times during the growing season over two experimental sites in the Czech Republic (Ivanovice and Lukavec). The field experiments on winter wheat included 12 levels of fertilisation (combination of organic and mineral fertilisers). Such an experiment design and the possibility of combining the data from two sites together increased the variability in our wheat yield dataset, which varied between 2.8 and 10.0 t/ha. Further, we used a machine learning method – namely gaussian process regression from the ARTMO toolbox to train two variants of models: a) combining the spectral data from both sites and from the multiple acquisition days and b) combining the spectral data from both sites for individual acquisition days.The results showed that it was feasible to predict wheat yield already at the beginning of April with R2 > 0.85. This promising result, however, requires more thorough validation and therefore we plan to include more data from other sites in the next steps.
|
|
|
Assessment of ecosystem functions of urban greenery
Zemek, František ; Pikl, Miroslav ; Pavelka, Marian ; Zapletal, Miloš ; Kokrda, Lukas ; Novotný, Jan ; Duchková, Helena ; Vačkářová, Davina ; Foltýnová, Lenka ; Cudlín, Pavel ; Kolařík, J. ; Holešová, H. ; Mácová, Kateřina ; Chalupová, O. ; Vojáčková, B. ; Mikulášek, J.
The methodology brings about new and practical approaches in assessment of five ecosystem functions and services of town greenery. It covers four regulation functions (microclimate regulation, carbon sequestration, pollution and noise reduction) and one information function (aesthetic and cultural-recreation). The calculus is based on a single greenery object evaluation using simplified biophysical and statistic models, however complying with all natural principles. The models utilize data standardly available in urban greenery passports and those which are accessible via map servers, e.g. meteorological situation and sociodemographic conditions.
|
|
| |
|
|
Assessment of the Estimates of Sun-induced Fluorescence in large masses of Vegetation
Rodriguez-Moreno, Fernando ; Zemek, František ; Pikl, Miroslav
The remote estimation of the sun-induced fluorescence has been successful in fields in which, in addition to the measurements for the verification, there is also a detailed characterization\nof the atmosphere, besides being an optimal environment, flat surface without interferences. Revealing the fluorescence is a process of precision because the minimum error tolerated forces each element of the chain to fulfill its function exactly and consistently. The detailed monitoring of a flight line is not feasible, hence the uncertainty in this process and the reason for the series of publications with the new estimation of fluorescence.
|
|
| |
|
|
Assessment of the sequestration capacity of vegetation by remote sensing methods in areas of reclaimed mining dumps
PIKL, Miroslav
The study aims at estimation and mapping the amount of carbon allocated in above ground biomass of wood and in organo-mineral soil horizon at sites where reclamation and spontaneous succession took place on spoil heaps after coal mining. Several categories of data have been used to meet the objectives, namely ground field measurements, laboratory analyses of soil samples, airborne hyperspectral data from VNIR region, and airborne LiDAR scanning data. The digital imagery analysis, GIS modeling and multivariation statistical methods were applied in data assessment. The results show that there is a 7 600 tons of carbon allocated in above ground wood biomass in the area of 209 ha, and 8 100?12 200 tons in the soil A horizon in the region of the same size. The results proofed: 1/ statistically significant negative relationships (p < 0,01) between slope and amount of soil carbon, where higher negative correlation was for broad leaved species; 2/ statistically significant difference (p < 0,05) between amount of soil carbon under broad leaved and needle classes and under different species, the highest between soils under Alnus sp. and Pinus sp.; 3/ statistically significant relationships (p < 0,05) between the amount of carbon allocated in the aboveground wood biomass and that in the soil A horizon under the needle leaved class and under the spontaneous wood vegetation.
|
|
|
Methods for identification of vulnerable zones for nitrogen leaching in arable land
Zemek, František ; Pikl, Miroslav ; Holub, Petr ; Klem, Karel ; Záhora, J. ; Vavříková, J.
The methodology demonstrates on the case study of a catchment (cca 200 km2 in size) how to identify the areas with high risk of nitrogen (N) leaching to underground water in arable land. It uses synergy between “conservative” characteristics of the catchment, such as orographic features, bedrock and soil hydrological properties, and the dynamic ones, which are represented by vegetation properties. The latter are identified through the imaging spectroscopy of wheat canopy obtained from airborne hyperspectral and/or Sentinel-2 multispectral data.
|
|
|
A methodology for assessing the impact of cattle grazing on soil properties, water quality and quantity, plant and invertebrates diversity
Fučík, P. ; Zemek, František ; Hakrová, P. ; Svobodová, M. ; Zajíček, A. ; Šlachta, M. ; Pikl, Miroslav ; Duffková, R. ; Mrkvička, J. ; Bystřický, V. ; Procházka, J. ; Skalický, M. ; Holubík, O. ; Moravcová, J. ; Novotný, Jan ; Skalická, J. ; Peterková, J. ; Musil, M. ; Šantrůček, J. ; Matoušková, V. ; Brom, J. ; Hanuš, Jan ; Novotná, K. ; Huislová, P.
This methodology introduces the approaches for assessment of soil, water, plant, forage and invertebrates characteristics of pastures as influenced by cattle grazing. These parameters are necessary to observe and evaluate in order to set the optimal grazing management of pastures for support of the sustainable land use and for mitigating the negative impacts of agriculture on the environment. Moreover, the proper management of pastures enables to improve livestock welfare, market challenges of foods, and strenghten and diversify various activities in rural areas. For the assessment of grazing-induced changes of aforementioned parameters, the focus is oriented towards different pasture geomorphological / hillslope zones, where grazing may exhibit in a distinct manner. The approaches are postponed based on literature survey as well as on knowledge acquired during a multi-branch research project. The methodology shows the principles for monitoring and evaluation of parameters regarding soil physics and chemistry, water quality and quantity, plant diversity and forage quality and ecology of selected invertebrate species. Beside this, approaches for application and interpretation of remote sensing methods and tools for detecting the changes in plant and soil pasture characteristics are introduced and discussed. The methodics is targeted on one hand for specialist from state administration offices (Ministry of Agriculture, Ministry of the Environment) as a background for evaluation of the effectiveness of subsidy programmes focused on grazing management. On the other hand, the approaches described in the methodics can serve the experts for an integrated assessment of grazing management on various pasture parameters or on the environment. The novelty of the methodics lies in the multi-level approach, which enables to assess the pasture parameters both individually or jointly, according to their mutual interrelatons and possible driving factors.
|
|
|
Hyperspectral image segmentation for estimation of biomass at reclaimed heaps
Pikl, Miroslav ; Zemek, František
This paper presents the preliminary results from a study that aims at estimation of above ground biomass and soil carbon content at reclaimed mining heaps in the Sokolov region. Two image segmentation methods are presented. We applied maximal likelihood (ML) and neural network (NN) classifi ers on airborne hyperspectral data. Th e objective of this part of the study was to prepare a land cover classifi cation of the region. Th e main focus was paid to discrimination of six classes with prevailing forest species cover. Th e classifi cation accuracy of the training sites was 93.75 % for NN and 79.12 % for ML respectively. But ML outperformed NN in overall classifi cation accuracy with 61.54 % compared to 40.9 % of NN. Th e more accurate results of the ML classifi er are probably infl uenced by properties of the training samples. Th e larger size of the training samples derived for ML enabled better representation of class histograms. Th e lower overall NN accuracy could result from high spatial resolution of HS data.
|
|
|
Alcohol Yield of Oat, Barley and Wheat Hydrolysates
PIKL, Miroslav
Major part of ethanol production for industrial use in Czech Republic is made from cereals. Effect of acid, alkali and enzymatic hydrolysis after thermopressure preparation takes part in the paper. The most promising alternative is enzymatic hydrolysis. Especially efficiency of different hydrolytic enzymes towards different species of cereals takes the main part of the simulations. Local farm products, oats, barley and wheat were prepaired by steam explosion in different pressure. After hydrolysis were detected the GE and after fermentation was observed the production of alcohol.
|
guest ::
RSS feed.