Aarhus University Seal / Aarhus Universitets segl

Amélie Marie Beucher

Mapping clay and soil organic matter for fields without samples using EMI and aerial imagery

Publikation: KonferencebidragPosterForskning


Several studies have shown that the apparent electrical conductivity (ECa) has a high correlation with the clay content of the soil. At the same time, bare soil imagery is highly useful for predicting soil organic matter (SOM) contents. However, both the ECa and the soil color are subject to spatial and temporal variations, for example due to variations in soil moisture. Therefore, models using EC and bare soil imagery need local samples for calibration in order to provide accurate predictions.
As an alternative, we propose a method using soil maps at regional or national extent and soil observations from other fields. All fields have ECa measurements from a DUALEM sensor and soil brightness from aerial imagery. The method establishes generalized relationships between the measured soil properties and ECa measurements and soil brightness using the observations from other fields. It then adjusts the predictions from this relationship, using a national map of soil properties.
We tested the method for predictions of topsoil clay and SOM contents at five agricultural fields in Denmark, using cross-validation between the fields. The method predicted clay contents with a higher accuracy than the national map for fields with clay contents higher than 10%. The method also predicted SOM contents with a higher accuracy than the national map, although nonlinear relationships between SOM and soil brightness lead to errors in some areas. However, the method had a lower accuracy than models trained from local observations for all fields.
Our study shows that it is possible to use ECa measurements and bare soil imagery to improve upon predictions of soil properties from regional and national maps. The method is mostly useful for providing preliminary estimates of soil properties for planning sampling strategies and fieldwork.
Udgivelsesår3 jun. 2019
StatusUdgivet - 3 jun. 2019
BegivenhedPedometrics - University of Guelph, Guelph, Ontaria, Canada
Varighed: 2 jun. 20196 jun. 2019


LokationUniversity of Guelph
ByGuelph, Ontaria

Se relationer på Aarhus Universitet Citationsformater

ID: 180123932