Sensor mapping of Amazonian Dark Earths in deforested croplands
Geoderma(2016)
Abstract
Amazonian Dark Earths (ADEs) are fertile soils for agricultural production as well as important archaeological resources for understanding the pre-Columbian past of the Neotropical lowland rainforest. ADEs are threatened by expanding land exploitation and there is a need to develop efficient approaches to soil mapping and analysis for documenting these soils. In this paper we assess the potential of satellite remote sensing and proximal soil sensing to map, predict and monitor ADEs in land affected by agro-industrial development. We use instruments based on portable x-ray fluorescence (PXRF) and electromagnetic induction (EMI) as well as high-resolution satellite data (Spot 6) for detailed soil surveys at a 10-ha ADE site now mainly used for soybean production on the Belterra Plateau, Pará, Brazil. We predict the regional occurrence of ADE in a c. 250km2 test area centred on the known ADE site São Francisco using satellite data. Multivariate adaptive regression splines models were parameterised for predictions of soil organic carbon (SOC), cation exchange capacity (CEC), phosphorus (P) and depth of the A horizon in ADEs from sensor data – both from individual sensors and in sensor combinations. Combining sensors gave the best validation results: the highest modelling efficiencies (E) were 0.70 (SOC), 0.88 (CEC) and 0.74 (for both P and A depth). The most powerful single proximal sensor outputs in the predictions were Sr from the PXRF data and magnetic susceptibility (MSa) as measured by the EMI instrument. In the regional satellite based model we located 17 previously unrecorded ADE sites >2ha. Ground control checks showed that 10 out of 11 sites were correctly classified. We conclude that these sensors are useful in studies of ADE in deforested cropland and provide new opportunities for detailed studies of the archaeological record.
MoreTranslated text
Key words
Terra preta,Remote sensing,Portable x-ray fluorescence,Electromagnetic induction,Soil classification,MARSplines
AI Read Science
Must-Reading Tree
Example
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined