"PhD Position: Environmental Assessments using Opportunistic Sample Data"

Are you passionate about spatial analysis and statistics of environmental properties, and are you not afraid of a challenge? Then this vacancy is just right for you!

Your research will contribute to effective environmental monitoring and mapping. When we have control over our spatiotemporal sampling of observational data locations, we first define the population. Next, we sample population units for (1) design-based estimation of population parameters, such as the mean and standard deviation of some target property or for (2) model-based prediction (i.e., mapping) of environmental properties such as forest biomass.

However, sampling and measuring at new locations can be very expensive. So we often have to rely on previously acquired environmental observations, which may be irregularly spread over the study area and fail to represent a substantial part of the target population. Such “opportunistic” data pose significant challenges for making valid inferences about population-level environmental metrics such as soil properties, biomass stocks, or map accuracy metrics. If the data is used for mapping environmental properties, the fitted model performs poorly on minority classes or parts of the map cannot be trusted because they correspond to areas not covered by the sample. The issue has received significant attention in recent literature, but with several ad hoc solutions proposed it still lacks a thorough theoretical foundation.

Your duties and responsibilities include:

• Synthesize theoretically grounded methods for inference and spatiotemporal prediction using opportunistic samples, and evaluate them alongside newly developed approaches.
• Integrating methods such as case weighting, anomaly detection, and model-based prediction (e.g., geostatistics and machine learning), using auxiliary geospatial or remotely sensed data.
• Quantifying uncertainty and correcting for spatial and sampling biases inherent in environmental observation systems.
• Target environmental properties such as above ground forest biomass, soil organic carbon and areas occupied by specific land cover types.
• Publishing your research in peer-reviewed scientific journal papers and presenting your results at conferences.

You will work here: The research is embedded within the Laboratory of Geo-Information Science and Remote Sensing (GRS), which is led by Prof. Sytze de Bruin. You will be co-supervised by Dr Nandika Tsendbazar and Dr Marc Russwurm.

Requirements: We seek a highly motivated PhD candidate having: a master’s degree in Geo-Information Science, Physical Geography, Remote Sensing, Applied Mathematics, Statistics or a closely related field; a strong methodological background covering spatial sampling, design-based estimation, model-based prediction (geostatistics, machine learning) and uncertainty assessment; proficiency in high-level programming languages such as Python or R; excellent scientific writing skills in English; a strong interest in environmental characterization; a proactive, inquisitive, enthusiastic, and creative mind-set. For this position your command of the English language is expected to be at C1 level.

Conditions of employment: Wageningen University & Research offers excellent terms of employment. A few highlights from our Collective Labour Agreement include: Partially paid parental leave; working hours that can be discussed and arranged so that they allow for the best possible work-life balance; there is a strong focus on vitality and you can make use of the sports facilities available on campus for a small fee; a fixed year-end bonus of 8.3%; excellent pension scheme. In addition to these first-rate employee benefits, you will receive a fully funded PhD position and you will be offered a course program tailored to your needs and the research team. The gross salary for the first year is € 2.901,- per month rising to € 3.707,- in the fourth year in according to the Collective Labour Agreements for Dutch Universities (CAO-NU) (scale P). This is based on a full-time working week of 38 hours. We offer a temporary contract for 18 months which will be extended for the duration of the project if you perform well.