|
|
Zpracování a interpretace výnosových map jako podklad pro agronomické rozhodování
Širůček, Petr
Diploma thesis was focused on data yield map processing, their interpretation and analysis of relations with another spatial data. Aims of the work were identification error values, statistical process yield data and with help statistical methods evaluate correlation between yield map and another spatial data. Methods were based on gathering yield data with combine harvester during two years in corn and malting barley in conditions of farm. With using GIS and predefined tolls were removed error values from yield data and created yield maps. Subsequently, yield maps were analyzed with vegetation index NDVI, digital elevation model (DEM), sensor measurements of soil electrical conductivity and historical yield maps. In results was achieved removing from 11,3 to 13,8 % yield points in average, increasing of average yield from 3,6 to 4,7 % and decreasing of variation coefficient between 7,0 and 8,6 % with comparing original yield data and depending on planting crops during data yield processing. In case of evaluation of relations between yield map and NDVI obtained from satellite remote sensing was achieved moderately correlation by corn (r = 0,44) and malting barley (r = 0,65) in average depending on times of imaging and vegetative phase of the crop. In the evaluation relations between yield maps and DEM and slope was achieved weak and very weak direct and indirect correlation. Likewise, by soil electrical conductivity and yield was achieved very weak to negligible correlation. Conversely, the highest correlation was achieved between yield map and historical yield map (r = 0,63) in average. Obtained information serve as starting point for making agronomy decisions in conditions of particular farm in case of replacement of creating yield maps by prediction of yield based on another spatial data.
|
guest ::
