Three centrally edited nationwide soil maps were published in Hungary between 1953 and 1988. Each of these soil maps has advantages, but serious drawbacks as well. Authors’ hypothesis was that the drawbacks of the individual soil maps are correctable with the help of other soil maps and with ancillary data. Therefore, the oldest soil map was digitized and a study was conducted for the harmonization of data on a 266 km2 area at Keszthely (near Lake Balaton) by using the CHAID classification tree method. CORINE land cover database, digital map of surface geology, digital elevation model and derived slope categories were used as ancillary data.The seven source maps contained 7–38 categories. After the intersection of all seven maps, the resulting file contained more than 50,000 polygons and nearly 14,000 category combinations. A variable — showing the probability of the category combinations in relation to the expected areas — was calculated. This was the target variable for classification by the CHAID method, using categories of the seven original maps as independent variables.0.5% of the total area was grouped into 13 less probable classes, which represent the inaccuracies of the initial maps. 99.5% of the total area was classified into 19 classes and some of them were further subdivided on the basis of the geological map. These classes were interpreted as eight WRB soil categories. The final soil map had much better spatial resolution than any of the initial soil maps, non-soil categories were interpreted as soil categories and spatial accuracy was successfully corrected with the proposed method.
The concept of precision agriculture is straightforward at the scientific level but even basic goals are blurred at the level of everyday practice in the Hungarian crop production despite the fact that several elements of the new technology have already been applied. The industrial and the service sectors offer many products and services to the farmers but crop producers do not get enough support to choose between different alternatives. Agricultural higher education must deliver this support directly to the farmers and via the released young graduates. The price of agricultural land must be higher if well-organized data underpin the production potential of the fields. Accumulated database is a form of capital. It must be owned by the farmers but in a data-driven economy its sharing will generate value for both farmers and the society as a whole. We present a methodological approach in which simple models were applied to predict yield by using only those yield data which spatially coincide with the soil data and the remaining yield data and the models were used to test different sampling and interpolation approaches commonly applied in precision agriculture. Three strategies for composite sample collection and three interpolation methods were compared. Multiple regression models were developed to predict yields. R2 values were used to select among the applied methods.
<p>&#160;</p> <p>In the hilly south-western region of Hungary a 23 hectare, slightly sloping agricultural plot was surveyed with various proximal sensing techniques, basically to support its precision management. An UAV-based hyperspectral &#160;sensor (Cubert UHD185) provided the spectral characteristics (in the form of a hyperspectral ortho-mosaic) together with the Digital Elevation Model of the plot. Detailed geophysical measurements were also carried out. Electrical conductivity was surveyed by a GF Instruments CMD-Explorer and a CMD-Explorer Mini. Gamma-ray data (U, Th, K &#233;s Total Count) was collected using a GF Instruments Gamma Surveyor Vario 2048-channel geo-physical gamma-ray spectrometer. Magnetic field was measured with a Geometrics G-858 magnetometer. Soil samples were collected in a 50 x 50 meter regular grid at 86 locations from the upper 0-30 cm. In the laboratory pH(KCl), consistency, salt, CaCO<sub>3</sub>, SOM, NO<sub>2</sub>-NO<sub>3</sub>-N, P<sub>2</sub>O<sub>5</sub>, K<sub>2</sub>O and Na content were measured.</p> <p>Spatial distribution of primary soil properties was predicted using different combinations of predictor datasets and various machine learning methods and with two differing concepts.</p> <p>While airborne hyperspectral data and terrain derivatives are available in spatially exhaustive (quasi-continuous) image format, underground geophysical data are basically generated as densely sampled point-like observations. According to general approach, these geophysical data are &#8220;simply&#8221; rasterized and then also used in spatially exhaustive form. In the recent study, we tested an alternative approach by using dense geophysical measurements as spatially non-exhaustive ancillary information to support the digital mapping of primary soil properties. First, soil features were estimated purely by geophysical auxiliary data as predictors, basically, establishing geophysical pedotransfer functions. Then predictions were used to infer spatially dense, estimated soil property data at geophysical measurement locations, which are more numerous by orders. Finally, soil property maps were derived by using (i) dense, inferred soil estimates as observation data and (ii) high-resolution spectral and terrain-based predictors. The results were evaluated and compared to those, which were produced using rasterized geophysical data.</p> <p>Our paper will present the comparison results, the experienced similarities and dissimilarities and discuss the applicability of dense geophysical measurements as spatially non-exhaustive ancillary information to support the digital mapping of primary soil properties.</p> <p>&#160;</p> <p>Acknowledgment: Our research has been supported by the Hungarian National Research, Development and Innovation Office (NRDI; Grant No: K 131820).</p>
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.