LGRB-BW BOPA: Near-surface clay content on arable land in Baden-Württemberg
Description
This dataset shows percentages of soil surface clay content modelled using Sentinel-2 time series from 2018, 2019, 2020 and 2021 and a Random Forest (RF) approach.
The sound content forecasts refer exclusively to agricultural land in Baden-Württemberg.
The RF modeling was carried out using the statistical software R. The reference data required for modelling consist of information derived from classical soil mapping (Landesamt für Geology, Raw Materials and Mining, LGRB), permanent observation areas (Landesamt für Umwelt Baden Württemberg, LUBW), the European LUCAS database (ESDA) and current field campaigns carried out by both the LGRB and the Federal Institute for Geosciences and Raw Materials (BGR). A total of 455 laboratory measurements on the clay content were available for the training, all of which come from grain size analyses using pipette method.
Applied to an independent sample of approximately 30 % of the total sample size, a certainness measure (R²) of 0.18.
The spatial ground resolution of the grid data is 10 m per pixel.
This dataset was carried out within the framework of the project BopaBW, funded by DLR and led by the State Office of Geology, Raw Materials and Mining (LGRB) (funding code:
50EW1703A) in cooperation with the Federal Institute for Geosciences and Raw Materials (BGR) and the Geoforschungszentrum Potsdam (GfZ).
Resources
| Name |
Format |
Description |
Link |
|
0 |
|
https://services.lgrb-bw.de/ms/lgrb_bopa?REQUEST=GetCapabilities&SERVICE=WMS&VERSION=1.3.0 |
Tags
- fernerkundung
- bodeneigenschaftenkarte
- maschinelles-lernen