Special issue Nutrient Cycling in Agroecosystems Journal

Context

Soil is the largest terrestrial sink of carbon, and therefore a major opportunity for the mitigation greenhouse gases emissions. While removing carbon from the atmosphere, sequestering more carbon in soil can lead to a range of benefits, such as improved fertility. This places agricultural soils, in particular, at the nexus of two major global challenges: food security and climate change. Consequently, there has been an increasing interest in novel technologies for measuring, mapping and monitoring of soil carbon and soil nutrients. But, how to organize sampling and monitoring of soil carbon and nutrients to achieve objective answers to key soil carbon management questions? What is happening with nutrients over large areas and over longer time scales? Which nutrients are in deficit and which might lead to supply risks? How significant is soil pollution and soil degradation and how is it related to soil fertility? Can machine learning frameworks (e.g. Tensor-Flow, H2O), data mining, High Performance Computing, and proximal and remote soil sensing technologies be combined to boost accuracy of carbon and nutrient maps without significant additional costs? What are the optimal statistical frameworks and best practices for producing reliable soil carbon and nutrient maps? How does the use of spatial soil information helps improve the value of soil resources, and support its communication to policy makers? How will such technologies impact food production and land management in 20 to 30 years? In which directions will technology and applications develop in the next 10 years?

This special issue invites researchers in the field of soil science, agronomy, ecology, data science and pedometrics to submit original research work connected with development and application of novel technologies and methods that can help answer some of the questions above. Our special interest is in using novel technologies such as soil sensing and image recognition, automated sensor networks, unmanned aerial vehicles (UAVs) and publicly available remote sensing products (such as NASA’s Landsat 7 & 8 and ASTER missions, ESA’s Sentinel 2 and other Copernicus land products, JAXA’s Advanced Land Observing Satellite ALOS, LiDAR, TanDEMx and similar missions), in combination with statistical learning, data mining and high performance computing.

Win Free registration!

First five (5) accepted papers for this special issue will be awarded a waived registration fees (names to be chosen by the contact author of the accepted paper) at the next OpenGeoHub Summer School 2019 (former GEOSTAT Summer School).

Scope

Specific topics include but are not limited to:

• Direct-to-digital soil mapping, from soil sensing to producing dynamic maps of soil carbon and nutrients,
• Integration of regional and global soil data, incorporation of scale into soil carbon and nutrient mapping,
• Ensemble methods / model averages for operational mapping,
• Estimation of uncertainty of spatial predictions in space and spacetime,
• Combination of expert-based soil evaluation and numeric approaches to soil mapping including soil sensing,
• Development of methods for predicting soil carbon and nutrients in space and spacetime (deriving soil carbon loss and gain),
• Development of sampling schemes for spatial monitoring of changes in soil carbon and nutrients possibly accounting for seasonal variation in the target soil property,
• Interdisciplinary developments spanning fields of pedometrics, econometrics and environmetrics,

Submission requirements

We require that all articles submitted for this issue come with documented computational steps (code) and/or data processing tutorials which are available publicly (github, Rpubs, Launchpad, Bitbucket or similar) and can be used to reproduce at least 2/3rd of key results (key tabular and graphical results). We also advise that the submitted works, as much as possible, involve using or distributing Open Data (meaning that the input and/or output data is registered under some of the Open Data licenses). Authors have full freedom to specify which of the key results are provided as reproducible research materials.

Submissions are now closed.

Reproducible research

Reproducibility check — conditions and workflows:

1. At least 2/3rd of key results (listed in the cover letter by the authors) must pass the reproducibility test. Key results should comprise between 5–12 items. Reproducibility test will be organized by the special issue editors who will also comments on directly on the code submitted together with the article.
2. Key results (including tables and graphs) that do not pass the reproducibility check will be returned to authors.
3. For the final submission, all input data, processing steps (the code) and final manuscript must match exactly before the paper can be accepted for publication (see below). This corresponds to Level 2 or 3 standards of the RR as defined in Nosek et al. (2015).