PostDoc on Soil Monitoring

INFOSOL Unit is looking for a post doctoral fellowship applicant for a period of 18 months. Infosol unit is working on the settlement of the French soil Information system and especially of the soil monitoring network (RMQS).

Soil monitoring is essential for the early detection of changes in soil quality. Such early detection should enable the design and implementation of policy measures to protect and maintain the sustainable use of soil so that it continues to deliver goods and services. Large dadaset are today available for France on various contaminants, among which trace elements (TE) are the parameters that will be investigated in this study. A classical issue when dealing with the results of soil monitoring system to map TE content is to separate effects coming from natural background from those
being from an anthropogenic source. Another issue is also to separate short-scale variation which can occur due to isolated short-scale processes such as pollution from industrial sites or geogenic anomalies. It is therefore a challenge to fit a statistical model describing the spatial variation of a soil property across a nation since such a model must account for variation at each of these scales. Moreover, RMQS have undertaken a single sampling only, and thus it remains inventories at present. A major issue is to assess whether possible changes in TE contents are detectable by soil monitoring taking into account the uncertainties caused by spatial heterogeneity, sampling methods and analytical errors. At a regional or national scale, it is necessary to assess the effect of the number of sites and of inherent soil spatial variability on the detection of a change in TE contents.
The successful candidate will be responsible for fitting robust linear mixed model which divides the spatial variation of soil properties between geological-, backgroundand short-scale processes and for the French territory. The fit and validation of the models will be performed using the RMQS dataset. The model will use on an algorithm bases upon robust estimators and predictors for identifying the outliers. Finally, predictions may be used to assess the ability of national network to detect possible changes by invoking the Central Limit Theorem.
Full details are provided on the website:

Required skills and knowledge
Required knowledge includes the trace elements cycle for terrestrial systems and specifically the factors of TE stocks changes, for natural and cultivated systems. Good statistical and geostatistical (for the purpose of using robust statistic and mixed models) skills would be appreciated. Knowledge of some scripting languages such as R, python or sql languages would help as well. Knowledge of English and experience in writing scientific papers is required.


You may also like...